Journal of Automation, Mobile Robotics and Intelligent Systems, vol. 15, no. 2 (2021) by Ł-PIAP

WWW.JAMRIS.ORG • pISSN 1897-8649 (PRINT) / eISSN 2080-2145 (ONLINE) • VOLUME 15, N° 2, 2021 Indexed in SCOPUS

wowe skrót

Łukasiewicz – Industrial Research Institute for Automation and Measurements PIAP

Journal of Automation, Mobile Robotics and Intelligent Systems A peer-reviewed quarterly focusing on new achievements in the following fields: • Fundamentals of automation and robotics • Applied automatics • Mobile robots control • Distributed systems • Navigation • Mechatronic systems in robotics • Sensors and actuators • Data transmission • Biomechatronics • Mobile computing Editor-in-Chief

Typesetting

Janusz Kacprzyk (Polish Academy of Sciences, Łukasiewicz-PIAP, Poland)

PanDawer, www.pandawer.pl

Advisory Board

Webmaster

Dimitar Filev (Research & Advenced Engineering, Ford Motor Company, USA) Kaoru Hirota (Japan Society for the Promotion of Science, Beijing Office) Witold Pedrycz (ECERF, University of Alberta, Canada)

Piotr Ryszawa (Łukasiewicz-PIAP, Poland)

Editorial Office ŁUKASIEWICZ Research Network – Industrial Research Institute for Automation and Measurements PIAP Al. Jerozolimskie 202, 02-486 Warsaw, Poland (www.jamris.org) tel. +48-22-8740109, e-mail: office@jamris.org

Co-Editors Roman Szewczyk (Łukasiewicz-PIAP, Warsaw University of Technology, Poland) Oscar Castillo (Tijuana Institute of Technology, Mexico) Marek Zaremba (University of Quebec, Canada)

The reference version of the journal is e-version. Printed in 100 copies.

Executive Editor ´ Katarzyna Rzeplinska-Rykała, e-mail: office@jamris.org (Łukasiewicz-PIAP, Poland)

Articles are reviewed, excluding advertisements and descriptions of products.

Associate Editor

If in doubt about the proper edition of contributions, for copyright and reprint permissions please contact the Executive Editor.

´ (Poznan ´ University of Technology, Poland) Piotr Skrzypczynski

Statistical Editor ´ Małgorzata Kaliczynska (Łukasiewicz-PIAP, Poland)

Editorial Board:

Mark Last (Ben-Gurion University, Israel) Anthony Maciejewski (Colorado State University, USA) Krzysztof Malinowski (Warsaw University of Technology, Poland) Andrzej Masłowski (Warsaw University of Technology, Poland) Patricia Melin (Tijuana Institute of Technology, Mexico) Fazel Naghdy (University of Wollongong, Australia) Zbigniew Nahorski (Polish Academy of Sciences, Poland) Nadia Nedjah (State University of Rio de Janeiro, Brazil) Dmitry A. Novikov (Institute of Control Sciences, Russian Academy of Sciences, Russia) Duc Truong Pham (Birmingham University, UK) Lech Polkowski (University of Warmia and Mazury, Poland) Alain Pruski (University of Metz, France) Rita Ribeiro (UNINOVA, Instituto de Desenvolvimento de Novas Tecnologias, Portugal) Imre Rudas (Óbuda University, Hungary) Leszek Rutkowski (Czestochowa University of Technology, Poland) Alessandro Saffiotti (Örebro University, Sweden) Klaus Schilling (Julius-Maximilians-University Wuerzburg, Germany) Vassil Sgurev (Bulgarian Academy of Sciences, Department of Intelligent Systems, Bulgaria) Helena Szczerbicka (Leibniz Universität, Germany) Ryszard Tadeusiewicz (AGH University of Science and Technology, Poland) Stanisław Tarasiewicz (University of Laval, Canada) Piotr Tatjewski (Warsaw University of Technology, Poland) Rene Wamkeue (University of Quebec, Canada) Janusz Zalewski (Florida Gulf Coast University, USA) ´ Teresa Zielinska (Warsaw University of Technology, Poland)

Chairman – Janusz Kacprzyk (Polish Academy of Sciences, Łukasiewicz-PIAP, Poland) Plamen Angelov (Lancaster University, UK) Adam Borkowski (Polish Academy of Sciences, Poland) Wolfgang Borutzky (Fachhochschule Bonn-Rhein-Sieg, Germany) Bice Cavallo (University of Naples Federico II, Italy) Chin Chen Chang (Feng Chia University, Taiwan) Jorge Manuel Miranda Dias (University of Coimbra, Portugal) Andries Engelbrecht (University of Pretoria, Republic of South Africa) Pablo Estévez (University of Chile) Bogdan Gabrys (Bournemouth University, UK) Fernando Gomide (University of Campinas, Brazil) Aboul Ella Hassanien (Cairo University, Egypt) Joachim Hertzberg (Osnabrück University, Germany) Evangelos V. Hristoforou (National Technical University of Athens, Greece) Ryszard Jachowicz (Warsaw University of Technology, Poland) Tadeusz Kaczorek (Białystok University of Technology, Poland) Nikola Kasabov (Auckland University of Technology, New Zealand) ´ Marian P. Kazmierkowski (Warsaw University of Technology, Poland) Laszlo T. Kóczy (Szechenyi Istvan University, Gyor and Budapest University of Technology and Economics, Hungary) Józef Korbicz (University of Zielona Góra, Poland) ´ University of Technology, Poland) Krzysztof Kozłowski (Poznan Eckart Kramer (Fachhochschule Eberswalde, Germany) Rudolf Kruse (Otto-von-Guericke-Universität, Germany) Ching-Teng Lin (National Chiao-Tung University, Taiwan) Piotr Kulczycki (AGH University of Science and Technology, Poland) Andrew Kusiak (University of Iowa, USA)

Publisher: ŁUKASIEWICZ Research Network – Industrial Research Institute for Automation and Measurements PIAP

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems Volume 15, N° 2, 2021 DOI: 10.14313/JAMRIS/2-2021

Contents 3

Investigation of a New Hovering Autonomous Underwater Vehicle for Underwater Missions Faryar Shamshiri Amirkolai, Reza Hasanzadeh Ghasemi DOI: 10.14313/JAMRIS/2-2021/8

Enhanced Clock Gating Technique for Power Optimization in SRAM and Sequential Circuit C. Ashok Kumar, B.K. Madhavi, K. Lal Kishore DOI: 10.14313/JAMRIS/2-2021/11

Light Exoskeleton Design with Topology Optimisation and FEM Simulations for FFF Technology Piotr Falkowski DOI: 10.14313/JAMRIS/2-2021/9

A Computational Model for Multi-Criteria Decision Making in Traffic Jam Problem Ali Naeem, Jabbar Abbas DOI: 10.14313/JAMRIS/2-2021/12

Unmanned Ground Vehicle Equipped with Ground Penetrating Radar for Improvised Explosives Detection Piotr Szynkarczyk, Józef Wrona, Mateusz Pasternak, Arkadiusz Rubiec, Piotr Serafin DOI: 10.14313/JAMRIS/2-2021/10

Articles

A New Approach to an Achievement Motivation System for the Choice of an Engineering High School and Field of Study Josef Malach, Dana Vicherková, Milan Chmura, Veronika Švrčinová DOI: 10.14313/JAMRIS/2-2021/13

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

Investigation of a New Hovering Autonomous Underwater Vehicle for Underwater Missions Submitted: 15th July 2021; accepted: 6th September 2021

Faryar Shamshiri Amirkolai, Reza Hasanzadeh Ghasemi DOI: 10.14313/JAMRIS/2-2021/8 Abstract It is impossible to implement very tasks by diver, because of complexity of underwater environment and high pressure in the deep sea. These tasks can just be done by a vehicle that includes all special requirements such as: high maneuverability, precise controllability, and especially hovering capability. Underwater robots are integral parts of the industry and marine science. The application of the underwater vehicles has increased with the development of the activities in deep sea. This paper presents a special Hovering type Autonomous Underwater Vehicle (HAUV) for underwater missions. To provide the most suitable and efficient formation of vehicle thrusters for reduction of complexity of control strategies and control of the most degrees of freedom, in this paper, a new thrusters’ configuration is investigated, in terms of number of the thrusters, position and the thrust direction of the thrusters. The state feedback controller is designed according to the linear dynamic model and then applied to the non-linear model to validate the controller performance. Designed controller consists of three controllers: horizontal plane controller, vertical plane controller and surge controller. The last controller is developed to control the forward speed. For examination of the system behavior in presence of environmental disturbance and hydrodynamic coefficients uncertainty, the robustness of controller is also investigated. Keywords: Hovering type Autonomous Underwater Vehicle (HAUV); Inspection; State Feedback Controller

1. Introduction Underwater inspection and examination is an essential task for maintenance and diagnosis of damage of underwater structures such as: ship, submarine and subsea pipelines. The remote inspection of ship hulls and marine structures has become a major concern to operators of vessels and ports, because of the vulnerability of such sites to relatively small mines or other devices placed to destroy or disrupt naval and commercial activities [1]. This has become extremely critical with the threat that ships entering ports and harbors for commerce may serve as carriers of explosives, deadly chemicals and other hazardous materials, with mass destruction in highly populated cities, national landmarks, and other drastic damages at the nation scale as potentially target activities [2].

One approach for underwater inspection is to use of a number of divers. This approach is time-consuming, dangerous and non-efficient. Presenting the autonomous underwater vehicles (AUV) is a good idea to bear on these problems. Utilizing AUV resolved some of the problems like divers’ health risks, but some major problems still remain. In many underwater tasks it is crucial to have a complete control and great maneuverability on the vehicle and to collect enough data to locate the vehicle as precisely as possible. All these allow moving the vehicle close to an underwater structure as closely as needed [3]. The hovering autonomous underwater vehicle (HAUV) is a novel underwater robot that combines the maneuverability of an ROV with the flexibility of autonomous operations, so as to efficiently perform detailed surveys of large marine structures such as floating vessels [4]. Lamp Ray was one of the primary commercial inspector robots which was introduced by Harris and Slate in 1999 [5]. Lamp Ray is a ROV that moves under human control. In 2002 Trimble and Belcher designed Cetus II which using altimeters could maintain a constant distance from the hull [6]. Odyssey IV was designed in cooperation of MIT with Desset et al, which is a low cost HAUV designed for unexpected quick discovery and detection surveys and saves energy for just 1 hour [7]. One of the successful robots of this kind is the HAUV jointly developed by Bluefin Robotics and MIT which was designed particularly for underwater ship hull inspections. The goal of this design was to achieve small, low-cost robot which can be used in shallow water with great accuracy [1, 4, 8]. Negahdaripour and Firoozfam in 2005 worked on vision system of a ship hull inspector ROV [9]. One of the major efforts on underwater inspection is to improve the data collection of the robot. This can be done by using better sensors and cameras, but another issue is the trajectory of the vehicle which can take the best out of the vision sensors if planned well. Therefore, path planning plays one of the key roles in inspection of the underwater structures which is dependent on the degrees of the freedom of the vehicle. Therefore, there are numerous publications on path planning of the robots, paving the way for a more precise and efficient performance of inspection around the propellers, stern and any other particular parts with high curves. Englot and Hover in 2010 worked Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

on inspection planning [10]. Similar researches are presented in [11-14]. For the first time, Gertler and Hagen in 1967 [15] developed the equation of motion for a submarine. This dynamic model has been improved through the years by Logan [16], Liang et al. [17], and Yang [18]. Designing a practical HAUV is a challenging job and requires a well-planned sequence each step of which has its particular challenges. This process starts with dynamic modeling of the vehicle, which is a difficult step because of the special physical and mechanical properties of water causing very high nonlinear behavior of the vehicle, and time varying hydrodynamic coefficients which are dependent of the vehicle shape and speed. Also, existence of unpredicted forces caused by the waves and underwater currents makes the dynamic modeling more difficult [19, 20]. This makes engineers develop numerous control architectures which fit to each mission and cover the problem of inexact and approximate dynamic model. Most of the underwater robots are controlled by classic control method like PID controller and nowadays fuzzy controller, adoptive controller, sliding mode controller and other modern control architectures which are desirable. Yuh [21] and Carven [22] cite a range of advanced controllers used in AUVs. In the past few years, PID controller was more preferable than the other controllers. Prestero [23], Pyo et al. [24], and Choiet et al. [25] used this method owing to its advantages like being easy to maintain and apply, but it suffers from problems like being so sensitive to disturbance and parameter uncertainty and being ineffective for MIMO systems. Blasuriya and Ura [26] and Silpa-Anan and Abdallah [27] used visual servo controller designed to control the vehicle by pursuing the objects or signs in the sea, but due to short vision range in shallow water of seaports, this method does not seem practical. One of the common choices for controlling AUVs is Fuzzy logic controllers. This method is more desirable for systems with complicated dynamic model, but it is very hard to apply for the difficulties like being a time consuming process for tuning parameters [28]. This controller is mostly used to control torpedo type AUVs instead of Hovering type AUVs or HAUVs, Nag et al. [29], Ma et al. [30], Ghanavati and Ghanbarzadeh [31] used fuzzy controllers for AUVs. Sliding mode controller is a robust controller which is developed to tackle the problem of uncertainty in hydrodynamic coefficients [32, 18], Yoerger and Slotine [33]. Healy and Lienard [34] used this controller for torpedo type AUVs, and Arshad [35] used this method for an HAUV. In practice, application of robust control methods like sliding mode controller is difficult due to hardware limitations and implementation difficulties. Controllers designed in State space yield advantages like easy application and tune, appropriateness for MIMO systems and suitability for nonlinear systems.

VOLUME 15,

Articles

2021

by Chin and Lau [36] and physical constants plus added mass matrix calculated by Eng et al. [37] for the RRC robot which is an open space frame ROV developed in NTU in order to inspect and maintain of underwater pipes. Dry weight of this robot is 115 kg and maximum depth that it can endure is 100m for now. This robot is a 6degree of freedom with 4 thrusters. Since this paper is concerned with developing a robot with a precise and robust control, we did some modification on location and thrust direction of the thrusters on this robot, and, also, we added 2 other thrusters. Thus, we changed the dynamic of the system to reach the desirable characteristics for complex underwater inspection. As a result, we could control 5 degrees of freedom and fix the Roll by adjustment of the distance between the center of buoyancy and center of gravity; moreover, with no thruster to change this angle, we are able to produce enough self-aligning static torque. In figure 1, we demonstrate the location and direction of each thruster; also, in figure 2, the VRML model of this robot is developed in V-realm to study the motion of the robot can be seen. Usually, two types of coordinate system used to describe the kinematics and dynamics of the body that includes the earth-fixed frame and body-fixed reference frame that connected to the body. In figure 3, the location of each two reference frames can be observed. The coordinates system set up is the basic and first step to dynamic modeling. Body fixed frame is a Cartesian coordinate system whose positive X direction is along forward speed of the robot which is known as Surge, and earth fixed reference frame is also a Cartesian coordinate where distances and orientations of the robot are measured with respect of this reference frame.

Fig. 1. Thruster’s position in new HAUV

2. Dynamic Modeling of Hovering Type Autonomous Underwater Vehicle

In this paper, in order to set up the dynamic model, we use the hydrodynamic coefficients calculated

N° 2

Fig. 2. Underwater robots model in V-realm

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

 m − Yv mx − N v  g  0   0

mx g − Yr I zz − Nr 0 0

N° 2

0 0  v   0 0  r  1 0  y    0 1 ψ 

(Yr − m)U  YvU 0 0  v   N U ( N − mx )U  0 0   r  v r g =  +  1 0 0 U y    0 1 0 1  ψ   Fig. 3. Coordinate systems With this configuration, force and moment vector can be calculated by equation (1).

0 0 0   F 1  1 1 0  0 0 1 1 0 0   F 2   0 0 0 0 1 1   F 3 =   0 0 0 0  F 4  0 0  0 0 0 0 D5 D6  F 5     0   F 6   D1 D2 D3 D4 0 F1 + F2     F3 + F 4     F5 + F6   0     F 5 × D5 + F 6 × D6   F 1 × D1 + F 2 × D2 + F 3 × D3 + F 4 × D4 

2021

(2)

F3 + F 4   F 1 × D1 + F 2 × D2 + F 3 × D3 + F 4 × D4      0   0   Equation (3) which is related to heave and pitch used to design vertical plane controller

(1)

Where the parameters D1, D2…D6 are distances of thrusters 1, 2… 6 to the center of buoyancy: D1 = −0.4, D2 = 0.4, D3 = 0.5, D4 = −0.5, D5 = 0.5, D6 = −0.5

With respect to figure 1 and figure 3, linearized equation of motion represented by McEwen and Streitlien [38] and Radzak and Arshad [39] will become like equation (2) and equation (3). Equation (2) related to Sway and Yaw is used to design horizontal plane controller.

 m − Z w  −mx − M g w   0  0 

−mx g − Z q I yy − Mq 0 0

( Z q + m)U 0  Z wU  M U ( M − mx )U 0 q g  w  1 0 0  0 1 0 

0 0 w   0 0  q  = 1 0  z    0 1  θ 

0  w   w( zb − zb )  q  z  U   0   θ 

(3)

F5 + F6   F 5 × D5 + F 6 × D6   +   0   0   Next, we rewrite equations (2) and (3) in form of equation (4) which is standard form of state space representation of the system.

= x Ax + Bu y= Cx + Du

(4)

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

After we substitute hydrodynamic and physical coefficients, and also constant speed of u=0.8 m/s in the equations, we set up matrixes A, B, C, D, state vector and input vector for equation (2), so we are able to represent the system in state space form. These matrixes can be shown as: 0.4810 −1.4534  0 0.2937 A=  1 0  1  0

0 0 0 0 0 0.8   0 1

0 0.0158  0 0.1245 −0.1245 0.1556 B=  0 0 0  0 0  0

C = [0 0 1 0] D=0

x = [v r

VOLUME 15,

N° 2

2021

3. Control Architecture The current controller is designed by pole placement method with thruster forces as input vector in state space. Firstly, an algorithm is developed in Matlab to design the state feedback controller upon the linearized system, and then the controller is applied to linearized system and nonlinear system. To control the forward speed of the robot, PD controller is used. Figure 3 demonstrates the block diagram of the system and controller.

0.0158  −0.1556 0   0 

y ψ]

u = [ F 1 F 2 F 3 F 4]

Fig. 4. Block diagram of the system and controller

Also, calculated matrixes A, B, C, D, state vector x and input vector u for equation (3) are shown in following: 2.5722 4.0800 0  0 0.3490 0 A=  1 0 0  1 0  0

 0.0443 0.0443  0.1498 −0.1498  B=  0 0    0 0  

C = [0 0 1 0] D=0

0  −33.7927 −0.8   1 

4. Results and Discussion The controller is designed and tested on the linearized model. Figure 5 shows a spiral maneuver which has been designed to take the controller to its limitation. This maneuver is obtained by giving a sine wave input to both vertical and horizontal controllers, and this wave has amplitude of 2.5 meter and a frequency of 28 deg/s. As shown in figure 5, the robot is able to accomplish the task precisely. Figure 6 demonstrates the force output of thrusters 3, 4, 5 and 6. In linearized model, the forward speed is constant, and its value is directly involved in the dynamic model. Therefore, thrusters 1 and 2 have not been involved in the moving, and they are not considered in this simulation.

x = [w q z θ ]

u = [ F 5 F 6]

At this point, we calculated state space representation of the system in vertical and horizontal plane, so we are able to design the controller in the next step.

Articles

Fig. 5. Linear model response to desired helical path

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

Fig. 6. Control inputs in the linear model simulation After the validation of the controller on the linear model, the controller has been implemented to the non-linear model. Same path as figure 5 has been given to the underwater robot, so we could compare and analyze the results. Figure 7 shows the desirable path and the robot path for this maneuver. In figure 8, you can see the output forces of the thrusters.

Fig. 8. Control inputs in the nonlinear model simulation

Fig. 7. Nonlinear model response to desired helical path

When an organization utilizes AUVs for underwater inspection, highest possibility of control on the vehicle is desirable. It is sometimes very useful and demanding to have a complete control on the an-

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

gles of orientation of the vehicle. In figures 9 and 10, a maneuver is presented in which two states are controlled at the same time. Y and 𝛳 degrees of freedom which represent lateral movement and rotation along y axis simultaneously are controlled. Such a maneuver allows for better 3D model developing, especially around the complex shaped parts like propeller of the ship or sonar dome; moreover, when a welding mission is expected, having such capability plays a key role. Figures 11 show the thrust outputs of thrusters for this maneuver.

