Solving Intuitionistic Fuzzy Multiobjective Linear Programming Problem

Page 1

Solving intuitionistic fuzzy multiobjective linear programming problem under neutrosophic environment

Article in AIMS Mathematics · February 2021 DOI: 10.3934/math.2021269

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/349538468
CITATIONS 0 READS 14 2 authors: Some of the authors of this publication are also working on these related projects: Modeling and Optimization Techniques under Neutrosophic Environment View project
17
Abdullah Ali H. Ahmadini Jazan University
PUBLICATIONS 5 CITATIONS SEE PROFILE
All content following this page was uploaded by Abdullah Ali H.
on 28 February 2021. The user has requested
the downloaded file.
Firoz Ahmad Indian Statistical Institute Kolkata
18 PUBLICATIONS 74 CITATIONS SEE PROFILE
Ahmadini
enhancement of

http://www.aimspress.com/journal/Math

Researcharticle

AIMSMathematics,6(5):4556–4580. DOI:10.3934/math.2021269 Received:24September2020 Accepted:03February2021 Published:22February2021

Solvingintuitionisticfuzzymultiobjectivelinearprogrammingproblem underneutrosophicenvironment

AbdullahAliH.Ahmadini1 andFirozAhmad2,3,*

1 DepartmentofMathematics,FacultyofSciences,JazanUniversity,Jazan,SAUDIARABIA

2 DepartmentofStatisticsandOperationsResearch,AligarhMuslimUniversity,Aligarh,INDIA

3 SQC&ORUnit,IndianStatisticalInstitute,203B.T.Road,Kolkata,700108,INDIA

* Correspondence: Email:firoz.ahmad02@gmail.com.

Abstract: Theexistenceofneutral /indeterminacydegreesreflectsthemorepracticalaspects ofdecision-makingscenarios.Thus,thispaperhasstudiedtheintuitionisticfuzzymultiobjective linearprogrammingproblems(IFMOLPPs)underneutrosophicuncertainty.Tohighlightthedegrees ofneutralityinIFMOLPPs,wehaveinvestigatedtheneutrosophicoptimizationtechniqueswith intuitionisticfuzzyparameters.Themarginalevaluationofeachobjectiveisdeterminedbythree differentmembershipfunctions,suchastruth,indeterminacy,andfalsitymembershipdegreesunder theneutrosophicenvironment.Themarginalevaluationofeachobjectivefunctioniselicitedby varioussortsofmembershipfunctionssuchaslinear,exponential,andhyperbolictypesofmembership functions,whichsignifiesanopportunityfordecision-makerstoselectthedesiredmembership functions.Thedevelopedneutrosophicoptimizationtechniqueisimplementedonexistingnumerical problemsthatrevealthevalidityandapplicabilityoftheproposedmethods.Acomparativestudyis alsopresentedwithotherapproaches.Atlast,conclusionsandfutureresearchdirectionsareaddressed basedontheproposedwork.

Keywords: intuitionisticfuzzyparameters;indeterminacymembershipfunction;neutrosophic optimizationtechnique;multiobjectivelinearprogrammingproblem

MathematicsSubjectClassification: 03B52,03F55,62J05,62J99

1.Introduction

Optimizationtechniqueshavemuchimportanceandpopularityinreallifewhilesolving optimizationproblems.Avarietyofmathematicalmodelsexistswhiledealingwithreal-lifescenarios ofdifferentproblems.Hence,themathematicalmodelscontainsomespecificobjectivefunction(s) alongwithasetofwell-definedconstraint(s)subjecttowhichtheproblemsaretobeoptimized(either

maximizedorminimized).Thus,amathematicalprogrammingproblemhavingmultipleconflicting objectivesunderasetofmixedconstraintsistermedasamultiobjectiveprogrammingproblem (MOPP).Inreallife,thedifferentfieldsofoptimizationsuchastransportationproblem,inventory control,assignmentproblems,portfoliooptimization,supplierselection,manufacturingsystem, supplychainmanagement,engineeringproblems,etc.containmorethanoneconflictingobjectives (eitherminimizationormaximization)andtaketheformofMOPP.Whiledealingwithmultiple objectives,itisnotalwayspossibletoobtainasinglesolutionthatoptimizeseachobjectiveefficiently. However,acompromisesolutionispossiblethatsatisfieseachobjectivesimultaneously.Therefore theconceptofcompromisesolutionisanimportantaspectandleadsinsearchoftheglobaloptimality criterion.Inthepastfewdecades,atremendousamountofresearchhasbeenpresentedinthecontext ofmultiobjectiveoptimizationtechniques.

Initially,thefuzzyset(FS)wasinvestigatedby[35],whichcontainsthedegreeofbelongingness (membershipfunctions)fortheelementintothefeasiblesolutionset.BasedonFS,[39]presentedthe fuzzyprogrammingapproach(FPA)forMOPPinwhichmembershipfunctionsrepresentthe marginalevaluationofeachobjective.InFPA,thedecision-makerssatisfactionlevelisachievedby maximizingthemembershipfunctionofeachobjective.Lateron,theextensionsofFPAsuchfuzzy intervalprogramming,fuzzystochasticprogramming,fuzzygoalprogramming,etc.havebeen presentedandsuccessfullyappliedtosolvetheMOPPaccordingtothenatureoftheproblemand appropriaterequirementofthetechniques.

Insomecases,onlythemembershipfunctionmaynotbethebestrepresentativeofthedegreeof belongingnessofanelementintothefeasiblesolutionset.However,thedegreeofnon-belongingness wouldbebettertorepresentthehesitationaspectsofanelementintothefeasiblesolutionset.Asa result,intuitionisticfuzzyset(IFS)waspresentedby[11],whichisthegeneralizationandextensionof theFS.TheIFSdealswithboththedegreeofbelongingnessandnon-belongingnessoftheelementinto thefeasiblesolutionsimultaneously.BasedontheIFS,P.P.Angelov[10]suggestedtheintuitionistic fuzzyprogrammingapproach(IFPA)tosolvetheMOPPinwhichthemarginalevaluationofeach objectivefunctionisdeterminedbythemembershipaswellasnon-membershipfunctions.InIFPA, thesatisfactionlevelofthedecision-makerisachievedbymaximizingthemembershipandminimizing thenon-membershipfunctionsofeachobjectivesimultaneously.Afterward,theextendedversionof IFPA,suchasintuitionisticfuzzystochasticprogramming,intuitionisticfuzzygoalprogramming,etc. havebeendevelopedandappliedtovariousdecision-makingproblems.

FurtherextensionofFSandIFShasbeenpresentedbecauseindeterminacydegreealsoexistin decision-makingprocesses.Tocopewiththedegreeofindeterminacy/neutrality,F.Smarandache[32] proposedtheneutrosophicset(NS).TheNSconsidersthreedifferentmembershipfunctionsnamely; thetruth(degreeofbelongingness),indeterminacy(degreeofbelongingnessuptosomeextent)and afalsity(degreeofnon-belongingness)membershipfunctionsfortheelementintoafeasiblesolution set.Recently,manyauthorshaveusedtheconceptofNSsuchas[2, 4, 5, 26]whilemakingtheoptimal decisions.Anindeterminacydegreearetheareasofignoranceofpropositions’pointsbetweenthetruth andafalsitydegree.Itmeansthattheindeterminacydegreesareindependentandcanbeoptimized simultaneously.Forexample,ifwecollectsomeinformationfromscholarsregardingajournaland thepossibilitythatthejournalisgoodis0.7,thejournalisnotgoodis0.6,anddonotknowaboutthe journalis0.3.Then,thiskindoflinguisticvaguenessorimprecisenessarebeyondthescopeofFSand IFSandconsequentlybeyondFPAandIFPAtodecisionmakingprocessrespectively.Therefore,itis worthtoincorporatetheindeterminacydegreewhiledealingwithMOPP.

AIMSMathematics

Volume6,Issue5,4556–4580.

4557

Acontinuouseffortisbeingmadebymanyresearchersorpractitionerstoobtainthebestpossible solutionofmultiobjectivelinearprogrammingproblems(MOLPPs).Alargepartoftheliteratureisfull offuzzy-basedoptimizationtechniquesforMOLPPs.Inthepastfewyears,thegeneralizedconcept ofafuzzysetisutilizedtosolvetheMOLPPs.Anintuitionisticfuzzy-basedoptimizationmethod isalsoimplementedbymanyresearchersandgainedawiderangeofapplicabilityandacceptability whileoptimizingtheMOLPPs.However,theexistingapproacheshavesomelimitationsandcanbe overcomebyapplyingtheproposedneutrosophicoptimizationtechnique.Thefollowingaresome essentialpointsthatensurearemarkablecontributiontothedomainofmultiobjectiveoptimization techniques.

• Anindeterminacydegreeistheregionofignoranceofpropositions’valuesbetweenthetruthand falsitydegrees.Thisaspectcanonlybemanagedwiththeneutrosophicoptimizationmethod.

• TheexistingmethodsofsolvingMOLPPs[19, 28, 36]consideredonlythemembershipfunction whereas[22, 30, 31]includedthemembershipaswellasnon-membershipdegreesofeach objectivefunction.Theydonotcovertheindeterminacydegreewhilemakingdecisions.We havesuccessfullycopewiththeconceptofneutralityandhencesuggestedindeterminacydegree alongwithmembershipandnon-membershipdegreessimultaneously.

• ThemethodsofhandlingMOLPPs[13, 17]onlytakecareofthedegreeofbelongingnessamong theparametersanddonotconsiderthedegreeofnon-belongingness.Therefore,thisstudy capturedtheconceptsofbelongingnessandnon-belongingnessdegreesamongtheparameters anddealtwiththehesitationaspects.

• Thepresentedapproachby[18]isapplicableonlyforsingle-objectiveproblems.Hencethe addressedmethodcanbeeasilyappliedtoMOLPPswithdifferenttypesofmembershipfunctions.

