Smart Distribution Network
InHand Wireless Overhead Line System
InHand Networks Inc.
Pioneering Industrial IoT for 17 Years
Global & Public company Beijing, China Virginia, US
Listed on NEEQ 430642 Backed by Top Silicon Valley VC
2014 2012
2007 2001
2018
Delivered industrial cellular modem for China State Grid
Founded as M2M startup, Pioneer of IoT
2005 Developed industrial router and remote monitoring solution for GE Healthcare; Now a global supplier of GEHC
Berlin, Germany
Became Schneider CAPP Technology Partner and Rockwell Encompass Product Partner
Complete IoT products and solutions proven by world class customers
Total Solutions for Industrial IoT IoT Products
Industrial Cellular Router
Cellular Data Terminal
Industrial Switch
Industrial Computer
Device Network Cloud Platform
Cloud Platform
Remote Maintenance and Service Platform
Solutions (include some products lines)
Distribution Overhead Line Analytic System
Smart Vending & Operation Platform
Machine Remote Monitoring System
France | Energy Energy Management Service
UK | Transportation Traffic Signal Control
Germany | Power Grid Distributed Energy Monitoring
China | Power Grid Distribution Automation Overhead Line Monitoring
Schneider Electric
Siemens
E.ON Energy, RWE
China State Grid
Used worldwide. Proven worldwide.
4
US | Healthcare Italy | Energy Generating Medical Equipment Monitoring Steam Boilers Preventive Maintenance Monitoring & Control
South Africa Industrial Automation
India | Oil & Gas Natural Gas Pipeline System Monitoring
GE Healthcare
Schneider Electric
Gail
Viessmann
01 / How IWOS works
Contents
02 / Solution Highlights 03 / Summary
60,000+ Line Sensors Deployed
6,000+ Power Lines Covered
15 Million+ Power Line Events & Waveforms
20,000+ Line Faults Detected 6 All figures are by the end of 2017.
InHand Wireless Overhead-Line System (IWOS) Fault Detection & Location • Grid Analytics • Preventive Maintenance
IWOS Analytics Platform • • •
How it works
Line Status Analysis Fault Detection Interface to DMS
Wireless VPN / Fiber Communication
Line Status Monitoring
Big-data Analysis
Panorama Monitoring
DMS Center Utility Pole # 25
Pole # 30
LinePoleSensors #
Pole # 177 Pole # 5
Pole # 94
Feeder Lines 110KV Substation
Pole #
Pole # 174
Pole # 2
Concentrator
Auto-triggered Waveform-recording Accurate Fault Locating
Pole # 28
InHand Networks
Notification to Line Patrol Team
Preventive Grid Analytics Powered by AI
• Integration Model based on Neural Network • Oscillography waveform classification based on AI algorithms • Accurately identify line faults and events including ground faults, short circuit faults, inrush current, power outage and etc. • Preventive analysis and pre-fault alerts
8
Basic Setup Site Configuration
• 3 Line Sensors form one group • Line Sensors communicate with concentrator via short range RF • Concentrator utilizes LTE/SSN to communicate with headend system
A
Design principle & Benefits
Line Sensors
B
RF
Concentrator
C
• Simplify line sensors, leave complexity to concentrator • Flexibility • Optimization • Time synchronization • Low cost • Easy to install & maintain • Future proof
Key Features Line Sensor
Line Sensor
• High precision current measurement: 0-800A ±1% • Electric-field strength measurement: ±1% • High sampling rate of oscillography waveforms: 12.8KHz (256 pts per cycle) • High precision time synchronization: 30us, enabling zerosequence current synthesis • Power optimization operation - Full-function mode: all features including waveform recording enabled; powered by line and supercapacitors - Low-power mode: waveform recording disabled; powered by line and battery • Power harvesting: 1A to support full-function mode; supercapacitors back for 12h • Water proof: IP67 • Light weight: < 1.25kg • Wide operation temperature: -40 ~ +70 °C
Concentrator Key Features • • • • • • • • Model A: Pole-mounted
WAN: LTE Field Area Network: RF, BLE Location and time synchronization: GPS Fault detection based on zero-sequence current synthesis Main power: 20W/30W solar panel Backup power: Lead–acid/Lithium battery IP55 protection Remote upgrade & maintenance
Concentrator Key Features
Model B: Line-mounted
• • • • • • • •
WAN: LTE Field Area Network: RF, BLE Location and time synchronization: GPS Fault detection based on zero-sequence current synthesis Main power: energy harvesting from the line, 6A Backup power: 24h Water proof: IP67 Light weight: <3kg
Field Installation
Headend System Key Features • Sensor/concentrator management • GIS-based real-time monitoring of line status and events • Deep Learning analytic system • Oscillography waveform classification • Fault alarms and preventive analysis • Reports • Load trend & statistics • power quality analysis • Fault statistics • System management
Artificial Intelligence is the Core Technology behind IWOS
Work condition classification flow
Power restore
Some features used by AI
Time domain
Frequency domain
Wavelet domain
Waveform signatures
Power Restore
Ground Fault Restore
Power Outage
Lightning Strike
Waveform signatures
Short Circuit Fault
Inrush
Other Condition
Other Condition
Accuracy of classification[1]
Total # of verified conditions: 2,924 sets Ground Faults: 522 sets Short Circuit Faults: 236 sets Inrush Current: 560 sets Lightning: 601 sets Power Restore: 524 sets Power Outage: 293 sets Other Conditions: 188 sets
Power restore
[1] Smart Distribution Network Operating Condition Recognition Based on Big Data Analysis. Fan Min , Zhang Bo , Yao Qiang , Zhang Jianliang , Huang darong , Han Qi. IEEE Conference on Data Driven Control and Learning Systems Conference(DDCLS), 2017.
Ground fault location process ① Test check if the Ground-fault event can be located
② Selection of fault line Select the fault line by comparison between lines
③ Fault location Locate the fault segment by comparison between nodes
故障定位模型结构
Integration Model based on DNN
Waveforms Preprocess
Output
故障定位模型结构
Training process
110kV Substation
Ground fault location example
Accuracy of ground fault location Result by ML
Not located
Full data set
3.56%
3.41%
Missing:
3.56%
False:
4.68%
Success:
92.76%
Locatable data set (94.92% of full data set): Success 89.35%
1.67%
located
Failed 3.01%
Missing:
3.71%
False:
3.14%
Success:
93.15%
Fault Can be located
Cannot be located
Our Achievements
Proved by installations in 18 China provinces and cities
20,000+ sets 60,000+ line sensors
China State Grid China Southern Grid
Achievements by Q2 2017 Statistic of IWOS systems under InHand Networks Support & Service (part of total IWOS installations) By Q2 2017 No. Power Utilities (Regional Level) No, of Substations
65 462
No. of Feeder Lines
2,159
No. of Installations
6,038
No. of Sensors No. of Power Line Events, with waveform data No. of Detection of Short Circuit Faults No. of Detection of Ground Faults
18,114 15,099,165 2,817 10,115
IWOS is: • an advanced and field-proven solution for fault location • optimized for cost and flexibility • enabled and highlighted by Machine Learning tech • keep evolving towards predictive maintenance
Wangjing Science Park
3900 Jermantown Road, Suite 150
Beijing, China
Fairfax, VA 22030
+86 (10) 64391099
+1 (703) 348 2988
Our Customers Smart Grid
Industrial Automation
Transportation & Security
29
Healthcare
Smart Vending
Finance
Distributor for Italy
MARCOM SRL Via Mezzacampagna 52 (int 29) Verona 37135 Italia info@marcomweb.it www.marcomweb.it