Fig. 9. Sinusoidal maneuver along y axis simultaneous with 𝛳 control

Fig. 11. Control inputs for the simultaneous control of 𝛳 and y

Fig. 10. 𝛳 control simultaneous with y control

Articles

Figures 12 and 13 are showing the results of controlling degrees of freedom z and 𝛳 at the same time, figure 12 represents 3D path of the robot and figure 13 represents controlling pitch at a fixed angle in this maneuver. As the robot increases depth up to 50 meters, the pitch angle will stay fixed at 20 degrees. This maneuver is more challenging than the previous one, since when the degrees of freedom y and 𝛳 are controlled, two different controllers on different planes are controlling the robot, that’s mean thrusters 4 and 5 are controlling the lateral movement and thrusters 5 and 6 are controlling the 𝛳 angle, but while controlling the z and 𝛳, both degrees of freedom must be controlled by thrusters 5 and 6, which exist at vertical plane controller. As the result shows, the controller is capable of accomplishing this maneuver accurately. Figures 14 show the thrust outputs of thrusters for this maneuver

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

Fig. 12. z control simultaneous with 𝛳 control

Fig 13. 𝛳 control simultaneous with z control

Fig. 14. Control inputs for the simultaneous control of 𝛳 and z

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

The next goal is to control y and ψ together. This is a very challenging maneuver to control these two degrees of freedom, because when thrusters 3 and 4 are controlling the lateral position of the vehicle at the same time, thrusters 1 and 2 are controlling both forward speed (surge) and yaw angle. This task has been done by using an algorithm that is developed to switch between the controllers any time that the error exceeds the allowed value. Figures 15 and 16 are showing this maneuver; at the same time, thrusters 5 and 6 are still able to control vehicle depth. This means that, in these maneuvers, three states are controlled simultaneously.

Fig. 15. y control simultaneous with ψ control

Fig. 16. ψ control simultaneous with y control

Fig. 17. Control inputs for simultaneous control of y and ψ

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

5. Robustness Examination To examine the robustness of the controller, we targeted two important parts: a) uncertainty of hydrodynamic coefficients and added mass, b) external disturbances A certain percentage of error is expected in estimation of hydrodynamic coefficient and added mass, so we multiplied this coefficient to a certain factor we named uncertainty factor in the non-linear system. In this paper, the uncertainty factor of 2 is applied to the hydrodynamic coefficients and added mass parameters. To test the controller against disturbance, three constant but different forces were applied to the system along x, y and z directions at a certain period of time (figure 18). For full robustness examination, a simulation has been utilized in which both parameters uncertainty and external disturbances are applied to the system. The desired path is the same one as figure 5. Figure 19 presents a trajectory of underwater robot under these conditions. The result shows, despite these severe conditions, the new underwater robot and presented controller acceptably follow the desired path. Figures 20 show the thrust outputs of thrusters for these conditions.

Fig. 18. External disturbance as forces

Fig. 20. Control inputs under external disturbance and parameter uncertainty

6. Conclusion Fig. 19. Tracking of desired path under external disturbance and parameter uncertainty

In this paper a highly maneuverable and controllable underwater vehicle was presented. The designed controller was able to control this multi-input, multioutput system so that the robot could track the complex trajectory. The control of multi state variables is a consequence of these presented underwater vehicle and controllers. All these characteristics cause that this underwater robot is a very suitable and practical choice for missions such as: underwater inspections (pipes, ship hull, dams, oil platforms …), rescue mission, welding and generally any mission which requires a quite maneuverable and hover capable robot. The underwater environment is very complex with Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

serious disturbance problems; therefore a robust controller is required for underwater robot; so in this paper, presented controller was tested to be robust.

AUTHORS

Faryar Shamshiri Amirkolai – Mechanical Engineering, Hakim Sabzevari University, Sabzevar, Iran.

Reza Hasanzadeh Ghasemi* – Mechanical Engineering, Hakim Sabzevari University, Sabzevar, Iran, Email: r.h.ghasemy@gmail.com. * Corresponding author

REFERENCES [1]

[2] [3]

[4]

[5]

[6]

[7]

[8] 12

F. Hover, J. Vaganay, M. Elkins, S. Willcox, V. Polidoro, J. Morash and others, “A vehicle system for autonomous relative survey of in-water ships”, Mar. Technol. Soc. J., vol. 41, no. 2, 2007, 44–55, 10.4031/002533207787442196.

U. C. Pati, 3D Surface Geometry and Reconstruction: Developing Concepts and Applications, 1st edition, IGI Global, 2012.

F. Shamshiri Amirkolai and R. Hasanzadeh Ghasemi, “Representation of an Autonomous Underwater Vehicle and Trajectory Controller design for in-water ship hull inspection”, Modares Mechanical Engineering, vol. 15, no. 10, 2016, 12–22 (In Persian). J. Vaganay, M. Elkins, D. Esposito, W. O’ Halloran, F. Hover and M. Kokko, “Ship hull inspection with the HAUV: US Navy and NATO demonstrations results”. In: Proc. of IEEE/MTS OCEANS Conference, 2006, 1–6, 10.1109/OCEANS.2006.307039. S. E. Harris and E. V. Slate, “Lamp Ray: ship hull assessment for value, safety and readiness”. In: Proc. of IEEE/MTS Oceans ‘99. Riding the Crest into the 21st Century. Conference and Exhibition., vol. 1, 1999, 493–500, 10.1109/OCEANS. 1999.799792.

G. M. Trimble and E. Belcher, “Ship berthing and hull inspection using the CetusII AUV and MIRIS high-resolution sonar”. In: Proc. of IEEE/MTS OCEANS Conference, vol. 2, 2002, 1172–1175, 10.1109/OCEANS.2002.1192132. S. Desset, R. Damus, F. Hover, J. Morash and V. Polidoro, “Closer to deep underwater science with ODYSSEY IV class hovering autonomous underwater vehicle (HAUV)”. In: Europe Oceans 2005, vol. 2, 2005, 758–762, 10.1109/OCEANSE. 2005.1513151.

J. Vaganay, M. Elkins, S. Willcox, F. Hover, R. Damus, S. Desset and others, “Ship hull inspection by hull-relative navigation and control”. In: Proc.

Articles

VOLUME 15,

[9]

N° 2

2021

of IEEE/MTS OCEANS Conference, vol. 1, 2005, 761–766, 10.1109/OCEANS.2005.1639844.

S. Negahdaripour and P. Firoozfam, “An ROV stereovision system for ship-hull inspection”, IEEE J. Oceanic Eng., vol. 31, no. 3, 2006, 551–564, 10.1109/JOE.2005.851391.

[10] B. Englot and F. Hover, “Inspection planning for sensor coverage of 3D marine structures”. In: Proc. of IEEE/RSJ International Conference, Intelligent, Robots and Systems, 2010, 4412–4417, 10.1109/IROS.2010.5648908. [11] B. Englot and F. S. Hover, “Sampling-Based Coverage Path Planning for Inspection of Complex Structures”. In: Proc. International Conference of Automated Planning and Scheduling, 2012, 29–37.

[12] G. A. Hollinger, B. Englot, F. Hover, U. Mitra and G. S. Sukhatme, “Uncertainty-driven view planning for underwater inspection”. In: Proc. of IEEE International Conference on Robotics and Automation, 2012, 4884–4891, 10.1109/ ICRA.2012.6224726.

[13] G. A. Hollinger, B. Englot, F. S. Hover, U. Mitra and G. S. Sukhatme, “Active planning for underwater inspection and the benefit of adaptivity”, Int. J. Robot. Res., vol. 32, no. 1, 2012, 3–18, 10.1177/0278364912467485.

[14] F. S. Hover, R. M. Eustice, A. Kim, B. Englot, H. Johannsson, M. Kaess and others, “Advanced perception, navigation and planning for autonomous in-water ship hull inspection”, Int. J. Robot. Res., vol. 31, no. 12, 2012, 1445–1464, 10.1177/0278364912461059. [15] M. Gertler and G. R. Hagen, “Standard equations of motion for submarine simulation”, Defense Technical Information Center, Fort Belvoir, VA, USA, 1967, 10.21236/AD0653861.

[16] C. L. Logan, “A comparison between H-infinity/mu-synthesis control and sliding-mode control for robust control of a small autonomous underwater vehicle”. In: Proc. of the Symposium on Autonomous Underwater Vehicle Technology, 1994, 399–416, 10.1109/ AUV.1994.518653. [17] X. Liang, J. Zhang, Y. Qin and H. Yang, “Dynamic Modeling and Computer Simulation for Autonomous Underwater Vehicles with Fins”, J. Comput. (Taipei), vol. 8, no. 4, 2013, 1058–1064.

[18] C. Yang, “Modular modeling and control for autonomous underwater vehicle (AUV)”, M.S. Thesis, Mechanical Enigineering, National University of Singapore, Singapore, 2008.

[19] F. S. Amirkolai and R. Hasanzadeh Ghasemi, “Designing a Trajectory controller in State Space for a Hovering type Autonomous Under-

Journal of Automation, Mobile Robotics and Intelligent Systems

water Vehicle”. In: 6th International Offshore Industies Conference, 2015 (In Persian).

[20] F. S. Amirkolai and R. Hasanzadeh Ghasemi, “Designing a discrete controller for an Autonomous Underwater Vehicle under sensoring malfunction”. In: 16th Marine Industries Conference, 2014 (In Persian). [21] J. Yuh, “Design and control of autonomous underwater robots: A survey”, Auton. Robots, vol. 8, no. 1, 2000, 7–24, 10.1023/A:1008984701078.

[22] P. J. Craven, R. Sutton and R. S. Burns, “Control strategies for unmanned underwater vehicles”, J. Navig., vol. 51, no. 1, 1998, 79–105, 10.1017/ S0373463397007601. [23] T. Prestero, “Verification of a six-degree of freedom simulation model for the REMUS autonomous underwater vehicle”, M.S. Thesis, Massachusetts Institute of Technology and Woods Hole Oceanographic Institution, Woods Hole, MA, USA, 2001.

[24] J. Pyo, H. G. Joe, J. H. Kim, A. Elibol and S. C. Yu, “Development of Hovering-Type AUV “cyclops” for Precision Observation”. In: Proc. of OCEANS, 2013, 1–5, 10.23919/OCEANS.2013.6741060.

[25] H. T. Choi, A. Hanai, S. K. Choi and J. Yuh, “Development of an underwater robot, ODIN-III”. In: Proc. 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems, vol. 1, 2003, 836–841, 10.1109/IROS.2003.1250733.

[26] A. Balasuriya and T. Ura, “Underwater cable following by Twin-Burger 2”. In: Proc. 2001 ICRA. IEEE International Conference on Robotics and Automation, vol. 1, 2001, 920–925, 10.1109/ ROBOT.2001.932668. [27] C. Silpa-Anan, S. Abdallah and D. Wettergreen, “Development of autonomous underwater vehicle towards visual servo control”. In: Proc. of the Australian Conference on Robotics and Automation, 2000, 105–110. [28] L. G. García-Valdovinos, T. Salgado-Jiménez, M. Bandala-Sánchez, L. Nava-Balanzar, R. Hernández-Alvarado and J. A. Cruz-Ledesma, “Modelling, Design and Robust Control of a Remotely Operated Underwater Vehicle”, Int. J. Adv. Robot. Syst., vol. 11, no. 1, 2014, 10.5772/56810.

[29] A. Nag, S. S. Patel and S. Akbar, “Fuzzy logic based depth control of an autonomous underwater vehicle”. In: Proc. of International Multi Conference on Automation, Computing, Communication, Control and Compressed sensing, 2013, 117–123, 10.1109/iMac4s.2013.6526393. [30] S. Ma, Y.-J. Pang, T.-D. Zhang and C. Lv, “Fuzzy S plane controller for motion control of underwater vehicles”. In: 2011 6th IEEE Conference

VOLUME 15,

N° 2

2021

on Industrial Electronics and Applications, 2011, 1640–1645, 10.1109/ICIEA.2011.5975853.

[31] M. Ghanavati and A. Ghanbarzadeh, “Control and Guidance of an Underwater Robot via Fuzzy Control Method”, International Journal of Advanced Design and Manufacturing Technology, vol. 4, no. 1, 2010, 25–32.

[32] L. A. Cooney, “Dynamic response and maneuvering strategies of a hybrid autonomous underwater vehicle in hovering”, M.S. Thesis, Massachusetts Institute of Technology, USA, 2009.

[33] D. R. Yoerger and J.-J. Slotine, “Robust trajectory control of underwater vehicles”, IEEE J. Oceanic Eng., vol. 10, no. 4, 1985, 462–470, 10.1109/ JOE.1985.1145131. [34] A. J. Healey and D. Lienard, “Multivariable sliding mode control for autonomous diving and steering of unmanned underwater vehicles”, IEEE J. Oceanic Eng., vol. 18, no. 3, 1993, 327–339, 10.1109/JOE.1993.236372. [35] M. R. Arshad and M. Y. Radzak, “Design and development of an autonomous underwater vehicle test-bed (USM-AUV I)”. In: ICARCV 2004 8th Control, Automation, Robotics and Vision Conference, 2004., vol. 1, 2004, 257–260, 10.1109/ ICARCV.2004.1468833.

[36] C. Chin and M. Lau, “Modeling and testing of hydrodynamic damping model for a complex-shaped remotely-operated vehicle for control”, Journal of Marine Science and Application, vol. 11, no. 2, 2012, 150–163, 10.1007/s11804-0121117-2. [37] Y. Eng, M. Lau and C. Chin, “Added mass computation for control of an open-frame remotely-operated vehicle: Application using WAMIT and MATLAB”, J. Mar. Sci. Technol., vol. 22, no. 2, 2013, 1–14, 10.6119/JMST-013-0313-2.

[38] R. McEwen and K. Streitlien, “Modeling and Control of a Variable-Length AUV”. In: 12th International Symposium on Unmanned Untethered Submersible Technology (UUST) Conference, 2001. [39] M. Radzak and M. Arshad, “AUV Controller Design and Analysis Using Full-State Feedback”. In: 9th WSEAS International Conference on Systems, Wisconsin, 2005.

Articles

Journal and Intelligent Systems Journal of of Automation, Automation,Mobile MobileRobotics Robotics and Intelligent Systems

VOLUME 2021 VOLUME 15, 15,N° N° 2 2 2021

LIGHT EXOSKELETON DESIGN WITH TOPOLOGY OPTIMISATION AND FEM SIMULATIONS FOR FFF TECHNOLOGY Submitted: 15th June 2021; accepted: 14th September 2021

Piotr Falkowski DOI: 10.14313/JAMRIS/2‐2021/9 Abstract: Among past years interest in robot‐assisted rehabilitation arose significantly; thus, constructions such as exoskele‐ tons are involved in this process much more often. As pa‐ tient’s bio‐signals may be included in a control loop of these devices, they may be also used to support the mo‐ tion of extremities in an everyday life. Therefore, a field of control over them stays a popular research topic. For this reason, an exoskeleton described in a paper was de‐ signed. The most important aim of a project was to ena‐ ble all anatomical movements within ranges required for the lifting of an object while minimising a mass of the device. The following paper consist of a concept of an exoskeleton and description of FEM simulations and to‐ pology optimisation applied to decrease the amount of material needed. Moreover, as an exoskeleton was built with FFF 3‐D printing technology, created parts are mo‐ delled orthotopically based on nominal mechanical para‐ meters of filaments and directions of their beams. The de‐ sign is complemented with a short description of control with EMG signals and analysis of load on a user’s muscu‐ loskeletal system. Keywords: exoskeleton, rapid prototyping, rehabilitation robotics, topology optimisation

1. Introduction Dynamic development of technology contributed to the creation of new �ields of study. Many of them are just combinations of conventional ones. Among ot‑ hers, medical and rehabilitation robotics became es‑ pecially prospective, due to current high interest in he‑ alth [5]. A signi�icant focus of rehabilitation robotics is de‑ dicated to exoskeletons [5]. These devices may be either used for physical therapy, supporting everyday mobility of a user or to affect other systems. They may be controlled mechanically or by tracking biosignals with methods such as EMG or EEG [11]. As different approaches may be used to control ro‑ bots with biosignals, they are still an important topic to research on. A presented project of an exoskeleton is designed to test possibilities of different EMG‑based methods, including direct control over motors. As the whole system supposed to be carried by a user, its mass has to be minimised. Therefore, innova‑ tive methods such as topology optimisation and FEM (�inite elements method) analysis are involved. More‑ over, at the early stage, an exoskeleton is being mo‑ di�ied within a lean cycle. Thus, it needs to be buil‑ 14

dable rapidly and not to generate high costs of pro‑ duction. For these reasons, 3D printing is considered as the most effective technology. However, combining 3D printing with computer strength analysis is not commonly applicable due to orthotropic material pro‑ perties based on a printout geometry. This paper pre‑ sents an approach towards this problem presented for the real case.

2. Concept of a Device

A designed exoskeleton (see �igure 1) of two dri‑ ven degrees of freedom (2 DOFs) supposed to sup‑ port lifting by an upper extremity of a human. It shall be usable by people suffering from musculoskeletal system’s disorders such as myopathies [15]. Its main aim is to test capabilities of control over such devices based on harder‑detectable EMG signals. However, a prototype of a robot mounted at the region of an ope‑ rator’s shoulder girdle may be applied also for the phy‑ sical works, e.g. in warehouses. According to the design constraints, an exoskele‑ ton shall be adjustable for most of the population (5th women percentile ‑ 95th men percentile) and have 7 DOFs, including 5 passive DOFs and 2 active DOFs, ena‑ bling a free motion of a limb within its whole anato‑ mic range. Complying with Polish legislation law, its load should also not exceed the maximum load of 40 N, permissible for a female worker [3]; however, con‑ struction of its end‑effector should enable carrying up to 380 N with arm muscles’ force. Moreover, its weight and maximum load should not result in a bigger load on a user’s shoulder girdle than lifting 8 kg with the straight limb [3]. The construction presented in �igure 1 consists of ten bodies, marked with B letters, twelve slide sleeves, marked with S letters and two servo‑motors. It is de‑ signed as closed loops and to work parallelly to the user’s extremity, what minimise torsion and bending stress. An elbow of an operator is supported at the S6/7 while their hand grabs a handle of the S11/12. The exoskeleton is attached to the shoulder girdle by neoprene belts assembled to B1 and it carries the load at TCP.

3. Antropomethric Model

Mechanics of the exoskeleton is design so to ena‑ ble its usage by most of the population within a whole range of motion in extremities’ joints. Therefore it is treated as a parallel mechanism to the multibody mo‑ del of human limb consisting of three segments ‑ an

Journal and Intelligent Systems Journal of of Automation, Automation,Mobile MobileRobotics Robotics and Intelligent Systems

2021 VOLUMEVOLUME 15, 15, N° N° 2 2 2021

Fig. 1. Schematic design of an exoskeleton arm, a forearm and a hand (see �igure 2). These are connected with shoulder by a 3 DOF joint, and then with each other by 2 DOF joints. The values of mechanical properties for different segments are estimated for chosen percentiles of the Polish population and afterwards used for dynamics simulations (see table 1, 2) [7] [10]. Considered ran‑ ges of anatomical motion of joints are compliant to the ISOM standards [12]. Moreover, a pattern of lifting with two human ex‑ tremities was kinematically modelled as well [14]. As a wrist joint was intended to stay rigid, all the parame‑ ters were de�ined for shoulder and elbow joints and used afterwards for inverse kinematics of a device as‑ sembled to the limb (see table 3).