• Thestudypresentedby[30, 36]doesnotpermittheversatilityofvaguenessdegree(shape parameters)inindeterminacydegrees,butwhileapplyingexponential-typemembershipfunction underneutrosophicenvironment,itcanbeavailed.

Therestofthepaperhasbeensummarizedasfollows:InSection 2,relatedresearchworkhave beendiscussedwhereasSection 3 representsthepreliminaryconceptsregardingintuitionisticfuzzy andneutrosophicsets.TheSection 4 representsmodelingofIFMOLPPs.Theproposedneutrosophic optimizationtechniqueispresentedinSection 5,whereasinSection 6,numericalillustration,and comparativestudywithotherexistingapproachesisdepictedtoverifytheperformanceofthe propoundedsolutionmethod.Finally,concludingremarksandfutureresearchdirectionispresented inSection 7

2.Literaturereview

Literaturesuggestslotsofresearchhasbeencarriedouttosolvethemultiobjectiveprogramming problemsusingfuzzy,intuitionisticfuzzy,andneutrosophicprogrammingapproaches.[20]presented asatisfyingsolutionapproachformultipleobjectives,whichcomprisethemorecriticalobjective attaintingthehighersatisfyingdegree.Amodifiedmethodbasedonthefuzzyrankingfunctionis suggestedby[33]andthecrispmodelsoobtained.Anuncertaingoalprogrammingapproachwas proposedby[21]forthesolutionofuncertainMOPP.P.Singh,etal.[28]discussedanefficient methodthatcarriesthequalityofthreedifferentapproaches,namely,fuzzyprogramming,goal programming,andinteractiveprogramming.Furthermore,thedevelopedalgorithmisappliedtothe

AIMSMathematics

Volume6,Issue5,4556–4580.

4558

4559 multiobjectivetransportationproblem.TheMOPPunderuncertaintywasaddressedby[38]and namedasanuncertainmultiobjectiveprogrammingproblem.Theknowledgeofefficiencyconcepts withdifferentforms,suchasexpected-value,properexpectedvalue,andtheirrelationshipswitheach other,havebeenestablishedundertheuncertainenvironment.Amultiobjectivefractional programmingproblemwaspresentedby[14],andanovellinearizationstrategictechniquehasbeen developedtosolvetheproblem.J.Dong,etal.[16]alsosuggestedanewapproachbasedonranking functionandTOPSIStosolvetheMOPPunderuncertainty.Anewapproachisproposedby[17]to solvethefullyfuzzyMOPPunderfuzzyparametersthatreducesthecomputationalcomplexityofthe problem.Asolutionapproachisalsodiscussedby[13]tosolvethefullyfuzzymultiobjectivelinear programmingproblemunderfuzzytrapezoidalparameters.S.L.Tilahu[34]developedafeasibility reductionapproachforhierarchicaldecision-makinginMOPP,andthepreferencefreemethod(ideal pointmethod)isappliedtosolveit.

Theintuitionisticfuzzygoalprogrammingapproachisdiscussedby[25]thathastakenthe advantagesofintuitionisticfuzzyset,goalprogramming,andinteractiveprocedures,tosolvethe MOPP.A.K.Nishad,etal.[23]addressedanintuitionisticfuzzygoalprogrammingtosolvethe MOPPunderimpreciseobjectivesandconstraintsand;implementedtosolvetheagricultural productionmanagementsystem.D.Rani,etal.[24]alsosuggestedanalgorithmtosolveMOPP underanintuitionisticfuzzyenvironment.Theproposedapproachhasbeencomparedwithother existingapproaches.S.K.Singh,etal.[30]alsoaddressedintuitionisticfuzzyMOPPundertriangular intuitionisticfuzzyparametersandmixedconstraintsand;solvedtheMOPPusingdifferent membershipfunctions.V.Singh,etal.[31]presentedthemodelingandoptimizationframeworkfor MOPPunderanintuitionisticfuzzyenvironmentwithanoptimisticandpessimisticpointofview. S.K.Singh,etal.[29]solvedthemulti-objectivemixedintegerprogrammingproblemunder intuitionisticfuzzyenvironmentandimplementedonthesupplychainplanningproblem.Recently,A. A.H.Ahmadini,etal.[9]investigateedanovelpreferenceschemeformultiobjectivegoal programmingproblemsunderdifferentsortsofmembershipfunctions.

Basedonneutrosophicsettheory,F.Ahmad,etal.[6]addressedamodifiedneutrosophic optimizationalgorithmforsupplychainmanagementinanuncertainsituation.S.Zeng,etal.[37] presentedmulti-criteriasupplierselectionframeworkbassedontheneutrosophicfuzzydataset. Recently,manyauthorshavecontributedtotheapplicationofneutrosophicoptimizaationtheoryin variousreal-lifeproblems.F.Ahmad[8]discussedthesupplierselectionproblemunderType-2fuzzy parametersandsolvedusingtheinteractiveneutrosophicprogrammingapproach.A.Y.Adhami,et al.[1]alsoproposedanovelPythagorean-hesitantfuzzyoptimizationtechniqueandapplieditto solvethetransportationproblem.F.Ahmad,etal.[7]performedastudyonEnergy-Food-Water securitynexusmanagementandappliedtheneutrosophicprogrammingapproachtosolvethe propoundedmodel.Hence,theproposedneutrosophicapproachemphasizesindeterminacyandtruth andfalsitydegreeswhilemakingthedecisions.Italsogeneralizesthefuzzyandintuitionistic fuzzy-basedapproachesunderdifferenttypesofmembershipfunctionssuchaslinear,exponential, andhyperbolicmembershipfunctions.

3.Preliminaryconcepts

Inthissection,wehavediscussedsomebasicconceptsrelatedtointuitionisticfuzzyand neutrosophicsets.

AIMSMathematics

Volume6,Issue5,4556–4580.

Definition1. [11](IntuitionisticFuzzySet)Considerauniversaldiscourse W;thenanintuitionistic fuzzyset(IFS) Y in W isdepictedbytheorderedtripletsasfollows: ˜ Y = {w,µY (w),νY (w) | w ∈ W} where µY (w): W → [0, 1]denotesthemembershipfunctionand νY (w): W → [0, 1]denotesthe non-membershipfunctionoftheelement w intotheset Y,respectively,withtheconditions 0 ≤ µY (w) + νY (w) ≤ 1.

µY (w)

and νY (w) =

Remark1. If a2 = a2 = a3 = a3 thenTrIFN Y = a1 ≤ a1 ≤ a2 ≤ a2 ≤ a3 ≤ a3 ≤ a4 ≤ a4 isreducedinto triangularintuitionisticfuzzynumber(TIFN).(seeFigure 1)

Figure1. Trapeziodalintuitionisticfuzzynumber.

Definition3. [22]ATrIFN Y = {(a1, a2, a3, a4; µY );(a1, a2, a3, a4; νY )} issaidtobeanon-negative TrIFNifandonlyiff a1 ≥ 0.

Definition4. [22]ArithmeticoperationsonTrIFNs. Assumethat A = {(a1, a2, a3, a4);(a1, a2, a3, a4)} and B = {(b1, b2, b3, b4);(b1, b2, b3, b4)} betwoTrIFNs. Then

AIMSMathematics Volume6,Issue5,4556–4580.

4560
=                       
Definition2. [30](Trapeziodalintuitionisticfuzzynumber)Anintuitionisticfuzzynumber ˜ Y issaid tobetrapeziodalintuitionisticfuzzynumber(TrIFN)ifthemembershipfunction µY (w)and non-membershipfunction νY (w)isgivenby                        
w a1 a2 a1 , if a1 ≤ w ≤ a2, 1, if a2 ≤ w ≤ a3, a4 w a4 a3 , if a3 ≤ w ≤ a4, 0, ifotherwise.
a2 w a2 a1 , if a1 ≤ w ≤ a2, 0, if a2 ≤ w ≤ a3, w a3 a4 a3 , if a3 ≤ w ≤ a4, 1, ifotherwise. where a1 ≤ a1 ≤ a2 ≤ a2 ≤ a3 ≤ a3 ≤ a4 ≤ a4 andisrepresentedby Y = {(a1, a2, a3, a4; µY );(a1, a2, a3, a4; νY )}.

Addition : A + pB = a1 + pb1, a2 + pb2, a3 + pb3, a4 + pb4; a1 + pb1, a2 + pb2, a3 + pb3, a4 + pb4 , p ≥ 0 a1 + pb4, a2 + pb3, a3 + pb2, a4 + pb1; a1 + pb4, a2 + pb3, a3 + pb2, a4 + pb1 , p < 0

Multiplication : ˜ A × ˜ B = a1b1, a2b2, a3b3, a4b4; a1b1, a2b2, a3b3, a4b4 , when a1, b1 ≥ 0

Definition5. [22](Accuracyfunction)Theaccuracyfunction φ forTrIFN A = {(a1, a2, a3, a4);(a1, a2, a3, a4)} canberepresentedasfollows: φ( ˜ A) = (a1 + a2 + a3 + a4 + a1 + a2 + a3 + a4) 8

Theorem1. Supposethat ˜ A and ˜ B betwoTrIFNs.Thentheaccuracyfunction φ : IF( R ) → R isa linearfunctioni.e., φ(A + B) = φ(A) + pφ(B)forall p ∈ R . Proof: Letusconsiderthat ˜ A = {(a1, a2, a3, a4);(a1, a2, a3, a4)} and ˜ B = {(b1, b2, b3, b4);(b1, b2, b3, b4)} betwoTrIFNs.Then,basedonthenatureof p,twodifferentcaseswillarise: CaseI: when p ≥ 0,then A + pB = a1 + pb1, a2 + pb2, a3 + pb3, a4 + pb4; a1 + pb1, a2 + pb2, a3 + pb3, a4 + pb4 Therefore, φ( ˜ A + ˜ B) = a1 + pb1 + a2 + pb2 + a3 + pb3 + a4 + pb4 + a1 + pb1 + a2 + pb2 + a3 + pb3 + a4 + pb4 8 = a1 + a2 + a3 + a4 + a1 + a2 + a3 + a4 8 + pb1 + pb2 + pb3 + pb4 + pb1 + pb2 + pb3 + pb4 8

= φ( ˜ A) + pφ( ˜ B)

CaseII: when p < 0,then A + pB = a1 + pb4, a2 + pb3, a3 + pb2, a4 + pb1; a1 + pb4, a2 + pb3, a3 + pb2, a4 + pb1 Suchthat, φ(A + B) = a1 + pb4 + a2 + pb3 + a3 + pb2 + a4 + pb1 + a1 + pb4 + a2 + pb3 + a3 + pb2 + a4 + pb1 8 = a1 + a2 + a3 + a4 + a1 + a2 + a3 + a4 8 + pb1 + pb2 + pb3 + pb4 + pb1 + pb2 + pb3 + pb4 8 = φ( ˜ A) + pφ( ˜ B). Ineachcase,wehaveprovedthat φ(A + B) = φ(A) + pφ(B).Thusaccuracyfunction φ islinear. Definition6. [2](NeutrosophicSet(NS))Supposethat X beauniversaldiscoursesuchthat x ∈ X.A neutrosophicset A in X canbestatedbythreemembershipfunctionsnamely,truth µA(x),indeterminacy λA(x)andafalsity νA(x),whichcanberepresentedbythefollowingexpressions: A = {< x,µA(x),λA(x),νA(x) > |x ∈ X}

AIMSMathematics Volume6,Issue5,4556–4580.