4. Design Cycle

An exoskeleton has been designed within itera‑ tion cycles preceded with anthropometric modelling and declaring project constraints. An outcome of the whole process after meeting initial requirements was a complex design of mechanics and printing out cho‑ sen parts of the prototype. 4.1. Dynamics Analysis

Motion of an exoskeleton is described with an equation 1 where rG (0) is position of a handle in a frame of reference of a shoulder girdle, R are rotration

matrices coresponding to the DOFs of either shoulder φS or elbow φE joints and the other parameters are the distances between points of intersection of joints’ rotation axes (see �igure 1). ( Ry (φSy )R Rz↑ (φSz ) [lA , 0, 0]T + rG (0) = Rx (φSx )R ) (1) Rx (φEx )R R↑z (φEz ) [lF + rG , 0, 0]T +R

Diverse kinematics and dynamics of a system is computed with respect to inertia of all bodies and gra‑ vitational forces. The motion of an end efector is given with trapesium pro�ile of velocity. �elected extremity’s joints are activated with asynchronic motion within their anatomic range, while the rotation movement of TCP around its vertical axis is locked. Moreover, TCP is constantly loaded with a force of 40 N. These simulations affected in a choice of the op‑ timal kinematic scheme from considered (diversi�ied in terms of the type of the �irst drive and construction of bodies B3 and B4). It required the lowest values of force/torque from the motors and has not limited the motion of a user’s elbow by the mechanism (see �igure 3). 4.2. Materials and Production Technology Selection

As an exoskeleton supposed to be easily modi�ied and cheap to build in the prototyping process, its ini‑ Articles

Journal andand Intelligent Systems Journal of ofAutomation, Automation,Mobile MobileRobotics Robotics Intelligent Systems

VOLUME 2021 VOLUME 15, 15,N° N° 2 2 2021

Fig. 2. Scheme of used model of a human upper extremity with involved parameters Tab. 1. Lengths [mm] and masses [kg] of segments of a human upper extremity

Percentile W5 M50 M95

lA 286 349 383

lF 189 235 273

lH 179 196 210

rG 115 117 125

rc A 125.84 153.56 168.52

rc F 81.27 101.05 117.39

dA 76.39 95.49 109.5

dF 66.23 72.52 77.70

Tab. 2. Main moments of inertia of segments of a human upper extremity [kg · m2 ] Percentile W5 M50 M95

IxxA 0.395 1.254 1.861

IxxF 0.317 0.822 1.144

IxxH 0.067 0.125 0.159

ω[rad/s] 3,0 2,3 1,7 3,0 2,0 2,4 2,6 2,0 1,0 1,2

ε[rad/s2 ] 23,0 22,4 12,5 23,0 17,8 18,3 19,5 19,4 8,2 8,2

IyyA 0.420 1.340 2.011

Tab. 3. Maximum kinematic parameters of shoulder and elbow joints during lifting with two limbs involved Joint Shoulder joint

Elbow joint

Type of motion Flexion Extension Adduction Abduction External rotation Internal rotation Flexion Extension External rotation Internal rotation

φ[o ] 135 25 60 5 20 ‑75 135 40 30 ‑25

Fig. 3. Chosen kinematic scheme of an exoskeleton

tial version was designed from thermoplastic mate‑ rial. Thus, it enabled 3‑D printing with FFF technology by using a simple Prusa MK3s printer. Main bodies (B) were to be made of a �ilament with high‑durability, Compositum ABS STTM [1], while all the sleeves (S) were to be made of a sliding �i‑ 16

Articles

IyyF 0.326 0.842 1.175

IyyH 0.054 0.103 0.130

rc H 67.16 87.22 98.04

IzzA 0.109 0.269 0.413

IzzF 0.061 0.128 0.186

mA 1.50 2.35 2.97

mF 0.78 1.21 1.54

mH 0.50 0.78 0.99

IzzH 0.024 0.125 0.159

lament, iglidur® I180‑PF [2]. However, outcomes of FEM analysis caused a change of the material for bodies into CFRP composite T300/914 [9] and divi‑ sion of sleeves’ structure into the inner supporting sleeve from T300/914 and outer sliding sleeve from iglidur® I180‑PF.

4.3. FEM Analysis

FEM analysis has been run in Autodesk Inventor 2018 environment separately for every body of an exo‑ skeleton (B1‑B10). Each of them was rigidly suppor‑ ted at the point of connection with a previous body and loaded with a maximum reaction force computed in the forward dynamics at the point of connection with the next body. Apart from reaction forces including in‑ ertia of parts, a gravitational acceleration was imple‑ mented into the study. To minimise the mass needed for strength, stiff‑ ness and durability constraints, an optimal direction of beams in printouts was chosen (see table 4, where orientation is described by the two letters standing for head‑bed surface according to CAD project, where the �irst letter describes an axis from the project aligned with printout’s beams). They were aligned with a vec‑ tor of maximum reaction force occurring in a body du‑ ring its simulated motion. Afterwards, every part was divided into tetrahe‑ dral elements with 8 nodes each (see table 4) and as‑

Journal of and Intelligent Systems Journal of Automation, Automation,Mobile MobileRobotics Robotics and Intelligent Systems

2021 VOLUMEVOLUME 15, 15, N° 2N° 2 2021

Tab. 4. Details of used FEM mesh grids and parts’ orientations Part

Nodes

10,238

B2 B3

Elements

Final material

Material of a prototype

Orientation

5,144

T300/914

Compositum ABS ST

1,358

T300/914

Compositum ABS ST

21,589

11,202

2,581

1,131

3,884

1,796

3,127

B5 B6

9,452

4,564

4,621

2,142

4,137

1,898

2,581

B10

1,131

3,418

1,562

T300/914 T300/914 T300/914 T300/914

Rm [Mpa] Fu [kN] E [Mpa] ν G [Mpa]

35.23

38.56

1.41

1.54

2716.44

2730.67

0.46

0.57

1023

977

zz 23.74 0.95

2375.11 0.44

906

min

(Ee ) u ∀Ee ∈ Ea

yz zy zy

Compositum ABS ST

The whole process was repeated until the �inal mass could not be reduced without splitting the part into halves or until results of FEM analysis of modi‑ �ied body were insuf�icient. Based on the refragmen‑ ted design, a new one was created by removing mate‑ rial from selected areas. The results of maximum mass reductions are presented in a table 6. Topological opti‑ misation decreased mass of an initially‑designed exo‑ skeleton by 15%. Tab. 6. Results of a strength analysis of the exoskeleton’s parts before and after topology optimisation Mass red.

m [g]

σred [Mpa]

ε[%]

d [mm]

51.2

367

4.911

0.004

0.044

0.48

205

12.782

0.001

0.0164

248

0.291

0.001

2.05

3431.82 1280

To make a design as light as it is possible, the to‑ pology optimisation module of Autodesk Inventor 2018 has been used as well. �reviously de�ined meshes and materials (see table 4) were used. FEM analysis was in prior to any modi�ication and then certain elements were removed from the mesh to keep elastic potential energy minimal according to the formula 2, where Ke is a stiffness matrix of an element e, Ee is a �ield of the stiffness of an element e and u is a vector of displace‑ ments [13]. N e=1 Ke

Compositum ABS ST

Part

Nominal

4.4. Topology Optimisation

(∑

Compositum ABS ST

T300/914

Tab. 5. Mechanical parameters of an orthotropic model of Compositum ABS STTM material

Compositum ABS ST

T300/914 T300/914

Compositum ABS ST

T300/914

signed to orthotropic material Compositum ABS STTM with strength parameters scaled according to experi‑ mental values for typical ABS [4], based on its nominal mechanical parameters (see table 5). The analysis was carried to check the design com‑ pliance with the strength and stiffness conditions, as signi�icant deviations occurring in the parts are com‑ pensated by their parallel connection to the human ex‑ tremity.

Direction

Compositum ABS ST

(2)

Additionally, constraints on some areas were ad‑ ded. Elements from the close neighbourhood of who‑ les and bearings were not removable. Moreover, the symmetry planes were de�ined as well. Also, every part had to stay in one piece at the end of optimisa‑ tion.

B4 B5 B6 B7 B8 B9

B10

0% 5%

16% 0% 0%

26% 0%

14% 0%

15% 0%

18% 0%

26% 0%

27%

172 163 173 137 184 223 192 208 177 123 101 248 184 267 194

1.949 2.178

14.185 43.189 28.459 51.143 5.56

24.868 27.046 16.927 16.253 0.291 0.852

27.846 35.154

0.001 0.001 0.001 0.03 0.02

0.038 0.04

0.019 0.021 0.012 0.011 0.001 0.001 0.02

0.024

0.002 0.002 0.016 0.001 0.016 0.088 0.065 0.39

0.537 0.024 0.034 0.001 0.001 0.101 0.127

4.5. Rapid Prototyping To validate the design rapidly, both sliding sleeves and parts from ABS were printed with Prusa MK3s printer. Elements made of Compositum ABS STTM kept high accuracy and the tolerance of their dimensions stayed within a range of ±1 mm; even though, the prin‑ ter was not covered with any enclosure. To achieve Articles

Journal and Intelligent Systems Journal of of Automation, Automation,Mobile MobileRobotics Robotics and Intelligent Systems

smaller roughness, the designs could be enhanced by additional outer trails and then ground. Howe‑ ver, this was not necessary in this case. Regions were the screws supposed to be were printed with 100% �ill to enable drilling and threading. Sleeves made of iglidur® I180‑PF required additional post‑machining. Their dimensional tolerance stayed within a range of ±2 mm; thus, the design was enhanced by additional outer trails and then ground signi�icantly. Thanks to this, relatively big roughness was decreased to the va‑ lue permissible for the designed bearings.

5. Human Compatibility 5.1. Control System

The exoskeleton was designed to broaden capabi‑ lities of direct control over such constructions with EMG signals. Thus, the electrodes were designated to be placed in the regions corresponding to four major muscles of a shoulder girdle (SP ‑ pectoralis major, AD ‑ deltoideus, TM ‑ teres major and LD ‑ latissimus dorsi) and two muscles of an arm (BB ‑ biceps brachii and TB ‑ triceps brachii) [6]. Then the simple schema‑ tic model of a control system was performed. Howe‑ ver, it had signi�icant biases caused by its dependence on measured and estimated anatomical parameters of a user, such as masses and lengths of limb’s segments. Therefore, it was ineffective for universal application by various operators. 5.2. Equipment Placement

Apart from the designed construction assembled to the region of shoulder girdle and blade by neoprene belts, the device consisted of a battery and a control system board. As these both need to be carried by a user, they affect the load on a musculoskeletal system. To minimise the forces affecting spine extensor a simulation based on Stotte’s model was run [8]. It as‑ sumed a point force load from the battery weight equal to 15 N and a point force load from the weight of a con‑ trol panel hardware equal to 5 N. Also to decrease a force affecting oblique muscles, minimisation of a mo‑ ment of force in a coronal plane was performed. Re‑ sults of both calculations determined that the battery and a control system board should be attached right above pelvis of a user on the other side of the torso than the exoskeleton (symmetrically regarding sagit‑ tal plane). Moreover, the battery should be placed in their back, while the board should stay in their front.

6. Conclusion

The designed exoskeleton is fully applicable for pre‑de�ined tests on the ef�icacy of a direct control ba‑ sed on EMG signals. Moreover, it may be also applied for patients with musculoskeletal disorders to support their lifting capabilities. A presented approach towards mechanical design involving orthotropic modelling of parts 3D printed within FFF technology from different materials, and their further topologic optimisation and FEM strength analysis, enabled signi�icant decrement of a mass of construction. If the device would be commercialised, 18

Articles

VOLUME 2021 VOLUME 15, 15,N° N° 2 2 2021

its sliding sleeves should be topologically optimised as well. As an initial design, it may also be signi�icantly im‑ proved. Further stages of the project might include complex dynamic user’s musculoskeletal system ana‑ lysis and deep UX research. Also, the bodies could be printed from a cheaper material, not necessarily in‑ volving FFF technology, and the chosen servo‑drives could be replaced by the ones generating higher tor‑ ques and rotational velocities. Apart from its initial aim, the exoskeleton could be used for tests on EEG capabilities or support lifting in warehouses after strengthening of bodies’ constructi‑ ons. AUTHOR Piotr Falkowski – Warsaw University of Techno‑ logy, ŁUKASIEWICZ Research Network – Industrial Research Institute for Automation and Measurements PIAP, Al. Jerozolimskie 202, 02‑486 Warsaw, Po‑ land, e‑mail: piotr.falkowski@piap.lukasiewicz.gov.pl, www: https://piap.pl/.

REFERENCES

[1] “Compositum ABS ST™”. http://www. corotechnology.com/english/compositum‑ filament‑series/compositum‑abs‑st/. Accessed on: 2021‑12‑21. [2] “iglidur� I180‑PF, �ilament do drukarki 3D”. https://igus.widen.net/content/ 5xyi4cs8uh/original/3DP_DS_iglidur_I180‑ PF_Product_Data_Sheet_EN_1.pdf. Accessed on: 2021‑12‑21.

[3] “Rozporządzenie Ministra Rodziny, Pracy i Polityki Społecznej z dnia 25 kwietnia 2017 r. zmieniające rozporządzenie w sprawie bezpie‑ czeń stwa i higieny pracy przy ręcznych pracach transportowych”. http://isap.sejm.gov.pl/isap. nsf/DocDetails.xsp?id=WDU20170000854. Accessed on: 2021‑12‑21. [4] M. Cader, Szacowanie wytrzymałości prototypów wytwarzanych przyrostowo metodą FDM, O�icyna Wydawnicza PIAP: Warszawa, 2016, (in Polish). [5] R. Dindorf, “Rozwó j i zastosowanie manipula‑ toró w i robotó w rehabilitacyjnych”, Pomiary Au‑ tomatyka Robotyka, vol. 7, no. 11, 2004, 5–9, (in Polish). [6] K. Fukuda, H. Tottori, K. Kameoka, T. Ono, and K. Yoshida, “A method for estimating axle weig‑ hts of in‑motion vehicles and its evaluation”. In: Proc. of the 41st SICE Annual Conference. SICE 2002., vol. 2, Osaka, Japan, 2002, 1014–1018, 10.1109/SICE.2002.1195309. [7] A. Gedliczka and P. Pochopień , Atlas miar czło‑ wieka: dane do projektowania i oceny ergonomi‑ cznej: antropometria, biomechanika, przestrzeń pracy, wymiary bezpieczeństwa, Centralny Insty‑ tut Ochrony Pracy: Warszawa, 2001, (in Polish).

Journal and Intelligent Systems Journal of of Automation, Automation,Mobile MobileRobotics Robotics and Intelligent Systems

2021 VOLUMEVOLUME 15, 15, N° N° 2 2 2021

[8] M. Gzik, Biomechanika kręgosłupa człowieka, Wydawnictwo Politechniki S� l�skiej: Gliwice, 2008, (in Polish).

[9] A. Lowe, “Transverse compressive testing of T300/914”, J. Mater. Sci., vol. 31, no. 4, 1996, 1005–1011, 10.1007/BF00352901.

[10] J. T. McConville, C. E. Clauser, T. D. Churchill, J. Cuzzi, and I. Kaleps. “Anthropometric Rela‑ tionships of Body and Body Segment Moments of Inertia”. Technical Report AFAMRL‑TR‑80‑ 119, Anthropology Research Project, Inc., Yellow Springs, Ohio, USA, 1980. [11] E. Mikołajewska and D. Mikołajewski, “Wykor‑ zystanie robotó w rehabilitacyjnych do uspraw‑ niania”, Niepełnosprawność ‑ zagadnienia, pro‑ blemy, rozwiązania, vol. 3, no. 4, 2013, 21–44, (in Polish).

[12] M. Pajor and P. Herbin, “Egzoszkielet koń czyny gó rnej ‑ model z wykorzystaniem rzeczywistych parametró w ruchu”, Modelowanie Inżynierskie, vol. 26, no. 57, 2015, 40–46, (in Polish).

[13] P. Poszwa and M. Szostak, “Topological optimi‑ zation of the design of products manufactured by injection molding of plastics”, Mechanik, vol. 90, no. 11, 2017, 948–950, 10.17814/mecha‑ nik.2017.11.151. [14] J. Rosen, J. Perry, N. Manning, S. Burns, and B. Hannaford, “The human arm kinematics and dynamics during daily activities ‑ toward a 7 DOF upper limb powered exoskeleton”. In: Proc. 12th International Conference on Advanced Robo‑ tics, ICAR 2005, Seatle, WA, USA, 2005, 532–539, 10.1109/ICAR.2005.1507460.

[15] M. A. Samuels, Manual of Neurologic Therapeu‑ tics, Lippincott Williams & Wilkins, 2004.

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

Unmanned Ground Vehicle Equipped with Ground Penetrating Radar for Improvised Explosives Detection Submitted: 15th July 2021; accepted: 20th September 2021

Piotr Szynkarczyk, Józef Wrona, Mateusz Pasternak, Arkadiusz Rubiec, Piotr Serafin DOI: 10.14313/JAMRIS/2-2021/10 Abstract: The article presents some objectives and results of the European Defense Agency Program on: Improvised Explosive Devices Detection (IEDDET Program). The goal of the article is to describe the work, results and recommendations regarding Unmanned Ground Vehicle (UGV) and Ground Penetrating Radar (GPR) - contributions within the MUSICODE project. Its scope and goals were presented, which are in line with the objectives of the IEDDET Program taking into consideration that the recommendations (for UGV and GPR) are related to the goals, not the results of MUSICODE project. There were described scenarios and the resulting changes in the structure implemented in the UGV - the FLORIAN robot which served as a sensors carrier including Ground Penetrating Radar (GPR). The main focus of the article is to find the answer to the research question: what is an impact of using the GPR to be mounted on the UGV to detect improvised explosive devices (IEDs) on the UGV construction and the GPR results. The structure of this radar was described and examples of tests results were presented. The summary presents recommendations for the construction of an unmanned land platform to carry sensors used in the work carried out in the MUSICODE project and conclusions regarding GPR, resulting from the experiences gained under the IEDDET Program. Keywords: Improvised Explosive Devices Detection, Unmanned Ground Vehicle, Ground Penetrating Radar

1. Introduction

The goal of the IEDDET program was to develop and demonstrate multi-sensor detection systems for detecting IEDs. As a result, it will enable the increase of the ability of the army to move quickly and safely on roads and off-road. Representatives of five countries participated in the IEDDET program: Austria, Belgium, the Netherlands, Norway and Poland. According to the schedule, the program has been completed in the beginning 2020 but the final demo (in the form of a symposium without all heavy equipment present) that was planned to be hosted in Austria in March 2020 was postponed to September 2020 due to the Covid-19 situation [18]. The program included the implementation of three projects. Each of the three projects concerned different stages of IEDs detection: early warning, remote detec-

tion as well as confirmation and identification (Fig. 1). The key issue was to improve overall capability by exchanging information between the three phases. Different requirements of each of the three phases mean that a separate project was launched to implement each of them under the program (Fig. 1): − early warning – Vehicle Mounted Early Warning of Indirect Indicators of IEDs (VMEWI3); − confirmation and identification – Confirmation, Identification and Airborne Early Warning of IEDs (CONFIDENT); − remote detection – UGV stand-off multi-sensor platform for IED component detection (MUSICODE).

Fig. 1. The IEDDET Programme – the components of IEDDET System [17] The goal of the VMEWI3 project was to detect indirect IEDs presence indicators using cameras that observe the area in front of a moving vehicle. The project was managed by the Netherlands Organisation for Applied Scientific Research (Nederlandse Organisatie voor Toegepast – Natuurwetenschappelijk Onderzoek, TNO) in cooperation with industry – medium and small enterprises, so-called SME, universities and research institutions. This group of entities included: The Armament and Defense Technology Agency (ARWT) from Austria, Royal Military Academy (RMA) from Belgium, Nederlandse Instrumenten Compagnie (Nedinsco), ViNotion BV, Technische Universiteit Eindhoven (TU/e), Quest Photonic Devices BV (Quest) from the Netherlands, PCO S.A. and the Military University of Technology, both entities from Poland. In the framework of the project there was developed and tested a system consisting of an Unmanned Ground Platform (UGV) with an optical multi-sensor IEDs early warning system.

Journal of Automation, Mobile Robotics and Intelligent Systems

This CONFIDENT project primarily aims were to confirm the detection and identification of significant IED components, including electronic components, explosives, chemical, biological, radiological and nuclear (CBRN) weapon and to provide complementary early warning capability. The project was carried out by a consortium led by the Armament and Defense Technology Agency (ARWT) from Austria, Royal Military Academy (RMA) from Belgium, Military University of Technology from Poland and The Norwegian Defence Research Establishment (Forsvarets forskningsinstitutt, FFI) from Norway. The technology demonstrators were mounted on remotely controlled platforms: land (UGV) and flying (UAV). The problem of how to use a given set of possibly heterogeneous unmanned aerial vehicles (UAVs) to provide protection to a moving convoy of ground vehicles was also in the interest of some scientists [4]. The MUSICODE project developed a multi-sensor remote detection system for IED components mounted on the Polish unmanned ground vehicle – FLORIAN, which is the final product of the Polish project – INNOTECH-K1/IN1/70/154619/NCBiR/12, developed by the Consortium consisting of: WB Electronics SA, HYDROMEGA Sp. z o.o. and the Military University of Technology. The MUSICODE project was coordinated by The Norwegian Defence Research Establishment (Forsvarets forskningsinstitutt, FFI) from Norway. The members of consortia were: the Armament and Defense Technology Agency (ARWT) from Austria, Royal Military Academy (RMA) from Belgium, the Netherlands Organisation for Applied Scientific Research (Nederlandse Organisatie voor Toegepast – Natuurwetenschappelijk Onderzoek, TNO), ŁUKASIEWICZ Research Network – Industrial Research Institute for Automation and Measurements PIAP, Military Institute of Technical Engineering (WITI), Consortia Electronica (Cons-el), Advanced Protection Systems (APS), Military University of Technology. The system developed in the project analyzed data coming from sensors scanning the space in front of the UGV – on the ground and in the ground. The tasks within the MUSICODE project were also performed using data coming from the VMEWI3 system. The Ground Penetrating Radar (GPR), one of the sensors that detected items in the ground and the FLORIAN – Unmanned Ground Vehicle (UGV) are the main subjects of the article to discuss and find the answer to the research question: what is an impact of using the GPR to be mounted on the UGV to detect improvised explosive devices (IEDs) on the UGV construction and the GPR results. There was used the control vehicle (CV) (Fig. 2) equipped with a control station (Fig. 1). There were some efforts in the area of MUSICODE project. In [3] there is presented a limited, operationally-focused overview of the current status of detection technologies to be focused on those technologies that directly detect the explosives, as opposed to those that detect secondary properties of the threat, such as the casing, associated wires or electronics. But one of the key issues in this area is detection of explosives. The design and fabrication process for

VOLUME 15,

N° 2

2021

an electrochemical – “electronic nose” type sensor to detect explosives is in the interest of some scientists [5]. In [6] the novel Ground Penetrating Radar (GPR) system to detect Improvised Explosive Devices (IEDs) was numerically evaluated that is composed of a transmitter placed on a vehicle and looking forward and a receiver mounted on a drone and looking downwards [6]. In literature it is possible to find not only military context of the usage of UGVs. In [7], the results of research on the identification of gaps in the current state-of-the-art that impede the use of UGVs in various scenarios are presented. The challenge of developing infrastructure with the use of UGVs has common aspects in both civilian and military areas. The article [7] presents the results of work that show the influence of UGV Operational Requirements (ORs) on technological aspects of autonomous operation and on UGV Human-Machine Interfaces (HMIs). The authors have based the presented approach on previous experiences in analyzing UGV operations in security and military contexts, where they are widely applied. There were some scientific activities in the world to be focused on the design and implementation of unmanned ground vehicle (UGV) for security and defense applications [8, 9, 10]. In [11], the concept of implementation of components of the remote control system to transform the main battle tank T-72 into Unmanned Ground Vehicle to perform mine clearance tasks and when crossing the river on its bottom, and some results of research of the system have been presented.

2. Research Goals and Methodology of the MUSICODE Project

The capability needs defined by the militaries inspired scientists to fulfill the Operational Requirements. Analysis of these requirements was the first step within the process to describe the intended use of the instrument to close the capabilities gap. Scenarios were described and functional and environmental requirements were derived from the scenarios. The requirements were defined to the demonstrator of technology to be possible to describe the anticipated future Route Clearance CONOPs (Concept of Operations) to discuss this issue with militaries [1]. The sensor systems and the UGV driving at an envisioned speed of 30 km/h generated a high demand in the wireless communication and such a speed was the one of the goals not the result of the project. A high data rate, large range and short delay time were crucial for efficient and safe operation. For this reason, a new control communication system was developed. The developed system includes independent channels for UGV control, environment visualization and for sensor data transmission. The main issues to solve were the required high speed of data transmission under the constraints of limited frequency bands, short delay time and EMC/EMI due to limited antenna space [1]. Related to this some modifications were required to the manned control vehicle (CV). The control vehiArticles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

Fig. 2. The MUSICODE project concept – the components of MUSICODE System

cle needed to be fitted with a control station (Fig. 1) for the UGV, a control station for remote sensor operation, a power source and all the necessary antenna systems. Since the UGV and sensors have to be operated at high efficiency while the control vehicle is moving, care has to be taken for comfort and ergonomics of the operating crew and the testing personnel. For this reason, a 3D model was developed and tested including all supporting frames and cable placements [1]. Finally, all parts have to be integrated into one working system. This includes the control vehicle (CV), the wireless communication between the control vehicle and the UGV, the UGV remote control and terrain visualization system and the sensor control and data acquisition (Fig. 2, 3). The second part of this task involved testing and debugging of the integrated system [1]. The general aim of the MUSICODE project was to develop and demonstrate a multi-sensor UGV for stand-off detection of IED components, where standoff as a result of realism was defined as distance to personnel, not to the sensors themselves [18]. The specific goals of the MUSICODE project were: a) to evaluate and demonstrate state-of-the-art sensors of the following types: (i) downward looking Ground Penetration Radar (GPR), (ii) metal detector, (iii) command wire detector and (iv) non-linear junction detector. The sensors can be partly new developments, and partly improvement of off-the-shelf sensors; b) to implement improved algorithms for signal processing and automated detection, both for individual sensors and in combination, and to present the results in a novel human-machine interface to the operator; c) to exploit available information from the early warning phase by reading data from an off-line common detection map / database, and to pass information to the same off-line database for use of information in the following identification and confirmation phase; Articles

d) to integrate the multi-sensor suite on an adapted UGV system for participation in joint test and evaluation trials.

Fig. 3. View of some components of VMEWI3 and MUSICODE projects The outcome of the project is at the 5th Technology Readiness Level (TRL) – technology demonstrator for multi-sensor platform for IED component detection. There is a need to distinguish between the Project’s goals and its results, which in some cases are not the same.

3. The FLORIAN Platform Construction – Description of Changes Adapting the Platform for the Implementation of Tasks Within the MUSICODE Project

The Unmanned Ground Vehicle (UGV) was the element merging the developed sensors into a IED detection system. Manned vehicle was used as a control vehicle (CV) to control UGV and sensors (Fig. 2) and to cooperate with control station (Fig. 1). The UGV used in the MUSICODE project was developed as a part of a national research program as a robot for heavy rescue tasks. As a result, it was directly used in the MUSICODE project. To fulfill the detailed specifications some modifications of the UGV were required. In its original configuration the UGV had two attachments – the manipulator and the

Journal of Automation, Mobile Robotics and Intelligent Systems

loader attachments. Their load capacities are 200 kg for manipulator and 1000 kg for loader attachment (with forklift). In the original version the UGV was controlled by a laptop plugged into the control console (Fig. 4b) [2]. The work conducted in the MUSICODE project identified several areas of the UGV which had to be modified. During the design development a particular focus was on Electromagnetic Compatibility (EMC) issues. There was defined the concept of reducing Electromagnetic Interferences (EMI) and the other works were done in the MUSICODE project required efforts in two areas: − UGV adaptation; − car adaptation for the role of the CV (Fig. 2).

VOLUME 15,

N° 2

2021

− development of a length measurement system for the UGV front suspension cylinders; − development of positioning system; − development of a data link; − development of a video link; − development of the UGV control software modification. The vehicle adaptation to be a control vehicle (CV) (Fig. 2) mainly included: − development of the UGV control station; − development of an operation station for sensors Human Machine Interface(HMI) to operate with the detection systems (Forsvarets forskningsinstitutt, FFI area of responsibility). At the very design stage, the supporting frame element did not foresee any noteworthy UGV modifications. The only modification was a quick-action coupling which would connect the UGV to the frame. The mechanical element integrating the frame and the UGV is a metal attachment connected to a mounting plate on the UGV. This is the only function of the mounting frame, meaning that the UGV design modification on the support frame was indirect (Fig. 5) [2].

WD Fig. 5. Frame model with elements of the detection system: GPR – Ground Penetrating Radar, WD – Wire Detector, MD – Metal Detector Fig. 4. View of the original version of the Florian platform: a) the Unmanned Ground Vehicle, b) the control console The general adaptation works included: − development of the supporting frame for the needs of the GPR, MD and WD mounting (Military Institute of Technical Engineering, WITI area of responsibility); − development of the modification of the coupling system of the sensors frame with the UGV; − development of nonlinear junction detection (NLJD) system fixing (Nederlandse Organisatie voor Toegepast Natuurwetenschappelijk Onderzoek, TNO area of responsibility); − development of the fixing elements of the computers and fixing elements of the processing measurement data system; − development of a power system for detection equipment; − modification of the UGV engine chamber cooling system;

In the final version of the support frame, the position of the longitudinal reinforcement beam was modified and closing elements of the open square profiles used were added. The change in the position of the longitudinal beam was determined by the weight of wire detector (WD) in order to achieve its stable operation while driving (Fig. 2, 5). The consequence of such loading was a greater vertical displacement of the corner of the frame and consequently a larger value of the torsion angle [2]. The change in the method of supporting the elements built on the support frame for MD, WD and the GPR (piano wheels) in relation to standard mounted tools resulted in the need to modify the existing locking system in the mechanical quick coupler. The UGV of the MUSICODE project in the original version (Fig. 4a) was equipped with a standardized type of mechanical quick coupler in accordance with SAEJ2513. This design is dedicated to the working tools of skid steer loaders. The upper edge of the subplate is shaped in a triangular form, thanks to which it adapts to the fastening elements of the working tool. Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

The final “locking” and “unlocking” of the work tool on the quick coupler occurs after turning the two levers and sliding the bolts from the bottom edge of the plate (Fig. 6) [2]. A.

Fig. 7. View of Unmanned Ground Vehicle with mounted sensors frame During the tests, sometimes the sensors frame was accidentally and uncontrolled disconnected from the quick coupler plate. This situation is highly undesirable. Therefore, it was necessary to modify the locking system in the quick coupler plate.

Fig. 8. View of the modernized quick-release plate locking system

Fig. 6. Locking mechanism of the UGV quick coupler plate: A – unlocked position, B – locked position

The originally mounted tool on the UGV (Fig. 4) was loaded with a moment of force that prevented disconnection of the tool from the UGV thanks to quick coupler. For testing purposes the original tool was removed from the base platform. In its place a new version quick coupler plate was mounted (Fig. 7). The sensors frame was attached to the base platform’s quick coupler plate using locking mechanism (Fig. 6). The attached sensors frame was equipped with two road wheels that play a supporting role. In this case, dynamic tests showed that it is possible for the sensor frame to be loaded vertically upwards [2]. Articles

As a result of the analysis, it was decided to change the length of the locking bolts. Because the conditions of the MUSICODE system do not require frequent disconnection of the sensors frame from the UGV, it was also decided to implement the mechanical screwed bottom bolt protection. The modernized locking system is shown in the Fig. 8 [2]. Conducting the necessary modification works on the platform (with particular emphasis on the GPR) required the collection of detailed requirements for individual detection subsystems. They concerned, inter alia, information such as: − the number of components included in each detection system; − external dimensions of each component; − mass of each component; − existing mechanical interfaces; − electrical power of all components; − existing electrical interfaces (location, wiring);

Journal of Automation, Mobile Robotics and Intelligent Systems

− requirements for the location of elements with respect to metal elements and with respect to other sensors (Electromagnetic Compatibility). The collected information enabled the development of a plan for the distribution of the main components of the detection system (Fig. 2). It was decided that: − the GPR, WD and MD will be attached to a support frame located on the front of the UGV; − NLJD will be built directly on the platform (instead of a manipulator); − all additional components (computers, measuring data processing elements, etc.) will be built above the UGV engine space; − due to the insufficient power of the UGV alternator, it was decided to use an additional source of electricity to power the detection system components. The decisions made allowed defining the available assembly spaces and conducting a preliminary functional – spatial – mass analysis. In the case of NLJD, the installation space at the front of the UGV (Fig. 9a) and the need to use existing mounting points (Fig. 9b.) are specified. a)

VOLUME 15,

N° 2

2021

tion on the UGV made it impossible to fit them using the current design possibilities. Particular emphasis was placed on the aspect of the area of visibility by teleoperation system cameras and not reducing the mobility of the robot. The analyzed areas are shown in the Fig. 10. a)

c) b)

d) Fig. 9. View of the space selected for mounting NLJD (a) and existing mounting points (b) The development of a support frame dedicated to WD, MD and the GPR belonged to the area implemented by Military Institute of Technical Engineering (WITI). Then, the analysis of the possibility of placing computers, measuring data processing components and other components necessary for the detection system on the platform was carried out. The considerable size and mass of devices necessary for installa-

Fig. 10. Platform view with analyzed zones: a) side view, b) isometric view, c) top view, d) front view Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

4.1. Hardware The GPR detector is composed of 12 modules. Each module consists of one microwave ultra wideband transmitter and two independent receivers, so it contains two separate radar channels. It gives totally 24 channels that cover width of 3 m scanning path. The transmitter of each module illuminates the ground using central antenna and receivers record the radio wave echoes by two side antennas. The planar ultra-wideband Vivaldi-type antennas of all co-operating radio modules make linear array with λ/4 offset [12, 13]. The antenna array is shown in the Fig. 11. main computing unit switch

... M1

R x24

R x23 T x0 1 2

R x22

R x21 T x11

λ/4

R x04

R x03 T x0 2

R x02

R x01 T x0 1

...

Fig. 11. Schematic view of antenna array and modules All channels are time and spatial separated. The single channel diagram is shown in the Fig. 12. It consists of triangular waveform oscillator driving precise voltage control oscillator that generates signal of 1 GHz bandwidth. Mixing it with the local oscillator signal 4 GHz and its doubled frequency copy, one can obtain doubling of the bandwidth [14]. 1-3 GHz

VCO 5-6 GHz

TWO

AD Rx

4 GHz

8 GHz

0-500 kHz

Fig. 12. Outlook of GPR bandwidth generation; TWO – triangular waveform oscillator, VCO – voltage-controlled oscillator, AD – amplitude detector

Articles

Rx Rx HF board Tx

optic Tx digital control

4. Ground Penetrating Radar – Construction and Field Data Processing

2021

The generated probing signal consists of two sub bands. The lower sub band is generated during an increase in voltage of TWO that drives VCO giving the frequency from 5 to 6 GHz; the upper sub band is making during a decrease TWO voltage and consequently frequencies from 6 to 5 GHz. In the first case, the signal from VCO is mixed with 4 GHz and in the second case with 8 GHz frequency. The switch of the frequencies follows the detection of minimum or maximum of TWO amplitude. As a result, the output signal is a sum of two sub bands 1-2 GHz and 2-3 GHz. The signal is coherent in a frame of sub bands. The transmitted signals are mixed with received echoes, so the frequency difference corresponds with depth of penetration and amplitude and phase gives an information about target existence and properties. Each channel has the following key parameters: - lower frequency fL~1GHz; - upper frequency fH ~ 3 GHz; - stepped frequency continuous wave modulation (SFCW); - number of frequencies N = 500 ; 4 MHz; - value of frequency step ∆f = - phase shift resolution ∆φ = 0.1 ; - time resolution tr< 0.4 ns; - maximum power of signal 20 dBm. The intermediate frequency signal is next pre-processed using dedicated fast electronic field-programmable gate array (FPGA) structures that provide the resulting data over standard protocol to the high-level software. The internal organisation of single module is shown in the Fig. 13. antenna connectors

The area (Fig. 10) was divided into four zones: 1 – right, 2 – back, 3 – top and 4 – left. Zones 1 and 4 were excluded from the possibility of their use due to the presence of access points to the drive system, control, startup system, communication and the UGV fuel supply system in their areas. Due to the mass distribution on the modified UGV and its stability, zone 2 was reserved for the new power source. Only area 3 was prospective, but only in its front.

N° 2

optic data link

FPGA DSP board Eth

power supply and driving board

Eth data link

power connector

military casing

Fig. 13. Internal organization of single module The hardware modules and antenna array are mounted inside dedicated metal-free casing. The space-time data necessary to mark the detections are generated by separate GPS system that send data markers for all collaborating detectors.

4.2. Signal Processing

The digitized IQ signals are submitted to the discrete Fourier transform (DFT) FPGA module. The transform in the such kind of radar plays the role of the range compression procedure. An exemplary spectrum of the signal is illustrated in the Fig. 14. It may be observed that the information on the ground’s re-

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

flectivity is contained in the transform domain. The frequency domain has been normalized to the sampling frequency range of the transform samples for positive frequencies. The GPR is designed to detect objects buried in shallow layers of the ground, therefore it does not require signals representing echoes from the whole range. Due to this only a selected set of the DFT samples is stored in the system and sent to the signal processing unit. 12

4 3

N° 2

2021

Im [x (n)] . 103

2 1 0 -1 -2

Re [x(n)] . 10

-3 -4 -5

200

400

600

800

1000

400

600

800

1000

n - sample number 1200 1400 1600

X( f ).10

2 0

SB(n)

-2

-4

Im [x (n)]

-6

-8

f /fs 0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

Fig. 14. An exemplary spectrum obtained in chosen channel The signal processing procedure starts with sorting of vectors corresponding to appropriate receiving channels, and aligned according to the time stamps obtained from GPS. The block diagram of the procedure chain is shown in the Fig. 15. Due to the comparatively fast movement of the carrier platform on rough ground surface along the tested route, the whole GPR modules and antennas strongly vibrate. This induces fluctuations of the received signals. It requires the filtration of those fluctuations in the slow time domain and, as the vibrations are of higher frequency than the changes in the ground that are of interest, a low-pass filter is applied with a –3 dB cut-off relative frequency of 0.12 [Hz/ Hz] that removes the vibrations. In the Fig. 16 an exemplary input signal and its filtered version is shown.

raw data

low-pass filtration

normalisation

sub-band signal combination

-10

Re [x(n)]

-12 -14 0

200

n - sample number 1200 1400

1600

Fig. 16. An exemplary signal before (a) and after lowpass filtration (b) The filtered signals are next weighted in order to equalize the amplification of the particular channels. The equalization is based on measurements of the mean magnitude of the noise signals in each channel. During the measurement, as has already been indicated, each module sends two successive linear frequency modulated signals in two separate sub-bands, a lower and a higher one. The combined bandwidth of the two sub-bands is equal to 2 GHz, and the sum of the signals from both the sub-bands allows to achieve the full range resolution of the GPR sensor. Therefore, the filtered and weighted complex signals are combined giving 24 signals that create a complete complex 3-D image of the ground’s reflectivity.

background echo cancellation

fast time integration

object detection

detections

Fig. 15. Block diagram of the signal processing chain

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

guard window guard window input signal

measurement window

measurement window cell under test

Fig. 17. Distribution of the measurement windows In order to extract the in homogeneities of the ground, that may appear dangerous objects, a background removal procedure is applied. This procedure bases on averaging the level of background echo signals around the tested chosen cell and subtracting it from the analyzed signal. The measurement is done by averaging the signal levels in two windows in the left and right neighborhood of the cell under test (CUT). In order to separate the measurement from the signal level in the CUT, two guard windows around the CUT are applied. There is shown in the Fig. 17 the distribution of the measurement windows in the algorithm. There are many variants of this procedure and majority of them base on averaging the background signals in measurement windows, there are also algorithms that base on the principle of a kind of a median filter that performs a sorting in the two measurement windows, and then finds the value on n0-th cell in the sequence taking it as the estimate of the background signal level. Those algorithms are very similar to the well known from classic radar technology CFAR (Constant False Alarm Ratio) algorithm with the general difference that the estimate of the background is not compared to any thresholds, due to the fact that the GPR should detect sudden lowering of the echo levels as well as peaks [15, 16]. The implemented algorithm is illustrated in the Fig. 18. The background elimination algorithm consists of two main steps: 1. For each n-th cell the estimates CL i CR of the background echo signal level in two windows (left and right) are computed, each window has the length of N, the edge of each window is at the distance of G cells from the CUT position;

input signal

1 2N Σ

input signal

Fig. 18. Block diagram of the background elimination algorithm 28

Articles

CL (n ) =

n − G −1

∑

SB k = n − N −G

(k ) ,

(1)

n+ N +G

CR (n ) = ∑ S B (k ) ,

(2)

k =n+G

2. The estimates are added up and divided by 2N, subsequently the average is subtracted from the signal in the CUT, the difference is stored in the n -th cell of the output signal:

S BE= (n ) S B (n ) −

C L (n ) +C R (n ) 2N

(3)

The next procedure of the signal processing chain is the signal integration in the fast time direction (towards the ground depth). The objects sought by the GPR, i.e., improvised explosive devices, in order to gain a proper effect, for example destruction of an armored vehicle, need to be of significant size. This means that their echo will dwell more than one range cell. In order to improve the signal-to-disturbance ratio an integration along the range direction is performed: p +1

S Int ( p ) = ∑ S BE (k ) , k = p −1

where p is the number of integrated range cells.

(4)

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

In the Fig. 19 there is presented the signal before and after the background elimination procedure. a)

Sf (n)

25 20 15 10 n - sample number 5

200

400

600

800

1000

400

600

800

1000

1200

1400

1600

1800

2000

SCFAR (n)

4 3 2 1 0 -1 -2 -3 -4

n - sample number

-5

200

1200

1400

1600

1800

2000

Fig. 19. The signal before (a) and after the background elimination algorithm (b)

channel number - road width = 3 m

After the integration a comparison of the signal with a threshold is performed and the detection report is generated. The report in JSON format is next sent to the fusion unit gathering information of all sensors of the platform. The additional subroutine allows to preview the detections. In the Fig. 20 there is shown an exemplary imaging generated using the subroutine. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

scanned distance of the road [m]

Fig. 20. An exemplary imaging of the detections at the distance of 70 m. The dots indicate suspicious objects The detection report contains also false alarms that should be eliminated in a final fusion process.