4561

4562 where µA(x),λA(x)and νA(x)arerealstandardornon-standardsubsetsbelongto]0 , 1+[,alsogivenas, µA(x): X → ]0 , 1+[, λA(x): X → ]0 , 1+[,and νA(x): X → ]0 , 1+[.Thereisnorestrictiononthe sumof µA(x),λA(x)and νA(x),sowehave, 0 ≤ sup µA(x) + λA(x) + sup νA(x) ≤ 3+

Definition7. [2]Supposethat X beauniverseofdiscourse.Asinglevaluedneutrosophicset A canbe representedasfollows: A = {< x,µA(x),λA(x),νA(x) > |x ∈ X} where µA(x),λA(x)and νA(x) ∈ [0, 1]and0 ≤ µA(x) + λA(x) + νA(x) ≤ 3foreach x ∈ X

Definition8. [5]Considerthattherebetwodifferentsinglevaluedneutrosophicsets W and Y.The unionof W and Y alsoresultsinasinglevaluedneutrosophicset Z,i.e., Z = (W ∪ Y),whosetruth

µZ (x),indeterminacy λZ (x)andfalsity νZ (x)membershipfunctionscanberepresentedasfollows:

µZ (x) = max(µW (x),µY (x))

λZ (x) = max(λW (x),λY (x))

νZ (x) = min(νW (x),νY (x))foreach x ∈ X.

Definition9. [5]Considerthattherebetwodifferentsinglevaluedneutrosophicsets W and Y.The intersectionof W and Y alsoresultsinasinglevaluedneutrosophicset Z,i.e., Z = (W ∩ Y),whose truth µZ (x),indeterminacy λZ (x)andfalsity νZ (x)membershipfunctionscanberepresentedasfollows:

µZ (x) = min(µW (x),µY (x))

λZ (x) = min(λW (x),λY (x))

νZ (x) = max(νW (x),νY (x))foreach x ∈ X

4.Intuitionisticfuzzymultiobjectivelinearprogrammingproblem

Mostoften,real-lifeproblemsexhibitsoptimizationofmorethanoneobjectivesatatime.The mostpromisingsolutionsetthatsatisfieseachobjectiveefficientlyistermedasthebestcompromise solution.Hencetheconventionalformofmultiobjectivelinearprogrammingproblem(MOLPP)with k objectivesisgivenasfollows(4.1):

Minimize O(x) = [O1(x), O2(x), , Ok(x)] s.t. J j=1 aij x j ≥ bi, i = 1, 2, ··· , I1, J j=1 aij x j ≤ bi, i = I1 + 1, I1 + 2, ··· , I2, J j=1 aij x j = bi, i = I2 + 1, I2 + 2, ··· , I. x j ≥ 0, j = 1, 2, ··· , J.

(4.1)

where Ok(x) = K k=1 ckj x j, ∀ k = 1, 2, ··· , K isthe k-thobjectivefunctionandislinearinnature, bi, ∀ i = 1, 2, , I and; x j, ∀ j = 1, 2, , J aretherighthandsidesandasetofdecision variables,respectively.

Definition10. Assumethat G bethesetoffeasiblesolutionfor(4.1).Thenapoint x∗ issaidtobean efficientorParetooptimalsolutionof(4.1)ifandonlyiftheredoesnotexistany x ∈ G suchthat, AIMSMathematics

Volume6,Issue5,4556–4580.

Ok(x∗) ≥ Ok(x), ∀ k and Ok(x∗) > Ok(x)forallatleastone k.

Definition11. Apoint x∗ ∈ G issaidtobeweakParetooptimalsolutionfor(4.1)ifandonlyifthere doesnotexistany x ∈ G suchthat, Ok(x∗) ≥ Ok(x), ∀ k = 1, 2, ··· , K.

InMOLPP(4.1),ifalltheparametersareintuitionisticfuzzynumberthenitistermedas intuitionisticfuzzymultiobjectivelinearprogrammingproblem(IFMOLPP).Itisassumedthatallthe parameterspresentinproblem(4.2)istrapeziodalintuitionisticfuzzynumber.Thusthemathematical formulationofIFMOLPPcanbestatedasfollows(4.2):

Minimize OIF (x) = OIF 1 (x), OIF 2 (x), ··· , OIF k (x) s.t.

J j=1 ˜ aIF ij x j ≥ ˜ bIF i , i = 1, 2, ··· , I1, J j=1 aIF ij x j ≤ bIF i , i = I1 + 1, I1 + 2, , I2, J j=1 aIF ij x j = ˜ bIF i , i = I2 + 1, I2 + 2, ··· , I. x j ≥ 0, j = 1, 2, , J

(4.2)

where OIF k (x) = K k=1 ˜ ckj IF x j, ∀ k = 1, 2, ··· , K isthe k-thobjectivefunctionwithtrapeziodal intuitionisticfuzzyparameters.

Withtheaidofaccuracyfunction(Theorem1)whichislinear,theIFMOLPP(4.2)canbeconverted intothefollowingdeterministicMOLPP(4.3):

Minimize O (x) = O1(x), O2(x), , Ok(x) s.t. J j=1 aij x j ≥ bi , i = 1, 2, , I1, J j=1 aij x j ≤ bi , i = I1 + 1, I1 + 2, ··· , I2, J j=1 aij x j = bi , i = I2 + 1, I2 + 2, ··· , I. x j ≥ 0, j = 1, 2, , J

(4.3) where Ok(x) = φ ˜ OIF k (x) = K k=1 φ ˜ ckj IF x j, ∀ k = 1, 2, , K; bi = φ bIF i and aij = φ ˜ aIF ij ,for all i = 1, 2, , I, j = 1, 2, , J arethecrispversionofalltheobjectivefunctionsandparameters. Theorem2. AnefficientsolutionfordeterministicMOLPP(4.3)isalsoanefficientsolutionfor IFMOLPP(4.2). Proof: Assumethat X = (x1, x2, ··· , xn) beanefficientsolutionofproblem(4.3).Then X isfeasible forproblem(4.3),itmeansthatthefollowingconditionwillhold: J j=1 aij x j ≥ bi , i = 1, 2, , I1, J j=1 aij x j ≤ bi , i = I1 + 1, I1 + 2, ··· , I2, J j=1 aij x j = bi , i = I2 + 1, I2 + 2, ··· , I. x j ≥ 0, j = 1, 2, , J

AIMSMathematics Volume6,Issue5,4556–4580.

4563

Sinceithasbeenprovedthat φ islinear(Theorem1),thenwehave

J j=1 φ aIF ij x j ≥ φ ˜ bIF i , i = 1, 2, ··· , I1, J j=1 φ ˜ aIF ij x j ≤ φ ˜ bIF i , i = I1 + 1, I1 + 2, , I2, J j=1 φ aIF ij x j = φ bIF i , i = I2 + 1, I2 + 2, , I x j ≥ 0, j = 1, 2, , J

Consequently,wehave

J j=1 ˜ aIF ij x j ≥ bIF i , i = 1, 2, , I1, J j=1 aIF ij x j ≤ bIF i , i = I1 + 1, I1 + 2, ··· , I2, J j=1 ˜ aIF ij x j = ˜ bIF i , i = I2 + 1, I2 + 2, , I. x j ≥ 0, j = 1, 2, ··· , J

Hence, X isafeasiblesolutionforproblem(4.2). Moreover,since X isanefficientsolutionforproblem(4.3),theredoesnotexistany X∗ = x∗ 1, x∗ 2, ··· , x∗ n suchthat Ok(X∗) ≤ Ok(X) ∀ k = 1, 2, ··· , K and Ok(X∗) < Ok(X)foratleastone k.Thuswehaveno X∗ suchthatMin K k=1 φ ˜ ckj IF x j ≤ Min K k=1 φ ˜ ckj IF x∗ j ∀ k = 1, 2, , K andMin K k=1 φ ckj IF x j < Min K k=1 φ ckj IF x∗ j foratleastone k. Since φ isalinearfunction(Theorem1),wehaveno X∗ suchthat Min K k=1 φ ˜ ckj IF x j ≤ Min K k=1 φ ˜ ckj IF x∗ j ∀ k = 1, 2, , K and Min K k=1 φ ckj IF x j < Min K k=1 φ ckj IF x∗ j foratleastone k Thus X isefficientsolutionforproblem(4.2).

Considerasingle-objectiveintuitionisticfuzzylinearprogrammingproblem(SOIFLPP)asgivenin problem(4.4):

Minimize ˜ OIF (x) s.t.

J j=1 aIF ij x j ≥ ˜ bIF i , i = 1, 2, ··· , I1, J j=1 ˜ aIF ij x j ≤ bIF i , i = I1 + 1, I1 + 2, , I2, J j=1 aIF ij x j = bIF i , i = I2 + 1, I2 + 2, ··· , I. x j ≥ 0, j = 1, 2, ··· , J.

(4.4)

Usingtheaccuracyfunctionwhichislinearinnature,SOIFLPP(4.4)canbeconvertedintothe followingdeterministicsingle-objectivelinearprogrammingproblem(SOLPP)(4.5):

Minimize O (x) s.t.

J j=1 aij x j ≥ bi , i = 1, 2, , I1, J j=1 aij x j ≤ bi , i = I1 + 1, I1 + 2, ··· , I2, J j=1 aij x j = bi , i = I2 + 1, I2 + 2, ··· , I. x j ≥ 0, j = 1, 2, , J

AIMSMathematics

(4.5)

Volume6,Issue5,4556–4580.

4564

Theorem3. AnoptimalsolutionforthedeterministicSOLPP(4.5)isalsoanoptimalsolutionfor SOIFLPP(4.4). Proof. SameasProof2. HencetheoptimalsolutionforanIFLPPcanbeeasillytransformedintodeterministicLPPwiththe aidofaccuracyfunction.Now,wewillproceedtowardsthesolutionproceduresoftheMOLPP(4.3) underneutrosophicenvironment.

5.Proposedneutrosophicfuzzyoptimizationtechnique

Manymultiobjectiveoptimizationtechniquesarepopularamongresearchers.Basedonthefuzzy set,differentfuzzyoptimizationmethodcameintoexistence.Inthefuzzyprogrammingapproach,the marginalevaluationofeachobjectivefunctionisdepictedbyonlythemembershipfunctionsandcan beachievedbymaximizingit.Theextensionofthefuzzyoptimizationmethodispresentedby introducingintuitionisticfuzzyoptimizationtechniques.Itiscomparativelymoreadvancedthanthe fuzzytechniquebecausethemarginalevaluationofeachobjectivefunctionisdepictedbythe membershipandnon-membershipfunctions,whichcanbeachievedbymaximizingthemembership andminimizingthenon-membershipfunctions,respectively.Thereal-lifecomplexitymostoften createstheindeterminacysituationwhilemakingoptimaldecisions.Apartfromtheacceptanceand rejectiondegreesinthedecision-makingprocess,theindeterminacydegreealsohasmuch importance.Thustocovertheindeterminacydegreeoftheelementintothefeasiblesolutionset,F. Smarandache[32]investigatedaneutrosophicset.Thename“neutrosophic”istheadvance combinationoftwoexplicitterms,namely;“neutre”extractedfromFrenchmeans,neutral,and “sophia”adoptedfromGreekmeans,skill/wisdom,thatunanimouslyprovidethedefinition “knowledgeofneutralitydegree”(see[2, 32]).TheNSconsidersthreesortsofmembershipfunctions, suchastruth(degreeofbelongingness),indeterminacy(degreeofbelongingnessuptosomeextent), andafalsity(degreeofnon-belongingness)degreesintothefeasiblesolutionset.Theideaof independentindetermincydegreedifferstheNSwithalltheuncertaindecisionsetssuchasFSand IFS.Theupdatedliteratureworksolelyhighlightsthatmanypractitionersorresearchershavetaken thedeepinterestintheneutrosophicresearchfield(see,[3–5, 26]).TheNSresearchdomainwould getexposureinthefutureandassistindealingwithindeterminacyinthedecision-makingprocess. ThisstudyalsofetchesthenovelideasofneutrosophicoptimizationtechniquesbasedontheNS.The marginalevaluationofeachobjectivefunctionisquantifiedbythetruth,indeterminacy,andfalsity membershipfunctionsundertheneutrosophicdecisionset.ThustheNSplaysavitalrolewhile optimizingthemultiobjectiveoptimizationproblemsbyincorporating,executing,andimplementing theindeterminacydegrees.

R.E.Bellman,etal.[12]firstpropoundedtheideaofafuzzydecisionset.Afterthat,itiswidely adoptedbymanyresearchers.Thefuzzydecisionconceptcomprisesfuzzydecision(D),fuzzygoal (G),andfuzzyconstraints(C),respectively.Herewerecallthemostextensivelyusedfuzzydecision setwiththeaidoffollowingmathematicalexpressions:

Consequently,wealsodepicttheneutrosophicdecisionset DN ,whichcontemplateoverneutrosophic objectivesandconstraintsasfollows:

AIMSMathematics

Volume6,Issue5,4556–4580.

4565
D = O ∩ C
DN = (∩K k
k)(∩I i
Ci) = (x
D(x)
D
x)
D
x))
=1O
=1
(
(

where

µD(x) = min

λD(x) = max

µO1 (x),µO2 (x),...,µOK (x)

µC1 (x),µC2 (x),...,µCI (x) ∀ x ∈ X

λO1 (x),λO2 (x),...,λOK (x) λC1 (x),λC2 (x),...,λCI (x) ∀ x ∈ X

νD(x) = max

νO1 (x),νO2 (x),...,νOK (x) νC1 (x),νC2 (x),...,νCI (x) ∀ x ∈ X

where µD(x),λD(x)and νD(x)arethetruth,indeterminacyandafalsitymembershipfunctionsof neutrosophicdecisionset DN respectively.

InordertodepictthedifferentmembershipfunctionsforMOLPP(4.3),theminimumandmaximum valuesofeachobjectivefunctionshavebeenrepresentedby Lk and Uk and;canbeobtainedasfollows:

Uk = max[Ok(x)]and Lk = min[Ok(x)] ∀ k = 1, 2, 3,..., K. (5.1)

Theboundsfor k-thobjectivefunctionundertheneutrosophicenvironmentcanbeobtainedas follows:

U µ k = Uk, Lµ k = Lk fortruthmembership (5.2) U λ k = Lµ k + sk, Lλ k = Lµ k forindeterminacymembership (5.3) U ν k = U µ k , Lν k = Lµ k + tk forfalsitymembership (5.4) where sk and tk ∈ (0, 1)arepredeterminedrealnumbersprescribedbydecision-makers.

5.1.Variousmembershipfunctions

InMOPPs,themarginalevaluationofeachobjectivefunctionisdepictedbyitsrespective membershipfunctions.Ingeneral,themostextensiveandwidelyusedmembershipfunctionisa linearoneduetoitssimplestructureandmoreaccessibleimplications.Thelinear-typemembership functioncontemplatesovertheconstantmarginalrateofsatisfactionordissatisfactiondegrees towardsanobjective.Mostcommonly,themarginalevaluationshavebeenevaluatedbyusinglinear membershipfunctiontoobtaintheobjective.However,itmaybepossibletorepresenteach objective’saspirationlevelwiththeaidofnon-linearmembershipfunction.Theflexiblebehaviorof non-linearmembershipalsofunctionswellenoughtodeterminethemarginalevaluationofobjectives satisfactiondegree.Italsodependsonsomeparameters’value,whichiswellenoughtoefficiently executetheDM(s)strategy.Anexponentialmembershipreflectsthesituationwhenthe decision-makerisworseoff concerninganobjectiveandchoosesahighermarginalrateof satisfaction.Thusbypreferringanexponentialmembershipfunction,thedecision-makercanalso reducedualitygapsbyselectingappropriateshapeparametersinvolvedinthedevelopmentof membershipfunction.Ahyperbolicmembershipfunctionshowstheflexiblecharacteristicbehavior fortheobjectivefunction.Thehyperbolicmembershipfunctionshowstheconvexitybehaviorovera portionoftheobjectiveandrevealsconcavitycharacteristicfeaturesfortheremainingportion.A worse-off situationofthedecisionconcerningagoalisdepictedbytheconvexportionofthe hyperbolicmembershipfunction,andthedecision-makerintendstogainahighermarginalrateof satisfactionlevel.Despitethis,abetter-off conditionofthedecisionconcerningagoalisdepictedby

AIMSMathematics

Volume6,Issue5,4556–4580.

4566

theconcaveportionofthehyperbolicmembershipfunction,andthedecision-makerintendstogaina lowermarginalrateofsatisfactionlevelconcerningthatgoal.Therefore,todealwiththeIFMOLPPs, onecanapplylinear,exponential,orhyperbolicmembershipfunctions,dependinguponthechoiceof decision-maker.

Thusthelinear,exponential,andhyperbolicmembershipfunctionsareconstructedunderthe neutrosophicenvironment.Eachofthemisdefinedforthetruth,indeterminacy,andafalsity membershipfunction,whichseemstobemorerealistic.

(O

(5.5)

(Ok(x

ν

(5.7)

Intheabovecase, L(.) k U (.) k forall k objectivefunction.Ifforanymembership L(.) k = U (.) k ,thenthe valueofthesemembershipwillbeequalto1(seeFigure 2).

Figure2. Diagrammaticrepresentationoftruth,indeterminacyandafalsitymembership degreeforobjectivefunction.

AIMSMathematics Volume6,Issue5,4556–4580.