5. Conclusion

The described UGV design and the results of GPR field data processing being one of the final results of the

N° 2

2021

MUSICODE project allowed to find the answer to the research question: there is an impact of using Ground Penetrating Radar to be mounted on Unmanned Ground Vehicle to detect improvised explosive devices (IEDs) on the UGV construction and the GPR results. The changes in the structure of UGV allowed to fulfil the requirements to be the carrier for IED detection sensors but with some limitations resulting from the basic structure of current version of the UGV. The UGV used in the project never demonstrated a driving speed of 30 km/h which was one of the goals. It is possible to achieve such a UGV speed by making design changes resulting from a change in the approach to the platform capability based on additional sources of information. The lessons learned knowledge can be implemented in the continuation works within this IED detection area. Further research should be focused on new design of UGV dedicated for the carrier of different sensors that should be used in IED detection missions and on development of new versions of the sensors with particular attention on GPR. The main goal of the new project should be development of a new generation of sensors to explore the possibilities of increasing the detection and identification of IEDs in complex tactical and operational conditions, and to develop a new solution of UGV. The specific objectives of the new project in the field should be: a) examining the possibilities of integrating early warning and remote detection systems on one reconnaissance platform. b) examining the possibility of direct cooperation of a group of robots for the detection and protection of convoys – early warning / remote detection as well as confirmation and identification – unification of the pace of operations; c) development of swarm technology and miniaturization of smaller and lighter sensors [18]; d) development of a new multi-task base platform that can act as a carrier of systems, in particular detection task; e) development of the suspension and control system (voice control) of the UGV as well as its executive mechanisms, including innovative manipulators with haptic control. f) development of a manipulator solution enabling confirmation and identification of IEDs placed in the ground or masked with heavy objects (necessary load capacity of 50-80 kg at a range of 3-4 m); g) development of a manipulator solution enabling rapid inspection of heavy goods vehicles and culverts without leaving the robot off the road crown; h) development of new control systems for inspection and intervention manipulators (with haptic and vector control); i) examining the possibilities of expanding the GPR system detection capabilities to detect of nonlinear junctions located on the route and detection of metal clusters (possibility of taking over the role of an inductive metal detector); Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

j) examining the possibility of mechanical stabilization of the position of GPR antenna. The new UGV should be modular and open architecture to be better adaptable to different missions including IED detection missions. There should be improved among others the suspension system to be useful for carrying the new generation of sensors including GPR, manipulators to be able to fulfil the requirements coming from the new generation of CONOPS.

VOLUME 15,

[5]

ACKNOWLEDGEMENTS

[6]

This work was supported by the European Defence Agency under PROGRAMME ARRANGEMENT (PA) No B 1465 GEM3 GP concerning the “IED Detection Programme” (IEDDET-programme).

[7]

AUTHORS Piotr Szynkarczyk – ŁUKASIEWICZ Research Network – Industrial Research Institute for Automation and Measurements PIAP, Al. Jerozolimskie 202, 02-486 Warsaw, Poland. Józef Wrona* – ŁUKASIEWICZ Research Network – Industrial Research Institute for Automation and Measurements PIAP, Al. Jerozolimskie 202, 02-486 Warsaw, Poland, Email: jozef.wrona@piap.lukasiewicz. gov.pl.

Mateusz Pasternak – Military University of Technology, gen. Sylwestra Kaliskiego 2, 00-908 Warsaw 46, Poland. Arkadiusz Rubiec – Military University of Technology, gen. Sylwestra Kaliskiego 2, 00-908 Warsaw 46, Poland.

Piotr Serafin – Military University of Technology, gen. Sylwestra Kaliskiego 2, 00-908 Warsaw 46, Poland. *Corresponding author

REFERENCES [1] [2] [3]

[4] 30

“The ‘MUSICODE’ Project”. PIAP, https://piap. lukasiewicz.gov.pl/en/research-projects/themusicode-project/. Accessed on: 2021-12-22. “Report on UGV with integrated sensors”, EDA, 2020. A. A. Faust, C. J. de Ruiter, A. Ehlerding, J. E. McFee, E. Svinsås and A. D. van Rheenen, “Observations on military exploitation of explosives detection technologies”. In: R. S. Harmon, J. Holloway and J. T. Broach (eds.), Proc. Volume 8017, Detection and Sensing of Mines, Explosive Objects, and Obscured Targets XVI, 2011, 10.1117/12.886391. S. C. Spry, A. R. Girard and J. K. Hedrick, “Convoy protection using multiple unmanned aerial ve-

Articles

[8]

[9] [10]

[11]

[12]

[13]

[14]

N° 2

2021

hicles: organization and coordination”. In: Proc. of the 2005, American Control Conference, 2005, 3524–3529, 10.1109/ACC.2005.1470519. S. M. Silva, J. D. Gamarra, C. A. Hernandez and J. F. Osma, “Design and fabrication of a sensor for explosives as a first step to an IED detection device”. In: 2014 IEEE 9th IberoAmerican Congress on Sensors, 2014, 1–4, 10.1109/IBERSENSOR.2014.6995515. M. Garcia-Fernandez, A. Morgenthaler, Y. Alvarez-Lopez, F. Las Heras and C. Rappaport, “Bistatic Landmine and IED Detection Combining Vehicle and Drone Mounted GPR Sensors”, Remote Sens., vol. 11, no. 19, 2019, 10.3390/ rs11192299. J. Czarnowski, A. Dąbrowski, M. Maciaś, J. Główka and J. Wrona, “Technology gaps in Human-Machine Interfaces for autonomous construction robots”, Autom. Construct., vol. 94, 2018, 179–190, 10.1016/j.autcon.2018.06.014. A. Bouhraoua, N. Merah, M. AlDajani and M. Elshafei, “Design and implementation of an unmanned ground vehicle for security applications”. In: ISMA’10 - 7th International Symposium on Mechatronics and its Applications, 2010. J. Łopatka, T. Muszyński and W. Polis, “Modułowe lekkie bezzałogowe platformy lądowe wsparcia”, Szybkobieżne Pojazdy Gąsienicowe, vol. 45, no. 3, 2017, 99–113, (in Polish). P. Klinkhachorn, A. S. Mercer, U. B. Halabe and H. V. S. GangaRao, “Unmanned Ground Vehicle for Autonomous Non-Destructive Testing of FRP Bridge Decks”. In: AIP Conference Proc., vol. 894, 2007, 1723–1730, 10.1063/1.2718172. A. Bartnicki, M. J. Łopatka, L. Śnieżek, J. Wrona and A. M. Nawrat, “Concept of Implementation of Remote Control Systems into Manned Armoured Ground Tracked Vehicles”. In: A. M. Nawrat (ed.), Innovative Control Systems for Tracked Vehicle Platforms, vol. 2, 2014, 19–37, 10.1007/978-3-319-04624-2_2. M. Pasternak, “Lambert W function application for construction of antipodal Vivaldi-type antenna”. In: 2018 14th International Conference on Advanced Trends in Radioelecrtronics, Telecommunications and Computer Engineering (TCSET), 2018, 624–627, 10.1109/TCSET. 2018.8336279. Nurhayati, G. Hendrantoro and E. Setijadi, “Mutual coupling and radiation pattern of Vivaldi antenna with slit”. In: Proc. of the 3rd International Conference on Communication and Information Processing - ICCIP ‘17, 2017, 296–300, 10.1145/3162957.3163056. M. Pasternak and P. Kaczmarek, “Continuous wave ground penetrating radars: state of the art”. In: XII Conference on Reconnaissance and Electronic Warfare Systems, vol. 11055, 2019, 84–89, 10.1117/12.2524524.

Journal of Automation, Mobile Robotics and Intelligent Systems

[15]

[16] [17]

[18]

VOLUME 15,

N° 2

2021

M. Kronauge and H. Rohling, “Fast Two-Dimensional CFAR Procedure”, IEEE Transactions on Aerospace and Electronic Systems, vol. 49, no. 3, 2013, 1817–1823, 10.1109/ TAES.2013.6558022. M. Labowski, P. Kaniewski and P. Serafin, “Motion Compensation for Radar Terrain Imaging Based on INS/GPS System”, Sensors, vol. 19, no. 18, 2019, 10.3390/s19183895. “EDA IED Detection Programme (IEDDET)”. M. Kalbarczyk, http://www.irsd.be/website/ images/images/Activites/Colloques/presen tation/2016-05-17/05-Mr-Marek-KALBAR CZYK.pdf. Accessed on: 2021-12-21. MUSICODE consortium and PMG, “MUSICODE D.1.2 – Project Final Report”, B 1465GEM3 GP, 2020.

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

Enhanced Clock Gating Technique for Power Optimization in SRAM and Sequential Circuit Submitted: 7th May 2021; accepted: 11th October 2021

C. Ashok Kumar, B.K. Madhavi, K. Lal Kishore DOI: 10.14313/JAMRIS/2-2021/11 Abstract Low power VLSI designs are having wide variety of application usage in real-time. VLSI circuits are analyzed with various power reduction strategies. Existing approaches are used the clock frequency control, switching activity and scaling factor for power reduction. The glitching problem and clock triggering issues are higher therefore; the proposed work utilized the improved circuit of clock gating technique. In this proposed work, the enhanced clock gating with D-latch model is constructed to obtain the less power consumption. The traditional clock gating technique is improved by adding clock triggering on LATCH circuit and adding buffer circuit between the source and load circuitry to reduce the clock switching issues like gitching and clocking activity. Here the SRAM and sequential counter circuits are designed to utilize the power reduction strategy for improving the performance. This is applicable for various applications in real world and utilizing the FPGA and DSP application specific circuits. Experimental results are analyzed to obtain the power reduction result of SRAM and sequential circuit. Area, power, and delay are obtained the better results as compared with the previous work. Overall, design is performed using Xilinx 14.2 ISE suit. Keywords: Enhanced clock gating; D-Latch gating; SRAM; sequential circuit; Area; Delay

1. Introduction

Power dissipation reduction issue in VLSI circuits plays a wide role for utilizing real-time applications. Power utilization in VLSI circuit as gatelevel exchanging actions are performed with various techniques. Gate level circuit might be used to select different combination of logics related to power utilization. The noise in digital circuit depends on the power utility, temperature, semiconductor property and so on. The power dispersal is the most part because of ground bouncing on the circuit. The lower bound of the power dissipation can be determined to utilizing data move limits of channel [1]. In gate level circuit, diverse mix of legitimate gateways may deliver same circuit yields however unique estimation of power utilization is obtained [2]. By maintaining a strategic distance from the delay at each information of gate logic, logical utility with module is determined. The semiconductor memory is separated into two sorts that are static and dynamic [3]. The static RAM uti-

lizes bipolar or MOS flip-flops and the dynamic circuit use MOSFETs and capacitors that are used to store the information. The SRAM is segmented to preparing arrangement of sensor nodes, which needs to fulfill the low-power necessity also [4]. Major objective is to reduce the power dissipation of VLSI circuits this utilizes many techniques and novel circuit approaches, which are studied in various reviews. Specifically, dissipated power is decreased utilizing various techniques that are voltage scaling, semiconductor scaling and the utilization of rest semiconductors. The 6T SRAM cell is to keep up the circuit execution and energy productivity with complex design. DFF with large register circuits on VLSI module mostly preferred the clock gating technique to reduce the leakage power [5]. Power dissipation reduction technique done with clock triggering strategy, gated clock generation with switching pulses, and synchronous clock generation circuit [6]. Power utilization is significantly expanding for Static Random Access Memory Field Programmable Gate Arrays along these lines lower power FPGA hardware and new CAD apparatuses are required. Clock-gating approaches have been applied in low force FPGA plans with just minor achievement in diminishing the complete normal force utilization [7]. The clock-gating strategy depends on the fractional reconfiguration and topological adjustments [8]. The arrangement depends on the powerful incomplete reconfiguration of the design memory outlines identified with the clock directing assets. D flip-flop utilizes tree-based clock drivers with gating to significantly decrease the stacking on dynamic clock drivers. Furthermore, D flip-flops are utilized to lessen the clock spikes and, in this manner, diminish the power utilization on the clock signal [9]. Design contains DFF to control the conveyance of the neighborhood clock signal “CLK “to the memory, and the “Lock signals along the way passing the worldwide clock source to the nearby clock signal are dynamic [10]. The yield of DFF and worldwide clock feeds to AND based RTL circuit, which produce neighborhood clock signal for memory [11]. Power utilization is drastically expanding for SRAM-FPGAs, thusly lower power FPGA hardware and new CAD devices are required. Clock-gating philosophies have been applied in low power FPGA with just minor accomplishment in diminishing the all-out power utilization [12]. The clock-gating method depends on inner halfway reconfiguration and topological changes. The arrangement depends on the powerful incomplete reconfiguration of the setup memory outlines identified with the clock steering assets [13].

Journal of Automation, Mobile Robotics and Intelligent Systems

The exchanging movement of the circuits can be utilized to locate the normal power scattered which thus helps in examining the speed performance. The exchanging movement can be restricted in VLSI plan by utilizing a method called check gating in simultaneous circuits by cutting the inactive patterns of flip flop [14]. Clock gating is the technique for adding additional rationale to infer a gated clock which is taken care of into the DFF. Delay Minimization and Power Minimization are two significant targets in the plan of optimal circuits [15]. Here the retiming is a viable method of postpone improvement of successive circuits. This depicts a calculation in RTL that discovers least spreading over tree for associated VLSI. The Bellmanford calculation is used at that point of examining the prim approach to focus on the synchronous circuitry [7]. Clock gating is acknowledged as the power advanced procedure as it decreases the power at framework level, RTL and gate level. More significant levels of advancements are accomplished more in RTL level than gate level, where tasks are completed in register blocks as restricted on logical gate circuit [16]. The fundamental point of the clock gating strategy is to remove the clock during the inactive patterns of flip flop [17]. Clock gating procedure is executed for three diverse cell types: 1) Latch based cell, 2) Flip-flop based cell, 3) Gate based cell. The drawback is that, for the positive edge set off counter when empower signal goes from 1 to 0 and when the clock is at rising edge, a glitch happens on account of the more prominent falling time span of empower signal [18] and [19]. The yield acquired is mistaken, in view of the previously mentioned reason. Any risk that happens when empower is equivalent to one is straightforwardly engraved on to the GCLK this is a precarious conduct of the circuit [20]. In this proposed work, the SRAM and sequential circuits are designed to utilize the improved clock gating technique for reducing the leakage power while performing simulation and synthesis. Latch based approach performed with DFF and logics are triggered using clock switching activity. Here the performance is designed to get the better performances of area, delay and power. Here static and dynamic power reduction is done effectively to achieve the better results than existing work. This paper summarized as follows. Section II describes the various reviews related to the power optimization in VLSI circuits. Section III provides the proposed logic with constructive algorithms and novel designs. Section IV presents the results and discussion. Finally, section V concluded with the proposed logic and future enhancement.

2. Literature Survey

Zamin Ali Khan, et al. (2011) [8] has presented the power consumed VLSI design with power optimization approach, which utilized the genetic algorithm. Booth multiplier VLSI design is constructed to test the power by triggering gated switching logic. Here the GA is used to find out the different combination of gates logic onto the power estimation to deter-

VOLUME 15,

N° 2

2021

mine the fitness value; these consequently reduce the power. Benchmark ISCAS-89 circuit is used and performance analysis depends on the gated logics and power consumption on that circuit. Physical design of VLSI circuit utilized the chip analysis and optimization using GA. Bo-Cheng C Lai, and Jiun-Liang L, (2017) [2] has presented the multiport memory logic on the RAM design with FPGA implementations. The use of Block RAMs (BRAMs) is a basic execution factor for multiported memory plans on FPGAs. Not exclusively does the exorbitant request on BRAMs block the utilization of BRAMs from different parts of a plan, however the complex steering among BRAMs and interconnection likewise restricts the frequency range. This presents a shiny new viewpoint and a more proficient method of utilizing a regular two peruses one compose memory as a 2R1W/4R memory. By abusing the 2R1W/4R as the structure block, this presents a various leveled plan of 4R1W memory that requires 25% less BRAMs than the past methodology of copying the 2R1W module. Recollections with more read/compose ports can be reached out from the proposed 2R1W/4R memory and the various leveled 4R1W memory. Nandita S, et al. (2015) [4] has presented the clock gating technique on VLSI circuits for power reduction strategy. The framework is a coordination of fundamental structure contains sensor framework, control units into existing power frameworks which could be actualized as Silicon on Chip (SoC) in VLSI circuits. VLSI circuits can be both sequential and combinational. In consecutive circuits, the clock is the significant wellspring of dynamic power utilization. The method of clock gating is utilized to lessen the clock power utilization by removing the inactive clock cycles. VHDL-based strategy, to embed the clock gating circuit and furthermore the unique power because of this is assessed. C Ashok Kumar, et al. (2020) [1] has presented the loss minimization strategy of VLSI circuits to analyze the power. The solitary chip planned to reduce low territory frameworks for bringing more proficient gadgets, which are more modest in size. Circuit produced at a very high rate and they devour a space parcel more than they used. The premier worry of VLSI engineers was Area, power performance and Cost. Power has consistently been an optional concern. Latest thing has given more weight to Area, Power and Delay because of versatile specialized gadgets. The high velocity calculation gadgets with complex usefulness are a developing pattern which is request to low power utilization. Sreenivasulu, et al. (2016) [13] has presented the optimized sequential circuit to reduce the power leakage by multi threshold CMOS circuitry. This method designed to gives lower leakage current and offers upgraded speed. It utilizes low edge voltage gadgets for low leakage and high limit voltage segments as rest semiconductors. These rest semiconductors are sufficient to disconnect the rationale modules from the stock, ground to lessen the spillage current. Furthermore, the most un-conceivable time for turn ON state in a circuit is essential worries for power utilization. Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

Xuan-Thuan N, et al. (2019) [3] has proposed the RAM based content addressable memory architecture and exemplary methods for decrease of dynamic on FPGA using hierarchical partitioning approach. In influence, significant factor in all-out power utilization the clock cycle, the update phase of an RAM-based of any VLSI circuit. Clock gating procedure empowers content-addressable-memory consistently endures saving of electrical power utilized by PC processors. It high idleness. Two essential drivers of such inacpower by turning on a practical tivity guarantees include: (1) the saving mandatory eradicating stagelogic blocktheclock, yet juststage when The survey of alongside composing andrequired. (2) the significant existing clock gating procedures and its points distinction in information width between thefocal RAMfurthermore, impediments with a demo of D flip lemon based CAM and the advanced frameworks. The design of RAM-based twofoldRandom CAM updates the low latency. and 4-bit Pseudo Binary Sequence Generator. A few RCAMs,Awhose information width goeshas from 8 Jagadeeswaran, et al. (2012) presented to 64 the bits,optimized were incorporated intobased a 256-bit frame-pulse power level sequential work for the assessment. triggering approach. Flip-flops are the significant Priya Singh and Ravi Goel, (2014) [7] has presentstorge components in all SOC's. They oblige the ed the comprehensive study of power optimization majority of the power that has been applied to the chip. technique in sequential circuit. Here the clock gating Flip-floptois improve quite possibly the power parts. It is performed the result. Lowutilization power VLSI to decrease the leakage in both circuitisisessential the most basic issues in thepower present ASICclock circulation and flip-flops. The power delay plan, as the element size is downsized and there is is principally because of the clock delays. The deferral a pressing requirement for power advancement. of flip-flop ought the to be limited for effective usage. Clock the gating is perhaps most rich and exemplary methods decrease of dynamic signifThisfor venture moves around ininfluence, supplanting regular icant factor in all-out utilization any expert slave basedpower on D-FF to a pulseofset offVLSI flip flop circuit.which Clockgoes gating procedure empowers savingfor of low about as a recognition substitute electrical power utilized by PC processors. It guaranpower applications. In this, semiconductor sizes and tees power saving by turning on a practical logic block heartbeat age circuit can be further diminished for clock, yet just when required. The survey of existing power saving. Here UMC CMOS 180 nm innovation is clock gating procedures and its focal points furtherin SPICE device plane for theflip structure. more, use impediments with to a demo of D lemon and

4-bit Pseudo Random Binary Sequence Generator. A Jagadeeswaran, et al. (2012) [14] has present3. Proposed Method ed the optimized power level based sequential pulse triggering approach. Flip-flops are the significant The proposed designThey of SRAM storge components in all SOC’s. obligeand thesequential macounter circuit is used for analyzing the jority of the power that has been applied to the power chip. by using improved clockthe gating technique. Hereparts. the Static Flip-flop is quite possibly power utilization It is essential decrease the leakage in bothwith Random toAccess Memory circuitpower is designed clock improved circulation andgating flip-flops. The power delay isbased clock technique using D-latch principally clock The deferral bufferbecause circuitry of of the gated clockdelays. generation to reduce the of the power flip-flop ought toand be limited for effective dissipation the counter circuit ofusage. sequential This venture moves around in supplanting regular logic is designed with improved clock gating expert slave based on D-FF to a pulse set off flip flop technique. The input enabled logic is performed on the which goes about as a recognition substitute for low circuitsIn and it issemiconductor integrated withsizes the improved powerregister applications. this, and clockage gating technique. this, the clock andfor enable heartbeat circuit can be In further diminished signal applied on the gated clock generation module power saving. Here UMC CMOS 180 nm innovation is circuit is structure. enabled with the AND use inusing SPICED-Latch device to planeand foritthe logic. Based on the triggering state of the clock switching, the gated clock is generated to the SRAM

VOLUME 15,

N° 2

2021

tion module using D-Latch circuit and it is enabled with the ANDcounter logic. Based the the triggering and sequential circuits.onWhen circuit isstate in of the clock switching, the gated clock is generated idle condition, the clock switching state occurs with to the SRAM and sequential counter circuits. When triggering problem this causes increased leakage. the circuit is in idle condition, the clock switching Therefore, the clock gated switching control-based state occurs with triggering problem this causes inlogic is designed utilize thethe SRAM sequential creased leakage. to Therefore, clockand gated switchcounter logic circuit withis improved gating ing control-based logic designed clock to utilize the technique. The proposed block diagram of power SRAM and sequential counter logic circuit with imreducingclock technique SRAM andThe sequential circuits proved gatingoftechnique. proposed block diagram reducing technique of SRAM and are using of thepower improved clock gating technique is given sequential in Figure 1. circuits are using the improved clock gating technique is given in Figure 1. Clk en

Improved clock gating gclk

SRAM circuit Input ‘A’

Sequential counter circuit

Power analyzer Performance results

Fig. 1. Proposed block diagram of power optimized Fig. 1. Proposed block diagram of power optimized SRAM and sequential circuits SRAM and sequential circuits The VLSI circuit optimization involved with various technique forcircuit improving the performance. The VLSI optimization involved withBy changing the width length the of the transistor,By the various technique for and improving performance. minimum impact is possible on the VLSI layout circhanging the width and length of the transistor, the cuit. While connecting source and load impedance, minimum impact is possible on the VLSI layout the buffer is used to reducing the number processing circuit. While connecting source and load impedance, stages. the buffer is used to reducing the number processing stages.