4567
µL k (Ok(x)),indeterminacy λL k
k
µ L k (Ok(x)) =              1 ifOk(x) ≤ Lµ k Uµ k Ok (x) Uµ k Lµ k ifLµ k ≤ Ok(x) ≤ U µ k 0
λL k
)) =             
Uλ k Ok (x) Uλ k Lλ k
ν L
=           
x
ν k
ν
5.1.1.Linear-typemembershipfunctionsapproach(LTMFA) Thelinear-typetruth ν
(x))andafalsity νL k (Ok(x))membership functionsunderneutrosophicenvironmentcanbefurnishedasfollows:
ifO
k(x) ≥ U µ k
1 ifOk(x) ≤ Lλ k
ifLλ k ≤ Ok(x) ≤ U λ k 0 ifOk(x) ≥ U λ k (5.6)
k (Ok(x))
0 ifOk(x) ≤ Lν k Ok (
) L
U
k L
k
ifL
k ≤ Ok(x) ≤ U ν k 1 ifOk(x) ≥ U ν k

Usingtheconceptof[12],wemaximizetheobjectivefunctiontoreachtheoptimalsolutionofeach objectives.Themathematicalexpressionforobjectivefunctionisdefinedasfollows:

Maxmink=1,2,3,...,K µ( ) k (Ok(x)) Minmaxk=1,2,3,...,K λ( ) k (Ok(x)) Minmaxk=1,2,3,...,K ν( ) k (Ok(x)) subjectto alltheconstraintsof(4.3)

(5.8)

Also,assumethat µL k (Ok(x)) ≥ α, λL k (Ok(x)) ≤ β and νL k (Ok(x)) ≤ γ,forall k. Usingauxiliaryparameters α,β and γ,theproblem(5.8)canbetransformedintothefollowingproblem (5.9): (LTMFA) Max α β γ subjectto Ok(x) + (U µ k Lµ k )α ≤ U µ k , Ok(x) (U λ k Lλ k )β ≤ Lλ k , Ok(x) (U ν k Lν k )γ ≤ Lν k , α ≥ β,α ≥ γ,α + β + γ ≤ 3 α,β,γ ∈ (0, 1) alltheconstraintsof(4.3)

(5.9)

Remark2. Theultimateaimofproblem(5.9)(LTMFA)manifeststhemaximizationofminimum possibilitylevel(truth)toacceptthebestpossiblesolutionand;minimizationofmaximumpossibility level(indeterminacy)and(afalsity)torejecttheworstpossiblesolutionbyconsideringallthe objectivefunctionsatatime.Itmeansthatwetrytodetermineasolutioninsuchawaythatit maximizestheminimumtruthdegree(acceptance)andminimizethemaximumindeterminacy (rejectionuptosomeextent)andafalsity(rejection)degreesbytakingallobjectivessimultaneously, toattaintheoptimalcompromisesolution.

Theorem4. Auniqueoptimalsolutionofproblem(5.9)(LTMFA)isalsoanefficientsolutionforthe problem(4.3). Proof: Supposethat x, α, β, γ beauniqueoptimalsolutionofproblem(5.9)(LTMFA).Then, α β γ > (α β γ) forany (x,α,β,γ) feasibletotheproblem(5.9)(LTMFA).Onthecontrary, assumethat x, α, β, γ isnotanefficientsolutionoftheproblem(4.3).Forthat,thereexists x

AIMSMathematics Volume6,Issue5,4556–4580.

4568
U
∗ (x∗ x) feasibletoproblem(4.3),suchthat Ok(x∗) ≤ Ok(¯ x)forall k = 1, 2, ··· , K and Ok(x∗) < Ok(¯ x)foratleastone k Therefore,wehave Ok (x∗) Lk Uk Lk ≤ Ok (¯ x) Lk Uk Lk forall k = 1, 2, ··· , K and Ok (x∗) Lk Uk Lk < Ok (¯ x) Lk Uk Lk foratleastone k. Hence,max k Ok (x∗) Lk Uk Lk ≤ (<)max k Ok (¯ x) Lk Uk Lk Supposethat γ
= max k
k Ok (x∗) Uk Lk ,then γ∗ ≤ (<)¯γ. Also,considerthat β∗ = max k Uk Ok (x∗) Uk Lk ,then β∗ ≤ (<) β Inthesamemanner,wehave Uk Ok (x∗) Uk Lk ≥ Uk Ok (¯ x) Uk Lk forall k = 1, 2, , K and Uk Ok (x∗) Uk Lk > Uk Ok (¯ x) Uk Lk forat leastone k. Thus,min k Uk Ok (x∗) Uk Lk ≥ (>)min k Uk Ok (¯ x) Uk Lk

Assumethat

Lk ,thisgives α β γ < (α

β

γ∗) whichmeansthatthe solutionisnotuniqueoptimal.Thus,wehavearrivedatacontradictionwiththefactthat x, α, β, γ is theuniqueoptimalsolutionof(LTMFA).Therefore,itisalsoanefficientsolutionoftheproblem (5.9).ThiscompletestheproofofTheorem4.

E k (Ok(x)),indeterminacy λE k (Ok(x))andafalsity νE k (Ok(x))membership functionsunderneutrosophicenvironmentcanbestatedasfollows:

(5.11)

ν

(5.12) where d isthemeasuresofvaguenessdegree(shapeparameter)andassignedbythedecision-makers.

νE k (Ok(x)) =

(5.13)

Followedbyproblem(5.8),weassumethat µE k (Ok(x)) ≥ α, λE k (Ok(x)) ≤ β and νE k (Ok(x)) ≤ γ,for all k.Usingauxiliaryparameters α,β and γ,theproblem(5.8)canbeconvertedintothefollowing problem(5.13): (ETMFA) Max α β γ subjectto e d Ok (x) Lµ k Uµ k Lµ k e d 1 e d ≥ α, e d Uλ k Ok (x) Uλ k Lλ k e d 1 e d ≤ β, e d Uν k Ok (x) Uν k Lν k e d 1 e d ≤ γ, α ≥ β,α ≥ γ,α + β + γ ≤ 3 α,β,γ ∈ (0, 1) alltheconstraintsof(4.3)

AIMSMathematics Volume6,Issue5,4556–4580.

4569
Uk
∗)
α∗ = min k
Ok (x
Uk
µ
µ E k (Ok(x)) =                    1 ifOk(x) ≤ Lµ k e d Ok (x) Lµ k Uµ k Lµ k e d 1 e d ifLµ k ≤ Ok(x) ≤ U µ
0
λE k (Ok(x)) =                    1 ifO
λ
e d Uλ k Ok (x) Uλ k Lλ k e d 1 e
λ
5.1.2.Exponential-typemembershipfunctionsapproach(ETMFA) Theexponential-typetruth                 
k
ifOk(x) ≥ U µ k (5.10)
k(x)
L
k
d ifL
k ≤
Ok(x) ≤ U λ k 0 ifOk(x) ≥ U λ k
0 ifOk(x) ≤ Lν k e d Uν k Ok (x) U
k Lν k e d 1 e d ifLν k ≤ Ok(x) ≤ U ν k 1 ifOk(x) ≥ U ν k

Remark3. If d → 0,thentheexponential-typemembershipfunctionswillbereducedintolinear-type membershipfunctions.

k ≤ O

 e

d Ok (¯ x) Lk Uk Lk e d 1 e d

 . If α∗ = min k 

) ≤ Ok(¯ x)forall k = 1, 2, ··· , K and Ok(x∗) < Ok(¯ x)foratleastone k. Consequently,wehave     

 ,then α∗ ≥ (>)¯α. Similarly,wehave Uk Ok (x∗) Uk Lk ≥ Uk Ok (¯ x) Uk Lk forall k = 1, 2, , K and Uk Ok (x∗) Uk Lk > Uk Ok (¯ x) Uk Lk foratleastone k Consequently,itgives e d Uk Ok (x ∗) Uk Lk e d 1 e d ≤ e d Uk Ok (¯ x) Uk Lk e d 1 e d forall k = 1, 2, ··· , K and e d Uk Ok (x ∗) Uk Lk e d 1 e d < e d Uk Ok (¯ x) Uk Lk e d 1 e d foratleastone k

k        e d

Hence,max k x) Uk

       ≤ (<        ,wehave β∗ ≤ (<) β. Again,byconsidering γ∗ = max k

Uk Ok (¯        ,weget γ∗ ≤ (<)¯γ. Thisgives α β γ < (α∗ β∗ γ∗),thatcontradictsthefactthat x, α, β, γ istheuniqueoptimal solutionoftheproblem(5.13)(ETMFA).Hence,theTheorem5isproved. 5.1.3.Hyperbolic-typemembershipfunctionsapproach(HTMFA) Thehyperbolic-typetruth µH k (Ok(x)),indeterminacy λH k (Ok(x))andafalsity νH k (Ok(x))membership functionsunderneutrosophicenvironmentcanbedepictedasfollows: µ H k (Ok(x)) =

Lk e d 1 e                1 ifOk(x) ≤ Lµ k 1 2 1 + tanh θk U µ k + Lµ k 2 Ok(x) ifLµ k ≤ Ok(x) ≤ U µ k 0 ifOk(x) ≥ U µ k

)max        e d Uk Ok (x ∗) Uk Lk e d 1 e d

(5.14) AIMSMathematics Volume6,Issue5,4556–4580.

4570
Ok
x∗) Lk Uk L
k
x
e d Ok (x ∗) Lk Uk Lk e d 1 e d ≥ e d Ok (¯ x) Lk Uk
e
Uk
e
d
x
Lk Uk
       e d Ok (x ∗) Lk Uk Lk e d 1 e d       
    
   
Theorem5. Auniqueoptimalsolutionofproblem(5.13)(ETMFA)isalsoanefficientsolutionforthe problem(4.3). Proof. Thiswillbeprovedbyarrivingatacontradiction. Supposethat x, α, β, γ beauniqueoptimalsolutionofproblem(5.13)(ETMFA)whichisnotan efficientsolutionfortheproblem(4.3).Then,thereexists x∗ (x∗ x) feasibletoproblem(4.3),such that Ok(x     
(
) Lk Uk Lk forall k = 1, 2, , K and Ok (x∗) Lk Uk Lk < Ok (¯ x) Lk Uk Lk foratleastone k. Hence,wehave
Lk
d 1 e d forall k = 1, 2, ··· , K and e d Ok (x
) Lk
Lk
d 1 e d > e
Ok (¯
)
Lk e d 1 e d foratleastone k. Thus,min k
≥ (>)min k
 e
d Ok (x ∗) Lk Uk Lk e d 1 e d
 d
            Assuming β∗ = max k
 e d Uk Ok (x ∗) Uk Lk e d 1 e d        e d Uk Ok (x ∗) Uk Lk e d 1 e d

0 ifOk(x) ≤ Lν k 1 2 1 + tanh θk Ok(x) U ν k + Lν k 2 ifLν k ≤ Ok(x) ≤ U ν k 1 ifOk(x) ≥ U ν k

(5.16) where θk = 6 Uk Lk , ∀ k = 1, 2,..., K. Movingwiththesameprocedureuptoproblem(5.8),wesupposethat µH k (Ok(x)) ≥ α, λH k (Ok(x)) ≤ β and νH k (Ok(x)) ≤ γ,forall k.Usingauxiliaryparameters α,β and γ,theproblem(5.8)canbe transformedintothefollowingproblem(5.17): Max α β γ subjectto 1 2 1 + tanh θk U µ k + Lµ k 2 Ok(x) ≥ α, 1 2 1 + tanh θk U λ k + Lλ k 2 Ok(x) ≤ β, 1 2 1 + tanh θk Ok(x) U ν k + Lν k 2 ≤ γ, α ≥ β,α ≥ γ,α + β + γ ≤ 3 α,β,γ ∈ (0, 1),θk = 6 Uk Lk , ∀ k = 1, 2,..., K alltheconstraintsof(4.3)

Equivalently,wehaveproblem(5.18)asfollows:

(HTMFA) Max α β γ subjectto θkOk(x) + tanh 1 (2α 1) ≤ θk 2 U µ k + Lµ k , θkOk(x) tanh 1 (2β 1) ≤ θk 2 U µ k + Lµ k , θkOk(x) tanh 1 (2γ 1) ≤ θk 2 U µ k + Lµ k , α ≥ β,α ≥ γ,α + β + γ ≤ 3, α,β,γ ∈ (0, 1),θk = 6 Uk Lk , ∀ k = 1, 2,..., K alltheconstraintsof(4.3)

(5.17)

(5.18)

Theorem6. Auniqueoptimalsolutionofproblem(5.18)(HTMFA)isalsoanefficientsolutionforthe problem(4.3).

Proof. Letusconsiderthat x, α, β, γ beauniqueoptimalsolutionofproblem(5.18)(HTMFA),but notanefficientsolutionfortheproblem(4.3).Thisgives,thereexists x∗ (x∗ x) feasibletoproblem (4.3),suchthat Ok(x∗) ≤ Ok(¯ x)forall k = 1, 2, ··· , K and Ok(x∗) < Ok(¯ x)foratleastone k. AIMSMathematics

Volume6,Issue5,4556–4580.

4571
=               
ν
k
=               
λH k (Ok(x))
1 ifOk(x) ≤ Lλ k 1 2 1 + tanh θk U λ k + Lλ k 2 Ok(x) ifLλ k ≤ Ok(x) ≤ U λ k 0 ifOk(x) ≥ U λ k (5.15)
H
(Ok(x))

5.2.Proposedsolutionalgorithm

Thestep-wisesolutionalgorithmissummarizedasfollows:

Step-1. FormulatetheIFMOLPPasgiveninproblem(4.2).

Step-2. Usingaccuracyfunction,obtainthecrispversionofIFMOLPPasgiveninproblem(4.3). Step-3. Solveeachobjectivefunctionindividuallyanddeterminetheupperandlowerboundbyusing Eq(5.1).

Step-4. Withtheaidof Uk and Lk,calculatetheupperandlowerboundfortruth,indeterminacyanda falsitymembershipunderneutrosophicenvirnmentasgiveninEqs(5.2)–(5.4).

Step-5. Elicitthedifferent-typesofmembershipfunctionsunderneutrosophicenvironmentbyusing Eqs(5.5–5.7),(5.10–5.12)and(5.14–5.16)accordingtodecision-makers’preferencerespectively.

Step-6. Developtheneutrosophicoptimizationmodelssuchas(LTMFA)or(ETMFA)or(HTMFA) underthegivensetofwell-definedconstraints.

Step-7. SolvetheneutrosophicMOPPs(5.9),(5.13)and(5.18)todeterminetheoptimalcompromise outcomesbyapplyingtheappropriatemethodsordifferentoptimizationsoftwarepackages.

6.Numericalillustrations

Followingnumericalexamplesaresolvedusingdifferentmethodsbymanyresearchers.Allthe crispMOLPPsarecodedinAMPLlanguageandsolvedwithhelpofsolverKintro10.3.0viaNEOS serverversion5.0on-linefacilityexecutedbyWisconsinInstitutesforDiscoveryattheUniversityof WisconsininMadisonforsolvingoptimizationproblems,see[15, 27].

Example1. [30]Thisnumericalexampleassumesthatthetrapezoidalintuitionisticfuzzynumber representsalltheparameters.Thusthemathematicalformulationoftheproblemcanbestatedas

AIMSMathematics

Volume6,Issue5,4556–4580.

4572
Simultaneously,tanh θk Uk + Lk 2 Ok(x∗) ≥ tanh θk Uk + Lk 2 Ok(¯ x) forall k = 1, 2, ··· , K and tanh θk Uk + Lk 2 Ok(x∗) > tanh θk Uk + Lk 2 Ok(¯ x) foratleastone k Furthermore,itgives, min k tanh θk Uk + Lk 2 Ok(x∗) ≥ (>)min k tanh θk Uk + Lk 2 Ok(¯ x) . If α∗ = min k 1 2 tanh θk Uk + Lk 2 Ok(x∗) + 1 2 ,then α∗ ≥ (>)¯α. Similarly,wehave β∗ = max k 1 2 tanh θk Ok(x∗) Uk + Lk 2 + 1 2 ,then β∗ ≤ (<) β and γ∗ = min k 1 2 tanh θk Ok(x∗) Uk + Lk 2 + 1 2 ,then γ∗ ≥ (>)¯γ Thus,weget α β γ < (α∗ β∗ γ∗).Thisarisesacontradictionwiththefactthat α β γ istheuniqueoptimalsolutionoftheproblem(5.18)(HTMFA).Hence,theTheorem6isproved.

follows:

Min OIF 1 (x) = ˜ 5IF x1 + ˜ 3IF x2 Min ˜ OIF 2 (x) = ˜ 2∗IF x1 + ˜ 7IF x2 s.t. ˜ 2@IF x1 + ˜ 4IF x2 ≥ 25IF 1∗IF x1 + 1@IF x2 ≥ 10IF ˜ 4IF x1 + 5IF x2 ≥ 50IF x1, x2 ≥ 0

(6.1)

TheintuitionisticfuzzyparametersrelatedtoExample1arefurnishedinTable 1.Usingthe accuracyfunction(Definition5),thecrispversionoftheProblem(6.1)canbedepictedasfollows:

Min O1(x) = 5 125x1 + 3 33x2 Min O2(x) = 2.37x1 + 6.2x2 s.t. 1 8x1 + 3 62x2 ≥ 23 75 0 9x1 + 1 06x2 ≥ 9 6 3.62x1 + 5.125x2 ≥ 51.8 x1, x2 ≥ 0.

(6.2) Table1. Example1:Intuitionisticfuzzyparameters. 25IF = (22, 25, 25, 25;18, 25, 25, 25) 10IF = (9, 10, 10, 10;8, 10, 10, 10) 50IF = (50, 50, 50, 55;50, 50, 50, 60) 5IF = (4, 5, 5, 6;4, 5, 5, 7) 3IF = (3, 3, 3, 4;3, 3, 3, 4 5) 2@IF = (1 5, 2, 2, 2;1, 2, 2, 2) 2∗IF = (2, 2, 2, 3;2, 2, 2, 4) 7IF = (7, 7, 7, 7 5;6, 7, 7, 8) 4IF = (3, 4, 4, 4;2, 4, 4, 4) ˜ 1@IF = (1, 1, 1, 1;0 5, 1, 1, 2) ˜ 1∗IF = (0 5, 1, 1, 1;0 2, 1, 1, 1 5) • UsingtheLTMFA(5.9),Problem(6.2)willbeequivalenttothefollowing(6.3): Max α β γ s.t. 5.125x1 + 3.33x2 + (U µ 1 Lµ 1 )α ≤ U µ 1 , 5.125x1 + 3.33x2 (U λ 1 Lλ 1 )β ≤ Lλ 1 , 5 125x1 + 3 33x2 (U ν 1 Lν 1)γ ≤ Lν 1, 2 37x1 + 6 2x2 + (U µ 2 Lµ 2 )α ≤ U µ 2 , 2.37x1 + 6.2x2 (U λ 2 Lλ 2 )β ≤ Lλ 2 , 2.37x1 + 6.2x2 (U ν 2 Lν 2)γ ≤ Lν 2, 1 8x1 + 3 62x2 ≥ 23 75 0 9x1 + 1 06x2 ≥ 9 6 3.62x1 + 5.125x2 ≥ 51.8 α ≥ β,α ≥ γ,α + β + γ ≤ 3 α,β,γ ∈ (0, 1) x1, x2 ≥ 0

(6.3) AIMSMathematics Volume6,Issue5,4556–4580.

4573

UsingtheETMFA(5.13),Problem(6.2)willbeequivalenttothefollowing(6.4):

Max α β γ s.t. e d 5 125x1+3 33x2 Lµ 1 Uµ 1 Lµ 1 e d 1 e d ≥ α, e d Uλ 1 5 125x1 3 33x2 Uλ 1 Lλ 1 e d 1 e d ≤ β, e d Uν 1 5 125x1 3 33x2 Uν 1 Lν 1 e d 1 e d ≤ γ, e d 2 37x1+6 2x2 Lµ 2 Uµ 2 Lµ 2 e d 1 e d ≥ α, e d Uλ 2 2 37x1 6 2x2 Uλ 2 Lλ 2 e d 1 e d ≤ β, e d Uν 2 2 37x1 6 2x2 Uν 2 Lν 2 e d 1 e d ≤ γ, 1.8x1 + 3.62x2 ≥ 23.75 0.9x1 + 1.06x2 ≥ 9.6 3 62x1 + 5 125x2 ≥ 51 8 α ≥ β,α ≥ γ,α + β + γ ≤ 3 α,β,γ ∈ (0, 1) x1, x2 ≥ 0.