3. Proposed Method

The proposed design of SRAM and sequential counter circuit is used for analyzing the power by using improved clock gating technique. Here the Static Random Access Memory circuit is designed with improved clock gating technique using D-latch based buffer circuitry of gated clock generation to reduce the power dissipation and the counter circuit of sequential logic is designed with improved clock gating technique. The input enabled logic is performed on the register circuits and it is integrated with the improved clock gating technique. In this, the clock and enable signal applied on the gated clock generaArticles

Fig. 2. RTL view of Bidirectional counter with Improved Clock Gating Technique The RTL schematic of Bidirectional counter circuit is shown in Figure 2. The load decoupling is also helpful for reducing the critical path on circuitry. By strengthening the switching activity, the clock signal variation may get better computation. The static power dissipation of CMOS circuit is reduced by capacitance utilization with better transition activity and the pre-charging the higher capacitance to improve the speed.

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

3.1 SRAM Circuit Synchronous RAM is designed with the improved logic of gated clock generation circuit to reduce the power leakage. D Latch registry with level delicates the positive lock passes contribution to yield on high stage, hold time on low enable states. Here the positive register tests on rising edge with sample input and the edge triggering is performed with the flip flops. Static power optimization with clock gated signal generation triggers the logic by switching activity and adding of BUFF logic on gclk. Here the Static power of RAM is controlled and optimized the power with effective utilization.

Fig. 3. SRAM circuit RTL schematic SRAM logic structure is designed to utilize the clock gating approach with RTL view (Figure 3). The clock gating strategy has been created to evade leakage power. When the framework is inactive, the clock switching timing of clock gating is TURN OFF. Explicitly for flip-flops, clock gating implies debilitating the clock signal when the info information doesn’t change away the information. It may be applied from the framework level where the whole useful unit can be specifically set into rest mode, or from the sequential circuit level where a few bits of the circuit are in rest mode while the rest of the logic blocks are working. In any case, Clock gating doesn’t come free of charge. Additional logics and interconnects are needed to produce the clock empowering signals.

3.2 Sequential Counter Circuit

Sequential counter circuit with bidirectional logic designed with the improved clock gating technique to reduce the power leakage. Here the RTL gate level performance is used for the gated clock circuits to improve the performance. Register with BUFFER utility on D-FF is shown in RTL view (Figure 4).

Fig. 4. RTL view of BUFF added D-FF Power consumption techniques mainly focus the static and dynamic to improve the performance of VLSI circuits. Aside from this short out and leakage assumes a crucial part in fixing the general energy utilization in a circuit. Static power is because of the quantity of intelligent a part utilized in the module and dynamic happens as aftereffect of number of transient states in the logics. Short circuit flows additionally happen in this situation of when both NMOS and PMOS semiconductors are in ON active state. Capacitive activity prompts dynamic energy utilization, which is most moving issue to bargain with as exchanging movement is high in consecutive circuits. If the circuit allows for timing logic with switching activity, the clock edges are reducing its existence on precharging state of capacitance.

3.3 Modified Clock Gating Technique

At the point when the present and next condition of the D flip failure is noticed, it is seen that when two nonstop information sources are indistinguishable, the D flip failure gives a similar incentive as yield. The clock cycle that is taken care of into the D flip lemon when the yield doesn’t shift is named as inactive clock cycles. To eliminate these, at the point when the flip lemon are of various qualities the EXOR entryway passes a yield 1 which is given as a yield to the AND entryway alongside a clock beat gave to the flip lemon for additional exchanging movement that happens during various patterns of the clock beats.

Pd = Cs (Vcc ) f clk 2

Where the cumulative value of switching clock is denoted as Cs and f_clk mentioned the frequency of clocking to reduce the dynamic power ‘Pd’ and Vcc denoted as power supply utility to the module.

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

Tab. 1. Area utility of SRAM logic with power optimization strategy SRAM-Device utilization Summary Slice Logic Utilization

Used

Slice Registers

Fig. 5. Block diagram of improved clock gating with D-FF The Clock gating wit DFF is given in the Figure 5. Setup time and hold time are continually influenced by their contributions as long as empower signal is affirmed. They are empowered, their substance changes promptly when their sources of info change. Flip-flops, then again, have their content change just either at the rising or falling edge of the empower signal. This empower signal is normally the controlling clock signal. After the rising or falling edge of the clock, the flip-flop content remaining parts steady even.

4. Results and Discussion

Thus, the design model of SRAM and sequential logic of counter circuit is effective utilization. Modification in clock gating technique reduces the leakage power than traditional approaches. The module comprises of one 8-input combinational logic-into table with four devoted registers and is viewed as a major structure of FPGA block.

Fig. 6. Simulation result of SRAM circuit The simulation result of SRAM is shown in Figure 6. Here the gated clock is generated and determines the better utility of performance. The –bit utilization memory is constructed to improve the synchronous logics and it is clock gating is improved the performance of area and delay utility. Here the Tables 1 and 2 show the results of area and delay.

Articles

Available

Utilization

93,120

Slice LUTs

46,560

Number used as Memory

16,720

Number used as logic Number of occupied Slices Number with an unused Flip Flop No. of fully used LUT-FF pairs slice register sites lost to control set restrictions Bonded IOBs

BUFG/BUFGCTRLs

Tab. 2. Delay report

2 3 1

46,560 11,640

1% 1%

16%

93,120

240

83% 11% 6%

Parameters of delay

Value

Minimum period:

1.257ns

Output required time after clock –maximum reach

1.148ns

Input arrival time before clock with maximum delay utility path delay - gclk Clock period Net delay

1.324ns

0.935ns 1.257ns 0.533ns

The power analysis is the major concern in proposed logic, which utilizes the better utilization on the synchronous RAM with XPower Analyzer tool in Xilinx ISE. Both Static and dynamic power is analyzed with the voltage and current utility, which is shown in Figure 7. When the register activity is TURN OFF, the input data need to register the gated clock and clock gets OFF state. The enable and clock applied on the register block, this is the enable clock gating signal.

Fig. 7. XPower analyzer result of SRAM

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

Sequential circuit power reduction is determined the power, area and delay. The ISim result of sequential logic with bidirectional counter is designed and it is shown in Figure 8. Fig. 9. XPower analyzer report of Sequential circuit Tab. 5. Comparison results Experiment

Fig. 8. Simulation result of Bidirectional counter with power optimal strategy Tab. 3. Device utilization of bidirectional-counter circuit Device Utilization Summary Slice Logic Utilization Number of Slice Registers Number of Slice LUTs

Number used as logic

Number used as Memory

Number of occupied Slices

Number with an unused Flip Flop Number with an unused LUT

Number of fully used LUT-FF pairs Number of slice register sites lost to control set restrictions Number of bonded IOBs

Number of BUFG/BUFGCTRLs

Used

Available

93,120

46,560

11,640

0 2 0 9

16,720

1% 0% 1%

18%

81%

240

10%

93,120

Utilization

0% 1% 3%

Device utilization summary determines the utility of area, which is given in Table 3. Various delays like path delay, net delay and gate delays are analyzed and it is given in Table 4. Tab. 4. Delay report of sequential circuit Parameter

Value

gclk

5.859ns

input arrival time before clock

1.551ns

Clk

Maximum output required time after clock Maximum combinational path delay Net delay

1.765ns 1.148ns 0.935ns 1.244ns

Bidirectional counter circuit uses the improved clock gating technique for power analysis, which is shown in Figure 9. Shifting and DFF register utilities are the major logic block of RTL view. Clock pair shared flip-flop based logic, low-swing DFF based logic and gated model of SRAM designs are compared to the proposed logic and it achieves the best result for proposed power reduction technique, which are shown in Table 5.

Power (W) [7]

Delay(ns) [21]

4-Bit Counter(no CG)

22.96

D flip flop (no CG)

14.299

4-Bit Counter(with CG) D Flip flop (with CG)

17.93

14.202

49 68

PRBS (no CG)

13.911

112

Proposed power reduction logic

1.065

PRBS (with CG)

12.5

As one of the most highly desired VLSI design fields, power consumption has grown to become one of the central study areas. Based on the connection between the triggering transition of the clock and the present and the next state function of the flip flop, the clock gating method that has been suggested would gate clocks. In latch-based clock gating procedure, a sensitive latch is utilized as the control component, to control the Enable pin, that is taken care of to the “AND” “OR” gate level for gating the clock signal. This latch is permitted to mirror the difference in Enable pin. The clock holds the estimation of empower signal from the dynamic edge of the clock till the idle edge of the clock. In the event that the “AND” is utilized for circuits working on certain edge of clock pulse. Therefore, the proposed clock gating based SRAM and Sequential circuit improves the result of area, delay and power than existing work.

5. Conclusion

Thus, it concludes that the design logic of SRAM and sequential counter circuit is improved the performance using improved clock gating technique. Here the power reduction strategy is performed with D-Latch based clock switching with triggering of RTL module in improved clock gating technique. The consequence of this kind of clock gating procedure on a D flip failure is as appeared. Power improvement, generally consigned to the combination, and situation and directing stages, has climbed to the System level also, RTL. HDL can thus utilize clock gating to turn off idle segments of the plan and decrease generally speaking dynamic power utilization. Thus, the RTL view of gate level examination determines the better result of SRAM and sequential Bidirectional counter circuit using improved clock gating approach. In future, the work may extend with the voltage limiting and managing approach for power reduction. Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

AUTHORS C. Ashok Kumar* – Department of Electronics and Communications Engineering, MLR Institute of Technology, Hyderabad, Telangana, India, Email: ashokkumar.cheeli@mlrinstitutions.ac.in. B.K.Madhavi – Siddhartha Institute of Engineering and Technology, Hyderabad, Telangana, India, Email: bkmadhavi2008@gmail.com. K. Lal Kishore – CVR College of Engineering, Hyderabad, Telangana, India, Email: lalkishorek@gmail. com. *Corresponding author

REFERENCES

[12]

[13]

[1]

C. Ashok Kumar, B. K. Madhavi and K. Lal Kishore, “Methods and Analysis for Low Power VLSI Design”, The International Journal of Analytical and Experimental Modal Analysis, vol. 12, no. 1, 2020, 3427-3435. [2] B.-C. C. Lai and J.-L. Lin, “Efficient Designs of Multiported Memory on FPGA”, IEEE Trans. on Very Large Scale Integr. (VLSI) Syst., vol. 25, no. 1, 2017, 139-150, 10.1109/TVLSI.2016.2568579. [3] X.-T. Nguyen, T.-T. Hoang, H.-T. Nguyen, K. Inoue and C.-K. Pham, “An Efficient I/O Architecture for RAM-Based Content-Addressable Memory on FPGA”, IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 66, no. 3, 2019, 472476, 10.1109/TCSII.2018.2849925. [4] N. Srinivasan, N. S. Prakash, Shalakha D., Sivaranjani D., S. Sri Lakshmi G. and B. B. T. Sundari, “Power Reduction by Clock Gating Technique”, Procedia Technology, vol. 21, 2015, 631-635, 10.1016/j.protcy.2015.10.075. [5] J. Shinde and S. S. Salankar, “Clock gating — A power optimizing technique for VLSI circuits”. In: 2011 Annual IEEE India Conference, 2011, 1-4, 10.1109/INDCON.2011.6139440. [6] J. Monteiro and S. Devadas, “Optimization Techniques for Low Power Circuits”. In: Computer-Aided Design Techniques for Low Power Sequential Logic Circuits, 1997, 81-96, 10.1007/978-1-4615-6319-8_5. [7] P. Singh and R. Goel, “Clock Gating: A Comprehensive Power Optimization Technique for Sequential Circuits”, International Journal of Advanced Research in Computer Science & Technology, vol. 2, no. 2, 2014, 321-324. [8] Z. A. Khan, S. M. Aqil Burney, J. Naseem and K. Rizwan, “Optimization of Power Consumption in VLSI Circuit”, International Journal of Computer Science Issues, vol. 8, no. 2, 2011, 648-653. [9] S. Nireekshan Kumar and J. Grace Jency Gnannamal, “Delay and Power Optimization of Sequential Circuits through DJP Algorithm”. In: Proc. of the World Congress on Engineering, vol. 1, London, U.K, 2008. [10] J. Monteiro, J. Rinderknecht, S. Devadas and A. Ghosh, “Optimization of combinational and

[11]

Articles

[14]

[15]

[16] [17] [18]

[19]

[20]

[21]

N° 2

2021

sequential logic circuits for low power using precomputation”. In: Proc. 16th Conference on Advanced Research in VLSI, 1995, 430-444, 10.1109/ARVLSI.1995.515637. P. Zhao, Z. Wang and G. Hang, “Power optimization for VLSI circuits and systems”. In: 10th IEEE International Conference on Solid-State and Integrated Circuit Technology, 2010, 639-642, 10.1109/ICSICT.2010.5667299. S.-H. Weng, Y.-M. Kuo and S.-C. Chang, “Timing Optimization in Sequential Circuit by Exploiting Clock-Gating Logic”, ACM Transactions on Design Automation of Electronic Systems, vol. 17, no. 2, 2012, 1-15, 10.1145/2159542.2159548. P. Sreenivasulu, K. Srinivasa Rao and A. Vinaya Babu, “Optimizing Power in Sequential Circuits by Reducing Leakage Current using Enhanced Multi Threshold CMOS”, Indian Journal of Science and Technology, vol. 9, no. 36, 2016, 10.17485/ijst/2016/v9i36/102601. A. Jagadeeswaran and C. N. Marimuthu, “Power Optimization Techniques for Sequential Elements Using Pulse Triggered Flip-Flops with SVL Logic”, IOSR Journal of VLSI and Signal Processing, vol. 1, no. 4, 2012, 31-36, 10.9790/42000143136. R. Samanth, C. Chaitanya and G. S. Nayak, “Power Reduction of a Functional unit using RT-Level Clock-Gating and Operand Isolation”. In: 2019 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER), 2019, 1-4, 10.1109/DISCO VER47552.2019.9008025. D. Mahesh Kumar and R. Kannusamy, “An Efficient Design of Low Power Sequential Circuit Using Clocked Pair Shared Flip Flop”, Int. J. Appl. Eng. Res., vol. 12, no. 2, 2017, 233-237. D. Kumar Sharma, “Effects of Different Clock Gating Techinques on Design”, Int. J. Sci. Eng. Res., vol. 3, no. 5, 2012. A. Nag and S. N. Pradhan, “An Autonomous Power and Clock Gating Technique in SRAM-Based FPGA”. In: V. Nath (eds.), Proc. of the International Conference on Nano-electronics, Circuits & Communication Systems, vol. 403, 2017, 1-14, 10.1007/978-981-10-2999-8_1. L. Sterpone, L. Carro, D. Matos, S. Wong and F. Fakhar, “A new reconfigurable clock-gating technique for low power SRAM-based FPGAs”. In: 2011 Design, Automation & Test in Europe, 2011, 1-6, 10.1109/DATE.2011.5763128. M. Tamilselvi, P. Vedhanayagi and K. Ramasamy, “Implementation of 13T SRAM Using Power Gated Techniques”, International Journal of Advanced Research in Electrical, Electronics and Instrumentation Engineering, vol. 6, no. S1, 2017, 147-156. M. Janaki Rani, “Leakage Power Reduction and Analysis of CMOS Sequential Circuits”, International Journal of VLSI Design & Communication Systems, vol. 3, no. 1, 2012, 13-23, 10.5121/ vlsic.2012.3102.

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

A Computational Model for Multi-Criteria Decision Making in Traffic Jam Problem Submitted: 19th May 2021; accepted: 6th September 2021

Ali Naeem, Jabbar Abbas DOI: 10.14313/JAMRIS/2-2021/12 Abstract In this paper, we apply a computational model for multicriteria decision making in traffic jam problems. First, we propose a system to determine the optimal shortcut road by reading the number of cars in each street using Radio Frequency Identification (RFID). Then, we have processed the data of traffic jam problems using Choquet integral with writing algorithm and computer program as a working procedure. Keywords: Multi-Criteria Decision Making, Fuzzy Measures, Choquet Integrals, Traffic Jams

1 Introduction

2 Multi-Criteria Decision Making Problem Consider a multi-criteria decision making problem that depends on n criteria (or points of view, attribute) described by the alternatives G1,…,Gn and a set of criteria X= {1, …, n}. An alternative is characterized by a value with respect to each criterion and is thus identified with a point in the Cartesian product G of the criteria, i.e. G = G1 ×… × Gn is the set of potential alternatives. The preference relation of the Decision Maker (DM) over alternatives is denoted by ⪰ . For any g, h ∈ G, g ⪰ h means that the DM prefers alternative g to h. A classical way is to model ⪰ with the help of an overall utility function z: G→R [1]:

∀ g , h ∈ G,

g ⪰h ↔ z ( g ) ≥ z ( h ) .

(1)

Traffic jam is the worst problem that drivers are sufClearly, z is a n dimensional function. An easy way fering in their daily lives, both in terms of delays in to construct z is to consider one-dimensional utility the traffic jams when going to work or the inconvefunction zi on each criterion and then to aggregate nience caused by road jam in cars on public holidays. them by a suitable operator: The problem of traffic congestion is not confined to = z ( g ) F z 1 ( g 1 ) , , z n ( g n ) ∀g ∈G regions without other regions, and thus countries al(2) ways seek to address it in all kinds of solutions. The theory of multi-criteria decision-making has gained where, F is called an aggregation function. the emergence of several new paradigms in the secAggregation functions (AFs) are mathematical ond half of the twentieth century (for more details functions to collect helpful data in multi criteria desee e.g. [1]). There are several multi-criteria decisioncision making. The input of AFs is several numerical making approaches that have been introduced to values and its output is a single value. study the problem of traffic congestion (see e.g. [2]). Based on the overall score by means of an aggreThe aim of this paper is to apply a computational gation function that takes into account the weights model for multi-criteria decision making in traffic jam of importance of the criteria, the alternatives can be problems, then use a system as a working procedure, ranked and the best alternative selected. A special to determine the optimal shortcut road using Radio type of aggregation function is fuzzy integral with Frequency Identification (RFID). The RFID system respect to non-additive measure. The fuzzy intehas already been adopted into various application argral [7], [8] is an appropriate tool to represents the eas (see, [3]-[6]). weights of criteria with non-additive measures. One In this paper, we propose a system that processed of the fuzzy integrals is Choquet integral with respect the data of traffic jam problems using Choquet inteto non-additive measure (fuzzy measure). The fuzzy gral to give the best result of the most appropriate integrals with respect to non-additive measures have road (optimal shortcut road) to go from the start been studied and applied in diverse fields (see, e.g. point to the endpoint. [9-17]). There are many types of non-additive measThe structure of this paper is as follows. Section 2 ures, one of them is the ℷ-fuzzy measure. The definiintroduces Multi Criteria Decision Making problem. In tion of ℷ-fuzzy measure is as follows. section 3, we propose a system to determine the optimal shortcut road in traffic jams. In section 4, we give Definition 1: Suppose P(G) be the power set of G, a practical example. In section 5, we finish the paper A set function is called ℷ-fuzzy measure if it satisfies with some conclusions. the following axioms: 1. g ℷ : P (G ) → 0,1 ∅ , then 2. If S ,T ∈P (G ) , S ∩T =

(

)

Journal of Automation, Mobile Robotics and Intelligent Systems

g g (S ∪ T=) g g (S ) + g g ( T ) + g g g (S ) g g ( T ) ,

(

1 g  i =1 n

)

(

(3)



)

1= 0 (6) g g {x 1 ,…, x n } =  ∏ 1 + gg g ( x i ) − 1  , g = 

This given a polynomial equation with respect to ℷ 1= +g

( 1 + g g ( x i )). ∏ i =1

(5)

The mapping f : G → R+ corresponds to the value that the supply of the source (i.e., f(xi ) = xi ), using, as before, xi to denote the ith input value and the fuzzy measure (gℷ : P(G) → [0, 1]) assigns importance to subsets of G.