UsingtheHTMFA(5.18),Problem(6.2)willbeequivalenttothefollowing(6.5):

Max α β γ s.t.

θ1(5 125x1 + 3 33x2) + tanh 1 (2α 1) ≤ θ1 2 U µ 1 + Lµ 1 , θ1(5.125x1 + 3.33x2) tanh 1 (2β 1) ≤ θ1 2 U µ 1 + Lµ 1 , θ1(5.125x1 + 3.33x2) tanh 1 (2γ 1) ≤ θ1 2 U µ 1 + Lµ 1 , θ2(2 37x1 + 6 2x2) + tanh 1 (2α 1) ≤ θ2 2 U µ 2 + Lµ 2 , θ2(2.37x1 + 6.2x2) tanh 1 (2β 1) ≤ θ2 2 U µ 2 + Lµ 2 , θ2(2.37x1 + 6.2x2) tanh 1 (2γ 1) ≤ θ2 2 U µ 2 + Lµ 2 , 1.8x1 + 3.62x2 ≥ 23.75 0 9x1 + 1 06x2 ≥ 9 6 3 62x1 + 5 125x2 ≥ 51 8

α ≥ β,α ≥ γ,α + β + γ ≤ 3 α,β,γ ∈ (0, 1) x1, x2 ≥ 0

(6.4)

(6.5)

Onsolvingtheneutrosophicoptimizationmodel(6.3),(6.4)and(6.5),thesolutionresultsare summarizedinTable 2.Itisclearthattheobjectivevaluesarequitebetterthan[30]methodforallthe

AIMSMathematics Volume6,Issue5,4556–4580.

4574 •

4575

membershipfunctions.Also,theproposedneutrosophic’sperformancemeasuresofvarious membershipfunctionscanbedepictedasHyperbolic>Exponential>Linear.Since[30]didnotusethe Exponentialmembershipfunctions;therefore,itisunjustifiabletoperformcomparisonsamongthe membershipfunction.However,ourproposedmethod’smaximumattaintmentofacceptancedegree isbetterachievedandrevealsitssuperiorityover[30]method.

Table2. Example1:Optimalsolutionresults.

MethodsMembershipObjectiveDeviationsfrom αβγ functionsvaluesidealsolutions

Linear(38.97,37.12)(8.82,5.86)0.76NANA

[30]Parabolic(43.02,39.79)(12.87,8.53)0.88NANA Hyperbolic(39.82,38.50)(9.67,7.24)1.00NANA Linear(38.56,38.28)(8.41,7.02)0.770.710.87

ProposedneutrosophicapproachExponential(39.42,38.15)(9.27,6.89)0.910.670.81 Hyperbolic(39.07,38.03)(8.92,6.77)1.000.510.76

NA:NotApplicable.

Example2. [22]Considerthefollowingnumericalillustration.

Min OIF 1 (x) = 2 i=1 3 j=1 c(1)IF ij xij

Min ˜ OIF 2 (x) = 2 i=1 3 j=1 c(2)IF ij xij

s.t.

2 i=1 3 j=1 xij ≤ (50, 70, 90, 110;40, 60, 100, 120)

2 i=1 3 j=1 xij ≤ (35, 55, 75, 95;25, 45, 85, 105) 2 i=1 3 j=1 xij ≥ (25, 45, 65, 85;15, 35, 75, 95) 2 i=1 3 j=1 xij ≥ (15, 25, 35, 45;10, 20, 40, 50) 2 i=1 3 j=1 xij ≥ (45, 55, 65, 75;40, 50, 70, 80) xij ≥ 0 ∀ i = 1, 2& j = 1, 2, 3

TheintuitionisticfuzzyparametersrelatedtoExample2aresummarizedinTable 3.Onsolving eachobjectiveindividually(Example2),wegettheupperandlowerbounds U1 = 8631.25, U2 = 3163.75, L1 = 7981.25and L2 = 2923.75forthefirstandsecondobjective functionsrespectively.OnimplementingtheneutrosophicMOPPs(5.9),(5.13)and(5.18),the solutionresultsaresummarizedinTable 4.Forthisexample,theobjectivevaluesarequitebetter than[22]methodforallthemembershipfunctions.Theperformancemeasuresoftheproposed neutrosophicofvariousmembershipfunctionscanbedepictedasHyperbolic>Exponential>Linear. Thus,theproposedneutrosophicoptimizationtechniqueoutperforms[22]methodforallthe membershipfunctions.However,themaximumattaintmentofacceptancedegreeisbetterachievedby ourproposedtechniquethatshowsthemostpromisingmethodover[22]method.

Table3. Example2:Intuitionisticfuzzyparametersfor˜

, 50, 70, 90)˜

(75, 85, 105, 115;70, 80, 110, 120)

13

(40,

,

,

,

,

,

(50, 60, 80, 100;45, 55, 95, 105)˜ c(1)IF 23 = (25, 35, 55, 65;20, 30, 40, 70) c(2)IF 11 = (6, 10, 12, 16;4, 8, 14, 18)˜ c(2)IF 12 = (25, 35, 40, 50;20, 30, 45, 55)˜ c(2)IF 13 = (14, 22, 26, 34;10, 18, 30, 38) c(2)IF 21 = (20, 30, 35, 45;15, 25, 40, 50)˜ c(2)IF 22 = (8, 14, 17, 23;5, 11, 20, 26)˜ c(2)IF 23 = (16, 24, 28, 36;12, 20, 32, 40)

AIMSMathematics

Volume6,Issue5,4556–4580.

c(1)IF
c(1)IF ij and˜ c(2)IF ij
11 = (15, 25, 35, 45;10, 20, 40, 50)˜ c(1)IF 12 = (45, 55, 65, 75;40
c(1)IF
=
c(1)IF 21 =
50
60
70;35
45
65
75)˜ c(1)IF 22 =

Table4. Example2:Optimalsolutionresults.

MethodsMembershipObjectiveDeviationsfrom αβγ functionsvaluesidealsolutions

Linear(9575,3270)(1593.75,346.25)0.50.5NA [22]Exponential(9563,3252)(1581.75,328.25)0.37750.0NA Hyperbolic(9575,3270)(1593.75,346.25)0.50.5NA Linear(9471,3189)(1489.75,265.25)0.740.610.67

ProposedneutrosophicapproachExponential(9302,3126)(1320.75,202.25)0.8350.560.53

Hyperbolic(9249,3107)(1267.75,183.25)0.880.210.36

NA:NotApplicable.

Example3. [22]Letusconsiderthefollowingnumericalexample.

Min ˜ OIF 1 (x) = 2 i=1 3 j=1 c(1)IF ij xij

Min ˜ OIF 2 (x) = 2 i=1 3 j=1 ˜ c(2)IF ij xij

s.t.

2 i=1 3 j=1 xij ≤ (20, 24, 28;18, 24, 32)

2 i=1 3 j=1 xij ≤ (15, 18, 24;12, 18, 30)

2 i=1 3 j=1 xij ≥ (16, 18, 22;14, 18, 24)

2 i=1 3 j=1 xij ≥ (8, 12, 16;6, 12, 20)

2 i=1 3 j=1 xij ≥ (11, 12, 14;10, 12, 18) xij ≥ 0 ∀ i = 1, 2& j = 1, 2, 3.

TheintuitionisticfuzzyparametersrelatedtoExample3aresummarizedinTable 5.Onsolving eachobjectiveindividually(Example3),wegettheupperandlowerbounds U1 = 428.25, U2 = 531.75, L1 = 378.00and L2 = 484.12forthefirstandsecondobjectivefunctions respectively.ByapplyingtheneutrosophicMOPPs(5.9),(5.13)and(5.18),thesolutionresultsare summarizedinTable 6.Onsolvingthefinalneutrosophicoptimizationmodel,theobtainedobjective valuesaremorepromisingthan[22]methodforallthemembershipfunctions.Theworkingefficiency oftheproposedneutrosophicmethodundervariousmembershipfunctionscanbedepictedas Hyperbolic >Exponential>Linear.Thus,theproposedneutrosophictechniqueisthemostprominent methodcomparedto[22]methodunderallthemembershipfunctions.Moreover,themaximum attaintmentofacceptancedegreeisbetterachievedbyourproposedtechniquethatcanbeapreferable methodover[22]method.

Table5. Example3:Intuitionisticfuzzyparametersfor˜

AIMSMathematics

Volume6,Issue5,4556–4580.

4576
c(1)IF ij and˜ c(2)IF ij . c(1)IF 11 = (4, 6, 8;2, 6, 10)˜ c(1)IF 12 = (5, 7, 9;3, 7, 11)˜ c(1)IF 13 = (6, 8, 10;4, 8, 12) c(1)IF 21 = (7, 9, 11;5, 9, 13)˜ c(1)IF 22 = (12, 14
16;10
c(1)IF
= (10
12
c(2)IF 11 = (3, 6, 9;0, 6, 12)˜ c(2)IF 12 =
10
c(2)IF
=
c(2)IF 21
12, 16;4
12
c
IF
,
, 14, 18)˜
23
,
, 14;8, 12, 16)
(7,
, 13;4, 10, 16)˜
13
(10, 15, 20;5, 15, 25)
= (8,
,
, 20)˜
(2)
22 = (10, 14, 18;6, 14, 20)˜ c(2)IF 23 = (12, 16, 20;8, 16, 24)

Table6. Example3:Optimalsolutionresults.

MethodsMembershipObjectiveDeviationsfrom αβγ functionsvaluesidealsolutions

Linear(454,546)(76,61.88)0.720.278NA [22]Exponential(437,531)(59,46.88)0.61380.0NA Hyperbolic(454,546)(76,61.88)0.93350.5NA Linear(427,526)(49,41.88)0.7430.250.47

ProposedneutrosophicapproachExponential(421,519)(43,34.88)0.7870.160.53 Hyperbolic(418,513)(40,28.88)0.9670.340.41

NA:NotApplicable.