Definition 2: Suppose that gℷ is fuzzy measures on G, whose elements are, refer to (x1, …, xn). The discrete choquet integral of a function f : G → R+ with respect to gℷ is defined by ∫C g ( = f) g

( f ( x (i ) ) − f ( x (i ∑ i N

−1)

)) g ( A( ) )

(6)

( )

{x ( ) ,…, x ( )} i

Arduino. An electronic development board (as in Figure 1), can be used by both professionals and beginners, and it is used in different fields. Arduino can communicate with the surrounding environment through a number of sensors and can influence its surroundings by controlling small motors or lights and other electronic parts. Arduino’s projects can be connected to its sensors and electronic parts only or can be connected to computer-based programs, such as Visual BASIC. Radio Frequency Identification (RFID). RFID is one of the most common applications in the recent period (Figure 2). It is widely found in the security system and is done by sending electromagnetic waves to special cards (Figure 3). The hardware design of the system for solving the problem of traffic jams is shown in the following figures (Figure 4 and Figure 5).

with f x (0) = 0

Fig. 1. Arduino Uno board

3 A Computational Model for Multi-Criteria Decision Making in Traffic Jam Problem 3.1 A System for Solving the Problem of Traffic Jams

In order to solve the problem of traffic jams, we must first try to understand its causes. While the causes and may differ from city to city, there are some common factors. For example, the increase in the number of cars and narrow roads are obvious reasons. Hence the idea of the solution of traffic jams through which the street jam is identified by knowing the number of cars in each street of the city that the user wants to pass. This process is done using the Radio Frequency Identification (RFID) technique, through which scores every car entering and go out the specified road. The procedure of statistical work for the proposed roads and choose the most appropriate road among these roads through the processing of data using multi-criteria decision making problems. Thus, the user can query the best road through an application designed for this purpose that is available for all smart mobile devices. Some countries use GPS technology to define the number of cars and regardless of the economic cost of this technology, where the cost of GPS is more than the cost of RFID. Furthermore, it is not available in all cars and it’s difficult for the normal driver to install it comparing with the RFID technology used in our system. Articles

2021

The Hardware design of the system can be classified into the following.

where (i) denotes that the indices have been permuted so that = 0 ≤ f x (1) ≤  ≤ f x ( n ) , and A (i ) :

N° 2

3.2 Hardware Design of the System

g ∈ ( −1,∞ )

In general, it can be shown that

VOLUME 15,

Fig. 2. RFID board

Fig. 3. Tags (RF Sender)

Fig. 4. Hardware for a small model

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

Fig. 5. A small city map

Fig. 9. Before choosing a path

Fig. 6. A small car with tag inside In the practical application will get the values of three criteria (Length, Time, and Traffic jam) automatically through the RFID reader (Figure 6, Figure 7) which will calculate the number of cars on the road, and calculate the rate of cars on the road by dividing the number of cars on the street capacity, and this value will represent the road jam. When the user wants to choose the best road, the application will draw the best streets on the map (Figure 10). The length of the street will be taken from the Google maps application, and the time will depend on the street length and the traffic jam. The hardware design of the system tools is showing in the following figures (Figure 6- Figure 9).

Fig. 10. The shortest path

3.3 Algorithm To simplify the notation in the algorithm, we use the following equation to calculate the discrete Choquet integral that is equivalent to the equation (6): ∫C g ( f ) = g

Fig. 7. Arduino connection

Fig. 8. Small city simulator

[x (i ) − x (i ∑ i =1

−1)

( )

] g g A( i )

where x(0) = 0 by convention, and A(i) = {1, …, n} Consider the vector of pairs ((x1, 1), (x2, 2), …, (xn, n)), where the second component of each pair is just the index i of xi. The second component will help to keep track of all permutations. Therefore, we use the following procedure for the calculation of the discrete Choquet integral Cg g (f ) .

Step 1: Sort the components of ((x1, 1), (x2, 2), …, (xn, n)) with respect to the first component of each pair in non-decreasing order. We obtain, ((x(1), i1), (x(2), i2), …, (x(n), in)), so that x(j) = xij and x(j) ≤ x(j+1) for all i. Let also x(0) = 0. Step 2: Let T = {1, …, n}, and S = 0. Step 3: For j = 1, …, n do a) S := S + [x(j)− x(j−1)] gℷ (T); b) T := T \{ij}

Step 4: Return S.

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

4 Practical Example We consider the problem of evaluation of traffic jams on roads with respect to three criteria, Length (L), Time (T), and Traffic Jam (J). This is usually done through a simple weighted amount, in which weights are coefficients of the importance of different criteria. Let us take some areas on the map that contain five roads (R1, R2, R3, R4, R5) to move from point (A) to point (B). The data of the five roads are given in the following table (Table 1). Tab. 1. The data of the five roads (R1, R2, R3, R4, R5) R1

Length (L)

Time (T)

Traffic Jam(J)

300

180

150

200

100

150

We can solve this problem by using a suitable fuzzy measure and the Choquet integral as follows.

1. The Length and Traffic jam criteria are more important than Time, we put the following weights on criteria taken individually.

(

)

(

)

(

)

= g g {Length} 0.7, = g g {Time} 0.5, g g = {Traffic Jam} 0.6.

Using equation (5), we get

(1 + 0.7g)(1 + 0.5g)(1 + 0.6g)

1+ g=

Then, the solutions of this equation are ℷ=–0.9103, ℷ =–4.1849, and ℷ=0. The only acceptable value is ℷ =–0.9103, but the other values are violated of constraints of ℷ-fuzzy measure. 2. Since length and traffic jam are redundant, the weight attributed to the length and the traffic jam should be less than the total weights of the length and traffic jams. By applying equation (3), g g {x 1 , x 3 } = ,

(

)

( ) ( ) ( ) ( ) g ({x , x } ) = g ({length , traffic jam} )

g g { x 1 } + g g {x 3 } + g g g {x 1 } g g {x 3 } , 1

= 0.8 + 0.5 + ( −0.9439 *0.8 *0.5 )= 0.917674

(

)

g g {length , traffic jam = } 0.917674 < 0.8 + 0.5 .

Similarly, the weight attributed to the time and length should be greater than the sum of individual weights (also, the same for time and traffic jam).

( (

) )

( (

)

g g {x 1 ,= x 2 } g g {length, time = } 0.881395 < 0.8 + 0.6, g g {x 2 , x= g g {time, traffic jam = } 0.82691 < 0.6 + 0.5, 3}

(

)

g g ( ∅ ) 0, = g g {length , traffic jam, time} 1. 42

Articles

N° 2

2021

3. By applying ℷ-fuzzy measure and Choquet integral for the first road (R1) we get the following result for (R1). 0 ≤ f ( x 3 ) ≤ f ( x 2 ) ≤ f ( x 1 ) , ∫C g ( = f) g

( f ( x (i ) ) − f ( x (i ∑ i N

−1)

)) g ( A( ) ) g

∫C g ( R 1 ) =f ( x 3 ) − f ( 0 )  g g ( x 1 , x 2 , x 3 ) g + f ( x 2 ) − f ( x 3 )  g g ( x 1 , x 2 ) + f ( x 1 ) − f ( x 2 )  g g ( x 1 )

∫C g ( R 1 ) =55 − 0 *1 + 100 − 55 *0.88139 g

+ 200 − 100 *0.7 = 164.66

Other results for the remaining roads (R2, R3, R4, R5) are shown in Table 2.

VOLUME 15,

Tab. 2. Choquet integral for the roads (R1, R2, R3, R4, R5) Length (L)

Time(T)

Jam (J)

Global evaluation (Choquet integral)

Road 2

300

180

249.17

Road 4

150

Road 1 Road 3

Road 5

200

150

100

150

50 60

55 70

80 90

164.66 124.35

127.35

130.35

The interpretation of Choquet integral with respect to scoring (from lower to above) on the road ranking confirms that the preference relation of the Decision Maker is R3  R4  R5  R1  R2 . Therefore, Road 5 is the optimal shortcut road according to the ranking roads. Thus, we get the predictable solution of the traffic jam problem for the appropriate road (optimal shortcut path) to go from the start point to the endpoint.

5 Conclusion

In this paper, we have described a computational model of the problem of multi-criteria decision making in traffic jams. Then, we have proposed in this model a system for solving the Multi-Criteria Decision Making Problem of traffic jams. The application of this system is high speed in data collection and processing, also the mechanism of the system is automatic. In the future, we plan to test our model in many areas, such as service, commercial, and security fields.

AUTHORS

Ali Naeem – Department of Applied Sciences, University of Technology, Baghdad, Iraq, Email: aliallamy. itu@gmail.com. Jabbar Abbas* – Department of Applied Sciences, University of Technology, Baghdad, Iraq, Email: 100033@uotechnology.edu.iq. *Corresponding author

Journal of Automation, Mobile Robotics and Intelligent Systems

REFERENCES [1] [2]

[3]

[4]

[5]

[6]

[7] [8] [9]

J. Figueira, S. Greco and M. Ehrogott, Multiple Criteria Decision Analysis: State of the Art Surveys, Springer New York, 2005, 10.1007/b100605.

N. Hao, Y. Feng, K. Zhang, G. Tian, L. Zhang and H. Jia, “Evaluation of traffic congestion degree: An integrated approach”, Int. J. Distrib. Sens. Netw., vol. 13, no. 7, 2017, 10.1177/1550147717723163.

D. Yue, X. Wu and J. Bai, “RFID Application Framework for pharmaceutical supply chain”. In: 2008 IEEE International Conference on Service Operations and Logistics, and Informatics, 2008, 1125–1130, 10.1109/SOLI.2008.4686568.

A. Hamdan, M. Chen and K. J. Rogers, “RFID application in the third-party logistics industry”. In: 2006 Technology Management for the Global Future - PICMET 2006 Conference, 2006, 2769– 2795, 10.1109/PICMET.2006.296871.

D. Manik, L. Toth and P. Dobrossy, “Analysis of RFId Application Through an Automotive Suppliers Production Processes”. In: 2007 International Symposium on Computational Intelligence and Intelligent Informatics, 2007, 177–181, 10.1109/ISCIII.2007.367385.

S.-W. Wang, W.-H. Chen, C.-S. Ong, L. Liu and Y.-W. Chuang, “RFID Application in Hospitals: A Case Study on a Demonstration RFID Project in a Taiwan Hospital”. In: Proc. of the 39th Annual Hawaii International Conference on System Sciences (HICSS’06), 2006, 10.1109/HICSS.2006.422.

VOLUME 15,

N° 2

2021

Soft Computing. ICAISC 2019. Lecture Notes in Computer Science, vol. 11508, 2019, 287–295, 10.1007/978-3-030-20912-4_27.

[13] R. I. Sabri, M. N. Mohammedali and J. Abbas, “An Application of Non-additive Measures and Corresponding Integrals in Tourism Management”, Baghdad Science Journal, vol. 17, no. 1, 2020, 172–177, 10.21123/bsj.2020.17.1.0172.

[14] M. Hesham and J. Abbas, “Multi-criteria Decision Making on the Best Drug for Rheumatoid Arthritis”, Iraqi Journal of Science, vol. 62, no. 5, 2021, 1659–1665, 10.24996/ijs.2021.62.5.28. [15] J. Abbas, “Shilkret Integral Based on Binary-Element Sets and its Application in The Area of Synthetic Evaluation”, Engineering and Technology Journal, vol. 33, no. 3, 2015, 571–577.

[16] J. Abbas, “Logical Twofold Integral”, Engineering and Technology Journal, vol. 28, no. 3, 2010, 477–483. [17] F. Kareem and J. Abbas, “A Generalization of the Concave Integral in Terms of Decomposition of the Integrated Function for Bipolar Scales”, Journal of Applied Sciences and Nanotechnology, vol. 1, no. 4, 2021, 81–90, 10.53293/ jasn.2021.3985.1065.

G. Choquet, “Theory of capacities”, Ann. Inst. Fourier, vol. 5, 1954, 131–295, 10.5802/aif.53.

M. Sugeno, “Theory of fuzzy integrals and its applications”, Ph. D. dissertation, Tokyo Institute of Technology, Tokyo, 1974.

A. Mendez-Vazquez, P. Gader, J. M. Keller and K. Chamberlin, “Minimum Classification Error Training for Choquet Integrals With Applications to Landmine Detection”, IEEE Trans. Fuzzy Syst., vol. 16, no. 1, 2008, 225–238, 10.1109/ TFUZZ.2007.902024.

[10] J. Abbas, “The Bipolar Choquet Integrals Based on Ternary-Element Sets”, Journal of Artificial Intelligence and Soft Computing Research, vol. 6, no. 1, 2016, 13–21, 10.1515/jaiscr-2016-0002.

[11] J. Abbas, “The Balancing Bipolar Choquet Integrals”, Int. J. Innov. Comput., Inf. Control, vol. 17, no. 3, 2021, 949–957, 10.24507/ijicic.17.03.949.

[12] J. Abbas, “The 2-Additive Choquet Integral of Bi-capacities”. In: L. Rutkowski, R. Scherer, M. Korytkowski, W. Pedrycz, R. Tadeusiewicz and J. M. Zurada (eds.), Artificial Intelligence and

Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

VOLUME 15,

N° 2

2021

A New Approach to an Achievement Motivation System for the Choice of an Engineering High School and Field of Study Submitted: 21st July 2021; accepted: 20th September 2021

Josef Malach, Dana Vicherková, Milan Chmura, Veronika Švrčinová DOI: 10.14313/JAMRIS/2-2021/13 Abstract: Achievement motivation is an important prerequisite for students’ engagement, overcoming study difficulties and, ultimately, successful completion of their studies. The aim of this study is to find out the current level of performance motivation of the population of students of engineering high schools, to compare its level with the standard for high school population and to find out whether there is a relationship between its level and choice. Two questionnaires were used in the research. The first was the School Achievement Motivation Questionnaire for Pupils. The theory of achievement motivation is then based on the concept of independence of the need for successful performance and the need to avoid failure. The resulting orientation of a person in a performance situation then depends on the predominance of one or the other tendency. The second questionnaire was focused on finding personal data, questions of school choice, satisfaction with school choice and other. The research was carried out at the beginning of 2020 and the sample consisted of over 900 students. Main results of the study: First, in the component of performance motivation NACH (need to achieve) the performance motivation of the group is statistically different from the group norm. In the NAF (the need to avoid failure) component, the results are statistically identical. Second, schools do not differ in the results for the NACH component, but differ in the NAF component. Third, in both components, the value of performance motivation for beginners and for final years in both components is significantly different from other years. Fourth, the student’s statement on the choice of study at an engineering school as a primary choice is not related to the values of NACH and NAF. Fifth, the level of student’s achievement motivation in both components is related to the student’s idea of studying. Keywords: Engineering High School Education, School Achievement Motivation, Need for Successful Performance, Need to Avoid Failure, Choice of School, Concept of Study

1 Introduction 1.1 Theoretical concept

Students’ achievement motivation is an important prerequisite for their engagement, overcoming study difficulties and, ultimately, successful completion of

their studies. The motivation to achieve goals not only leads individuals to pursue work they perceive to be valuable, it also prompts them to compete with others [2]. This drive may come from an internal or external source. Achievement motivation is intrinsic when it is sparked by an interest or enjoyment in the task itself. It is organic to the person, not a product of external pressure. Achievement motivation can be instead extrinsic when it comes from outside the person. Common sources of extrinsic motivation among students are rewards like good marks, or praise from parents and teachers. [9] Hustinx et al. [5] summarize the views of older authors (McClelland et al. [7]; Heckhausen [3]; and others) in the thesis: “Individuals with a high achievement motivation set standards of excellence, show clear affect in connection with evaluation of their performance, and display a high level of aspiration in terms of achievement goals” (p. 561). Hustinx et al found out by research, that achievement motivation shows a certain degree of stability, but this stability decreased to rather low values when intervals between measurements increased up to four years. Alternatively, it is possible that achievement motivation, even though it may have some characteristics of a stable trait, is a characteristic that partially depends on, and is activated by, situational factors.” (p. 576) According to research Steinmayr and Spinath [18] motivation constructs nearly explained as much unique variance in general school performance as intelligence. The achievement motivation can be treated as an important psychological predictor of graduates’ future success or failure and according to McClelland and other authors it should be intentionally trained. [6] Theories of achievement motivation have significantly evolved over the last several decades, and research grounded in these theories influences and informs teaching practices, parent involvement activities in schools, and educational interventions targeted at students, administrators, teachers, and parents. [8] Since the 1950s, performance motivation has been in the focus of pedagogical and psychological research, which usually examines it as an independent phenomenon of the school population. For high engineering school students, motivation is also a prerequisite for obtaining professional qualifications for the professions that are usually defined by national qualifications frameworks.

Journal of Automation, Mobile Robotics and Intelligent Systems

1.2. Current Research on Achievement motivation of secondary school students In all PISA countries and economies except Belgium and Switzerland, disadvantaged students have lower levels of achievement motivation than advantaged students. On average across OECD countries, immigrant students reported higher achievement motivation than non-immigrant students. Achievement motivation is positively related to performance at school and to life satisfaction. On average across OECD countries, students in the top quarter of the index of achievement motivation score 37 points higher in science and reported 0.7 point higher life satisfaction (on a scale from 0 to 10) than students in the bottom quarter of the index. [9] Tamilselvi and Devi [19] found out in a smaller sample of 100 Indian higher secondary school students, that the achievement motivation of higher secondary school students from the selected government schools in Coimbatore District is found to be associated with the type of family they come from, whereas, their gender, subject group, locality of the school they study are not associated with their achievement motivation. Achievement motivation of the students is not associated with the level of education, occupation and annual income of their parents. Also, Pawar [11] ascertained, that male and female secondary school students were found to have same level of academic achievement motivation. The expected result was, that urban secondary school students have high level of academic achievement motivation than rural students. For a similarly large sample of college students Shekhar and Devi [15] detected significant difference between the achievement motivation of male and female college students and between the achievement motivations among science and arts stream students. Whereas females have higher achievement motivation compared to males and science stream students have significantly higher achievement motivation (AM) compared to arts stream students. Sarangi [14] also confirms, that boys have marginally better AM than girls. In terms of the theoretical concept of AM his other results are interesting: „In case of relationship between Achievement Motivation and Academic Achievement, it is observed from the study there is no significant relationship between AM and AA (Acadademic Achievement) in case of tribal students and boys students. On the other hand, there is significant relationship between AM and AA in case of non-tribal students, girl-students, rural and urban students. Hence the study revealed that the achievement motivation enhances the academic achievement of the students (p. 144). Pavlas [10] determined the level of performance motivation in a sample of 116 students from two different secondary schools, one of which was a sports grammar school and the other industrial high school, using a standardized questionnaire LMI (Leistungsmotivationsinventar), which in its current form is suitable only for counseling or research (Sedláková and Knapová [17]). He used the obtained values for a total of 17 dimensions of performance motivation

VOLUME 15,

N° 2

2021

(each of which is saturated with 10 items) to determine possible differences between technically oriented students and between humanities-oriented students. He found that there were statistically significant differences between these schools in the dimension of pride of performance and orientation to status in favor of students of the Sport grammar school. A statistically significant difference was found in the dimension of willingness to learn in favor of students of the Secondary Industrial School. There are no statistically significant differences in any dimension among the women of these schools. There are statistically significant differences between these schools in the willingness to learn dimension in favor of men in industrial school. Men of industrial school have a higher score of performance motivation compared to women of both types of schools in the dimension of “difficulty preference”. Overall, he states that high school students are motivated on average in the overall score of performance motivation. Poledňová, Stránská and Neidobová [13] examined the performance motivation of high school students in relation to their social position in the class. Differences between achievement motivation scores of students with different social positionsin the class (as given by the combination of influence and popularity) proved non-significant. Bakadorova, Hoferichter and Raufelder [1] proved by comparative research, that students from both Montréal and Moscow compare their levels of achievement to the performance of their peers, which motivates them to perform better. In six countries, Pavelková, Hrabal and Hrabal [12] found that the fear of failure to a certain extent acts as a positive motivating factor, while in a strong form it paralyzes individuals and weakens performance. Smith and Karaman [16] notes that many studies examining performance motivation, as a predictor of performance, work with relatively small groups and provide mixed results. Therefore, they develop and validate a broad and unique achievement motivation measure consisting of 36 items assessing concept Contextual Achievement Motivation in multiple settings (School, Work, Family, and Community). Although some studies have dealt with the performance motivation of high school students, no survey of the level of performance motivation was recorded in the population of students of secondary technical schools.

2. Own Research on Achievement Motivation of High School Students

2.1 Aim, Research Questions and Hypotheses of Study The aim of this study is to find out the current level of performance motivation of the population of students of secondary technical schools in the MoravianSilesian Region of Czech Republic, to compare its level with the standard for general secondary school population and to find out whether there is a relationship Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

between its level and choice. The study finds answers to the following research questions: 1. Does the level of performance motivation of the research population in both components differ from the norm for the secondary school population? 2. Does the level of performance motivation of the individual schools constituting the research set differ in their two components? 3. Does the level of performance motivation in both components change with gradual years of study? 4. Is there a relationship between the performance motivation in both their components and the student’s declaration of choice as a primary choice in engineering school? 5. Is there a relationship between the performance motivation in both their components and the student’s declaration of conformity of the current study with the idea of it before entering school? Hypotheses were formulated on the research questions: H1: The level of performance motivation of the research sample in both their components is the same as the norm for the secondary school population. H2: The level of performance motivation in both their components does not differ between the individual schools in the research set. H3: The level of performance motivation in both of its components is higher at the beginning and end of studies compared to the other years. H4: There is no correlation between the performance motivation in both their components and the student’s declaration on the choice of study as a primary choice. H5: The level of performance motivation in both their components and the student’s declaration of conformity of the current study are related to the concept.

2.2. Methods

Two questionnaires were used in the research. The first was the School Performance Motivation Questionnaire for Pupils [4], which builds on Atkinson’s approach. In his model, Atkinson unites the findings of performance motivation research and anxiety research. The theory of performance motivation is then based on the concept of independence of the need for successful performance and the need to avoid failure. The need for successful performance and the need to avoid failure are the basis of performance orientation, further consisting of the degree of attractiveness of the performance activity to individuals and the subjective probability of the expected outcome. The resulting orientation of a person in a performance situation then depends on the predominance of one or the other tendency. The second questionnaire compiled by the authors of the study was focused on finding personal data about students, questions of school choice, satisfaction with school choice and other didactic variables. The research was carried out at the beginning of 2020 and the research sample consisted 46

Articles

VOLUME 15,

N° 2

2021

of over 900 students of secondary technical schools in the Moravian-Silesian Region of the Czech Republic.

2.3. Results

The questionnaire includes 12 questions. Questions 1-6 saturate the NACH indicator (need to achieve, need for successful performance). Questions 7-12 saturate the NAF indicator (the need to avoid failure, or the fear of failure). Both NSP and NAF indicators are determined for each respondent as the sum of the point score for each of these indicators. The point score is called the raw score. The contribution of the questionnaire item of each sub-question is the following rule to the interval scale: answer a - maximum (almost always / always / a lot, etc.) (5 points), b (4 points), c (3 points), d (2 points), e – minimal (not at all / almost never / never, etc.) (1 point). Each respondent therefore has its own raw score value for NACH and NAF. The NACH standard for all secondary school pupils is determined on the basis of the arithmetic average of the graw scores of all partial NACHs calculated for all respondents. Its value is 19.41. The NAF standard for all secondary school pupils is determined on the basis of the arithmetic average of the graw scores of all partial NAFs calculated for all respondents. Its value is 18.03. The average values of the total raw scores of NACH and NAF at individual secondary schools with a mechanical programms. (Tab. 1) are shown in Fig. 1. For reasons of data protection, these schools will be marked with the name of the city in which they are located. The average values of the total raw scores of NACH and NAF for individual year of study (Tab. 2) are shown in Fig. 2. Tab. 1. Total raw scores of NACH and NAF indicators for respondents by individual institutions Mechanical high school students in the city

NACH

NAF

Frýdek–Místek

20,50

18,49

Opava II.

19,21

18,90

Opava I.

19,59

Krnov

17,95

19,00

Ostrava

17,54

18,98

17,56

Tab. 2. Total raw scores of the NACH and NAF indicators for respondents for individual year of study Year

NACH

NAF

19,79

17,59

19,35

18,27

2. 3.

18,97

18,42

Journal of Automation, Mobile Robotics and Intelligent Systems

,21,0

19,36

17,94

VOLUME 15,

,20,0 ,21,0 ,19,0 ,20,0 ,18,0 ,19,0 ,17,0 ,18,0 ,16,0

,17,0 Frýdek-Místek

Opava I. Opava II. NACH standard

Krnov NAF standard

Ostrava

,16,0Fig. 1. Total raw scores of NACH and NAF indicators by individual institutions in Opava I. NACHOpava II. standards Krnov Ostrava comparison with and NAF Fig. 1.Frýdek-Místek Total raw scores ofstandard NACH andNAF NAF indicators by NACH standard

individual inand comparison NACH andin ,20,0Fig. 1. Totalinstitutions raw scores of NACH NAF indicators bywith individual institutions NAF standardscomparison with NACH and NAF standards ,19,0 ,20,0 ,18,0 ,19,0 ,17,0 ,18,0 ,16,0 ,17,0 ,16,0

2. NACH standard

3. NAF standard 3.

4. 4.

Fig. 2. Total raw scores of theNACH NACH and NAF indicators for individual standard NAF standard years of study in comparison with the NACH and NAF standards

Fig. 2. 2. Total raw scores the NACHof andthe NAF NACH indicators and for individual of study in Fig. Total raw ofscores NAFyears indicators comparison with the NACH and NAF standards for of study in comparison with the The individual hypotheses hadyears to be divided into part a and part b to consider the partial indicators NACH or NAF. The hypotheses were tested using IBM Statistics version NACHofand NAF standards

26 software support. The hypotheses hadhypothesis to be divided and part b to consider 1aH0: the partial Hypothesis 1aH. Null aboutinto the part meana value of the parameter: The indicators of NACH or NAF. Theinhad hypotheses were tested using IBM Statistics level of The performance motivation the NACH component is statistically significantly hypotheses to be divided into part aversion and 26 software identical to support. the mean value of 19.41 of the standard for the secondary school part b 1aH. to consider the of NACH or Hypothesis Null hypothesis aboutpartial the mean indicators value of the parameter: 1aH0: The population. level of performance motivation in the NACH component is statistically significantly Alternative about the mean value of the parameter: 1aH1:Statistics The level of NAF. Thehypothesis hypotheses were tested using IBM identical to the mean value 19.41 component of the standard the secondary school performance motivation in theofNACH is notfor statistically significantly version software population. identical with 26 the mean value of support. 19.41 of the standard for the secondary school Alternative hypothesis about the mean value of the parameter: 1aH1: The level of population. performance motivation in the by NACH component is pnot statistically**significantly From the p-value, determined the mean value test, = 0,0010303 (p <0,05 *, Null hypothesis the with the mean of 19.41 of the thesignificance standard forlevel the αsecondary school pidentical <0,01Hypothesis **, p <0,001 ***),value it 1aH. follows that at =about 0, 01, the null population. mean value of the parameter: 1aH0: The level of From the p-value, determined by the mean value test, p = 0,0010303 ** (p <0,05 *, performance motivation inthethe NACHlevel component is p <0,01 **, p <0,001 ***), it follows that at significance α = 0, 01, the null

statistically significantly identical to the mean value of 19.41 of the standard for the secondary school population. Alternative hypothesis about the mean value of the parameter: 1aH1: The level of performance motivation in the NACH component is not statistically significantly identical with the mean value of 19.41 of the standard for the secondary school population. From the p-value, determined by the mean value test, p = 0,0010303 ** (p <0,05 *, p <0,01 **, p <0,001 ***), it follows that at the significance level α = 0, 01, the null hypothesis is rejected in favor of the alternative hypothesis. Therefore, it is true that the level of performance motivation in the NACH component is not statistically significantly identical with the mean value of 19.41 of the standard for the secondary school population.

Hypothesis 1bH. Null hypothesis about the mean value of the parameter: 1bH0: The level of performance motivation in the NAF component is statistically significantly identical to the mean value of 18.03 of the standard for the secondary school population. Alternative hypothesis about the mean value of the parameter: 1bH1: The level of performance motivation in the NAF component is not statistically significantly identical to the mean value of 18.03 of the standard for the secondary school population.

N° 2

2021

From the p-value, determined by the mean value test, p = 0.54736 (p> 0.05), it follows that at the level of significance α = 0.05, the null hypothesis is not rejected. Therefore, the level of performance motivation in the NAF component is statistically significantly identical to the mean value of 18.03 of the standard for the secondary school population.

Hypothesis 2aH. Null hypothesis: 2aH0: The level of performance motivation in the NACH component is statistically significantly the same within schools. Alternative hypothesis: 2aH1: The level of performance motivation in the NACH component is statistically significantly different within schools. From the p-value, determined by the non-parametric Kruskal-Wallis test, p = 0.003488 ** (p <0.05 *, p <0.01 **, p <0.001 ***), it follows that at the significance level α = 0,01 the null hypothesis is rejected in favor of the alternative hypothesis. Therefore, the level of performance motivation in the NACH component is statistically significantly different within schools.

Hypothesis 2bH. Null hypothesis: 2bH0: The level of performance motivation in the NAF component is statistically significantly the same within schools. Alternative hypothesis: 2bH1: The level of performance motivation in the NAF component is statistically significantly different within schools. From the p-value, determined by the non-parametric Kruskal-Wallis test, p = 0.08841 (p> 0.05), it follows that at the significance level α = 0.05, the null hypothesis is not rejected. Therefore, it is true that the level of performance motivation in the NAF component is statistically significantly the same within schools.

Hypothesis 3aH. Null hypothesis: 3aH0: The level of performance motivation in the component of NACH is statistically significantly identical at the beginning and at the end of the study compared to other years of study. Alternative hypothesis: 3aH1: The level of performance motivation in the NACH component is statistically significantly higher at the beginning and at the end of the study compared to other years of study. From the p-value, determined by the one-sided non-parametric Mann-Whitney test, p = 0.04264 * (p <0.05 *, p <0.01 **, p <0.001 ***), it follows that at the significance level α = 0,05 the null hypothesis is rejected in favor of the alternative hypothesis. Therefore, it is true that the level of performance motivation in the NACH component is statistically significantly higher at the beginning and at the end of the study compared to other years of study. Hypothesis 3bH. Null hypothesis: 3bH0: The level of performance motivation in the NAF component is statistically significantly identical at the beginning and at the end of the study compared to other years of study. Alternative hypothesis: 3bH1: The level of performance motivation in the NAF component is statistiArticles

Journal of Automation, Mobile Robotics and Intelligent Systems

cally significantly higher at the beginning and at the end of the study compared to other years of study. The p-value, determined by the one-sided non-parametric Mann-Whitney test, p = 6.72 × 10-4 *** (p <0.05 *, p <0.01 **, p <0.001 ***), shows that at the significance level α = 0.001, the null hypothesis is rejected in favor of the alternative hypothesis. Therefore, it is true that the level of performance motivation in the NAF component is statistically significantly higher at the beginning and at the end of the study compared to other years of study.

Hypothesis 4aH. Null hypothesis: 4aH0: There is no statistically significant relationship between performance motivation in the NACH component and the student’s statement on the choice of study at an engineering school as a primary choice. Alternative hypothesis: 4aH1: There is a statistically significant relationship between performance motivation in the NACH component and the student’s statement on the choice of study at an engineering school as a primary choice. From the p-value, determined by the non-parametric Mann-Whitney test, p = 0.91791 (p> 0.05), it follows that at the significance level α = 0.05 the null hypothesis is not rejected. Therefore, it is true that there is no statistically significant relationship between performance motivation in the NACH component and the student’s statement on the choice of study at an engineering school as a primary choice.

Hypothesis 4bH. Null hypothesis: 4bH0: There is no statistically significant relationship between performance motivation in the NAF component and the student’s statement on the choice of study at an engineering as a primary choice. Alternative hypothesis: 4bH1: There is a statistically significant relationship between performance motivation in the NAF component and the student’s statement on the choice of study at an engineering school as a primary choice. From the p-value, determined by the non-parametric Mann-Whitney test, p = 0.03917 * (p <0.05 *, p <0.01 **, p <0.001 ***), it follows that at the significance level α = 0, 05 the null hypothesis is rejected in favor of the alternative hypothesis. Therefore, there is a statistically significant relationship between performance motivation in the NAF component and the student’s statement on the choice of study at an engineering school as a primary choice.

Hypothesis 5aH. Null hypothesis: 5aH0: There is no statistically significant relationship between performance motivation in the NACH component and the student’s statement on the coincidence of current study and the idea. Alternative hypothesis: 5aH1: There is a statistically significant relationship between performance motivation in the component of NACH and the student’s statement on the coincidence of the current study and the idea. Articles

VOLUME 15,

N° 2

2021

From the p-value, determined by the non-parametric Mann-Whitney test, p = 2,157 × 10-15 *** (p <0,05 *, p <0,01 **, p <0,001 ***), it follows that at the level of significance α = 0.001, the null hypothesis is rejected in favor of the alternative hypothesis. Therefore, it is true that there is a statistically significant relationship between performance motivation in the NACH component and the student’s statement on the coincidence of the current study and the idea. Hypothesis 5bH. Null hypothesis: 5bH0: There is no statistically significant relationship between performance motivation in the NAF component and the student’s statement on the coincidence of current study and the idea. Alternative hypothesis: 5bH1: There is a statistically significant relationship between performance motivation in the NAF component and the student’s statement on the coincidence of current study and the idea. From the p-value, determined by the non-parametric Mann-Whitney test, p = 5.846 × 10-7 *** (p <0.05 *, p <0.01 **, p <0.001 ***), it follows that at the level of significance α = 0.001, the null hypothesis is rejected in favor of the alternative hypothesis. Therefore, it is true that there is a statistically significant relationship between performance motivation in the NAF component and the student’s statement on the coincidence of current study and the idea.

3. Discussion

The results of our research cannot be compared with the results of other researchers, as the individual studies worked with other evaluation tools. However, they can be compared with a standardized standard for the relevant population. In this view, it is a fundamental finding that in the component of performance motivation – NACH (need to achieve, need for successful performance) the performance motivation of the sample is statistically different from the group norm. In the NAF indicator component (the need to avoid failure, or the fear of failure) the results are statistically identical. In two out of five examined schools, the school values in both components are always higher than the group standard and in three schools lower. Schools do not differ in the results for the NACH component, but differ in the NAF component. In both components, the value of performance motivation for beginners and for final years is significantly different from other years. The effect of “novelty” and desired study is probably manifested in the first year in the highest value of performance motivation in the NACH component and the lowest value of NAF among all four years. In the second year of study, on the other hand, the lowest value of NACH and the highest of NAF turns out. With the following years, NACH rises slightly and NAF decreases. This finding probably explains the students’ interest in successfully completing their studies and entering practice with the acquired qualification or continuing their university studies. It has not been confirmed that the student’s statement on the choice of study at an engineering school as a primary choice is related to the values of NACH and NAF. This could be explained by the lack of prefer-

Journal of Automation, Mobile Robotics and Intelligent Systems

ence for secondary technical studies or by the standard quality of basic education, which allows a wider choice of fields of secondary education for graduates of basic education. On the other hand, if this link were to be confirmed, it would mean that teachers would pay increased attention to pupils for whom secondary mechanical school was another professional choice and are not intrinsically motivated to adequate school performance. It was confirmed that there is a statistically significant relationship between performance motivation in the NACH performance motivation component and also in the NAF component with the student’s declaration of the coinciding of the current study with the idea of it. This would testify in favor of the system of professional orientation of youth, resp. also in favor of the care of parents for the professional choice of children, which provide those interested in studying at secondary schools with information about the field and its study, which are then confirmed by the pupils’ educational practice.

4. Conclusion

Engineering production in developed countries has a great perspective in the era of Industry 4.0 and faces a shortage of skilled labor. The work of secondary school teachers with the results of their students’ performance motivation measurement could positively influence not only the updating and modification of the curriculum, but also the methods used in the theoretical and practical component of their vocational training. It is very important to lead pupils and students to find a positive form of adaptation to the social environment, to strengthen motivation that is related to the diverse focus of human activity, such as technical study. It is possible to develop pupils’ motivation for technical studies by updating (awakening) their needs (cognitive, performance, social). Pupils’ cognitive needs can be developed by problem tasks in the field of technical everyday reality. It is appropriate to support project teaching, in which there is enough space for problem-based teaching, manual competition, programmed learning, creative tasks, research activities (technical experiments), brainstorming discussions and sharing examples of good technically oriented everyday practice.

Acknowledgement

This study was written within the project of the Technology Agency of the Czech Republic called “Education in engineering and its optimisation for the needs of the labour market”, registration number TJ 02000083, carried out at the Faculty of Education at the University of Ostrava between 2019 and 2021.

AUTHORS

Josef Malach* – Faculty of Education, Department of Pedagogy and Andragogy, University of Ostrava, Ostrava, Czech Republic, Email: josef.malach@osu.cz.

VOLUME 15,

N° 2

2021

Dana Vicherková – Faculty of Education, Department of Pedagogy and Andragogy, University of Ostrava, Ostrava, Czech Republic, Email: dana.vicherkova@ osu.cz.

Milan Chmura – Faculty of Education, Department of Pedagogy and Andragogy, University of Ostrava, Ostrava, Czech Republic, Email: milan.chmura@osu.cz. Veronika Švrčinová – Faculty of Education, Department of Technical and Vocational Education University of Ostrava, Ostrava, Czech Republic, Email: veronika.svrcinova@osu.cz. *Corresponding authors

REFERENCES [1]

[2] [3] [4] [5]

[6] [7] [8] [9] [10]

O. Bakadorova, F. Hoferichter and D. Raufelder, “Similar but different: social relations and achievement motivation in adolescent students from Montréal and Moscow”, Compare: A Journal of Comparative and International Education, vol. 50, no. 6, 2020, 904–921, 10.1080/03057925.2019.1576122. M. V. Covington, “Goal Theory, Motivation, and School Achievement: An Integrative Review”, Annu. Rev. Psychol., vol. 51, no. 1, 2000, 171–200, 10.1146/annurev.psych.51.1.171. H. Heckhausen, “Motivationsanalyse der Anspruchsniveau-Setzung”, Psychol. Forsch., vol. 25, no. 2, 1955, 118–154, 10.1007/ BF00422332, (in German). V. Hrabal and I. Pavelková, Školní výkonová motivace žáků. Dotazník pro žáky., Národní ústav odborného vzdělávání, 2011, (in Czech). P. W. J. Hustinx, H. Kuyper, M. P. C. van der Werf and P. Dijkstra, “Achievement motivation revisited: new longitudinal data to demonstrate its predictive power”, Educ. Psychol., vol. 29, no. 5, 2009, 561–582, 10.1080/01443410903132128. S. Kołodziej, “The role of achievement motivation in educational aspirations and performance”, General and Professional Education, vol. 1, 2010, 42–48. D. C. McClelland, J. W. Atkinson, R. A. Clark and E. L. Lowell, The achievement motive, AppletonCentury-Crofts, 1953, 10.1037/11144-000. J. Meece and Ch. Agger, “Achievement Motivation in Education”. In: Oxford Research Encyclopedia of Education, 2018, 10.1093/acrefore /9780190264093.013.7. OECD, PISA 2015 Results (Volume III): Students’ Well-Being, OECD Publishing, 2017, DOI: 10.1787/9789264273856-en. I. Pavlas, “Achievement motivation of High School Students”, Paidagogos – Journal of Education in Contexts, vol. 2015, no. 1, 2015, (in Czech). Articles

Journal of Automation, Mobile Robotics and Intelligent Systems

[11]

[12] [13]

[14] [15]

[16]

[17] [18]

[19]

S. Pawar, “A Study of Academic Achievement Motivation among Secondary School Students”, Scholarly Research Journal for Interdisciplinary Studies, vol. 4, no. 36, 2017, 6646–6651, 10.21922/srjis.v4i36.10009. I. Pavelková, K. Hrabal and V. Hrabal, “Comparison of the Sources of Motivation of the Pupil´s Learning Activity”, Pedagogika, vol. 60, no. 3-4, 2010, 97–107, (in Czech). I. Poledňová, Z. Stránská and H. Niedobová, “Achievement Motivation of Secondary School Students in Relation to Their Social Position in the Class”, Problems of Psychology in the 21st Century, vol. 8, no. 1, 2014, 61–70, 10.33225/ ppc/14.08.61. C. Sarangi, “Achievement Motivation of the High School Students: A Case Study among Different Communities of Goalpara District of Assam”, J. Educ. Pract., vol. 6, no. 19, 2015, 140–144. C. Shekhar and R. Devi, “Achievement Motivation across Gender and Different Academic Majors”, Journal of Educational and Developmental Psychology, vol. 2, no. 2, 2012, 105–109, 10.5539/jedp.v2n2p105. R. Smith and M. A. Karaman, "Development and Validation of the Contextual Achievement Motivation Measure", International Journal of Psychology and Educational Studies, vol. 6, no. 3, 2019, 16-26. J. Sedláková and L. Knapová, “Dotazník motivace k výkonu: Recenze metody”, TESTFÓRUM, vol. 5, no. 8, 2017, 19–24, 10.5817/TF2017-8132, (in Czech). R. Steinmayr and B. Spinath, “The importance of motivation as a predictor of school achievement”, Learning and Individual Differences, vol. 19, no. 1, 2009, 80–90, 10.1016/ j.lindif.2008.05.004. B. Tamilselvi and S. Devi, “A study on achievement motivation of higher secondary students in Coimbatore district”, International Journal of Advanced Education and Research, vol. 2, no. 3, 2017, 81–84.

Articles

VOLUME 15,

N° 2

2021

wowe skrót

Publisher: Łukasiewicz – Industrial Research Institute for Automation and Measurements PIAP