7.Conclusions

Inthispaper,aneffectivemodelingandoptimizationframeworkfortheIFMOLPPispresented underneutrosophicfuzzyuncertainty.Theproposedoptimizationtechniquehasbeendesignedunder theneutrosophicenvironment,consistingofindependentindeterminacydegreeindecision-making processes.Thesatisfactorydegreeofdecision-makerisalwaysmaximizedindifferentformstoachieve thebestpossiblecompromisesolutionofeachobjectivefunction.Wehavealsodevelopedavarietyof membershipfunctionstodeterminethebestpossibleoptimalsolution.Theuseoflinear,exponential, andhyperbolicmembershipfunctionsmoreflexibleandrealisticindecision-makingprocessesdue totheindeterminacydegreewhileobtainingtheoptimalcompromisesolution.Theproposedmodels areimplementedonthreedifferentexistingnumericalexamples,andacomparativestudyisdone. Thesolutionresultsobtainedbyourproposedneutrosophicoptimizationtechniqueoutperformsas comparedtoothersforallthediscussednumericalexamples.Anampleopportunitytoselectthe differentsolutionresultsisalsodepictedbyincorporatingvarioussortsofmembershipfunctionsunder theneutrosophicenvironment.Inthecaseofexponential-typemembershipfunction,asetofdifferent optimalsolutionresultscanbegeneratedbytuningtheshapeparameters.

Severaladvantagescanbeavailedbyapplyingtheproposedapproachefficientlyinbothnumerical problemsaswellasreal-lifeapplications.

• Theproposedneutrosophicoptimizationtechniquesdealwiththeindeterminacydegree,whichis theareaofignoranceofpropositions’valuesbetweentheacceptanceandrejectiondegrees.

• Theproposedmultiobjectivemodelingapproachdealswiththeintuitionisticfuzzyparameters, whichinvolvesmembershipaswellasnon-membershipfunctionsandaremorerealisticas comparedtofuzzyparameters.

• Differentsortsofmembershipfunctionsundertheneutrosophicenvironmentprovidean opportunitytoselectthedesiredmembershipfunctionsaccordingtodecision-makers’choices.

• Theshapeparameter(d)inexponential-typemembershipfunctionsdepictsthedegreeof uncertaintyatadifferentlevelandprovidesmoreflexibilityindecision-makingprocesses.

• Anamplescopeforgeneratingavarietyofcompromisesolutionsetsbytuningtheparametersin indeterminacydegreesisalsoabenchmarkingadvantageoftheproposedapproach.

• Itcanbeeasilyimplementedondifferentreal-lifeproblemssuchastransportation,supplier selection,supplychain,manufacturing,inventorycontrol,assignmentproblems,etc.

Thepropoundedstudyhassomelimitationsthatcanbeaddressedinfutureresearch.Various metaheuristicapproachesmaybeappliedtosolvetheproposedneutrosophicmodelasafuture

AIMSMathematics

Volume6,Issue5,4556–4580.

4577

4578 researchscope.Thediscussedmodelscanbeappliedtovariousreal-lifeapplicationssuchas transportationproblems,supplierselectionproblems,inventorycontrol,portfoliooptimization,etc.

Acknowledgments

AllauthorsareverythankfultotheEditor-in-Chief,anonymousGuestEditor,andpotential reviewersforprovidingin-depthcommentsandsuggestionsthatimprovedthereadabilityandclarity ofthemanuscript.ThefirstauthorisverythankfultoJazanUniversityforprovidingtheonlineaccess ofvariouspaidjournals.

Conflictofinterest

Allauthorsdeclarenoconflictofinterest.

References

1. A.Y.Adhami,F.Ahmad,Interactivepythagorean-hesitantfuzzycomputationalalgorithmfor multiobjectivetransportationproblemunderuncertainty, Int.J.Manage.Sci.Eng.Manage., 15 (2020),288–297.

2. F.Ahmad,A.Y.Adhami,Neutrosophicprogrammingapproachtomultiobjectivenonlinear transportationproblemwithfuzzyparameters, Int.J.Manage.Sci.Eng.Manage., 14 (2019a), 218–229.

3. F.Ahmad,A.Y.Adhami,Totalcostmeasureswithprobabilisticcostfunctionundervaryingsupply anddemandintransportationproblem, OPSEARCH, 56 (2019b),583–602.

4. F.Ahmad,A.Y.Adhami,F.Smarandache,Singlevaluedneutrosophichesitantfuzzy computationalalgorithmformultiobjectivenonlinearoptimizationproblem, NeutrosophicSets Syst., 22 (2018).

5. F.Ahmad,A.Y.Adhami,F.Smarandache,Neutrosophicoptimizationmodelandcomputational algorithmforoptimalshalegaswatermanagementunderuncertainty. Symmetry, 11 (2019).

6. F.Ahmad,A.Y.Adhami,F.Smarandache,15-modifiedneutrosophicfuzzyoptimizationmodelfor optimalclosed-loopsupplychainmanagementunderuncertainty,2020.

F.InSmarandache,M.Abdel-Basset, OptimizationTheoryBasedonNeutrosophicandPlithogenic Sets,AcademicPress,343–403.

7. F.Ahmad,S.Ahmad,M.Zaindin,A.Y.Adhami,.Arobustneutrosophicmodelingand optimizationapproachforintegratedenergy-food-watersecuritynexusmanagementunder uncertainty, Water, 13 (2021),121.

8. F.Ahmad,Supplierselectionproblemwithtype-2fuzzyparameters:Aneutrosophicoptimization approach, Int.J.FuzzySys.,(2021),1–21.

9. A.A.H.Ahmadini,F.Ahmad,Anovelintuitionisticfuzzypreferencerelationsformultiobjective goalprogrammingproblems, J.Intell.FuzzySys.,(2021),1–17.

10. P.P.Angelov,Optimizationinanintuitionisticfuzzyenvironment, FuzzySetsSys., 86 (1997), 299–306.

AIMSMathematics

Volume6,Issue5,4556–4580.

11. K.T.Atanassov,Intuitionisticfuzzysets, FuzzySetsSys., 20 (1986),87–96.

12. R.E.Bellman,L.A.Zadeh,Decision-makinginafuzzyenvironment, Manag.Sci., 17 (1970), B–141–B–164.

13. S.K.Bharati,S.R.Singh,Abhishekh,Acomputationalalgorithmforthesolutionoffullyfuzzy multi-objectivelinearprogrammingproblem, Int.J.Dyn.Control, 6 (2018),1384–1391.

14. C.T.Chang,Agoalprogrammingapproachforfuzzymultiobjectivefractionalprogramming problems, Int.J.Sys.Sci., 40 (2009),867–874.

15. E.Dolan,Theneosserver4.0administrativeguide.Tech.Technicalreport,Memorandum ANL/MCS-TM-250,MathematicsandComputerScienceDivision,ArgonneNationalLaboratory, Argonne,IL,USA,2001.

16. J.Dong,S.Wan,Anewmethodforsolvingfuzzymulti-objectivelinearprogrammingproblems, IranianJ.FuzzySyst., 16 (2019),145–159.

17. A.Ebrahimnejad,Aneffectivecomputationalattemptforsolvingfullyfuzzylinearprogramming usingmolpproblem, J.Ind.ProductionEng., 36 (2019),59–69.

18. A.Ebrahimnejad,J.L.Verdegay,Anewapproachforsolvingfullyintuitionisticfuzzy transportationproblems, FuzzyOptimizationDecisionMaking, 17 (2018),447–474.

19. A.Gupta,A.Kumar,(2012).Anewmethodforsolvinglinearmulti-objectivetransportation problemswithfuzzyparameters, Appl.Math.Model., 36 (2012),1421–1430.

20. S.Li,C.Hu,Satisfyingoptimizationmethodbasedongoalprogrammingforfuzzymultiple objectiveoptimizationproblem, Eur.J.Oper.Res., 197 (2009),675–684.

21. B.Liu,X.Chen,Uncertainmultiobjectiveprogramminganduncertaingoalprogramming, J. UncertaintyAnal.Appl., 3 (2015),10.

22. S.Mahajan,S.Gupta,Onfullyintuitionisticfuzzymultiobjectivetransportationproblemsusing differentmembershipfunctions, Ann.Oper.Res.,(2019),1–31.

23. A.K.Nishad,S.R.Singh,Solvingmulti-objectivedecisionmakingprobleminintuitionisticfuzzy environment, Int.J.Syst.AssuranceEng.Manage., 6 (2015),206–215.

24. D.Rani,T.Gulati,H.Garg,Multi-objectivenon-linearprogrammingprobleminintuitionistic fuzzyenvironment:Optimisticandpessimisticviewpoint, Expertsystemswithapplications, 64 (2016),228–238.

25. J.Razmi,E.Jafarian,S.H.Amin,Anintuitionisticfuzzygoalprogrammingapproachforfinding pareto-optimalsolutionstomulti-objectiveprogrammingproblems, ExpertSyst.Appl., 65 (2016), 181–193.

26. R.M.Rizk-Allah,A.E.Hassanien,M.Elhoseny,Amulti-objectivetransportationmodelunder neutrosophicenvironment, Comput.Electr.Eng., 0 (2018),1–15.

27. N.Server,State-of-the-ArtSolversforNumericalOptimization,2016.

28. P.Singh,S.Kumari,P.Singh,(2017).Fuzzyefficientinteractivegoalprogrammingapproachfor multi-objectivetransportationproblems, Int.J.Appl.Comput.Math., 3 (2017),505–525.

29. S.K.Singh,M.Goh,Multi-objectivemixedintegerprogrammingandanapplicationina pharmaceuticalsupplychain, Int.J.Prod.Res., 57 (2019),1214–1237.

AIMSMathematics

Volume6,Issue5,4556–4580.

4579

30. S.K.Singh,S.P.Yadav,Intuitionisticfuzzymulti-objectivelinearprogrammingproblemwith variousmembershipfunctions, Ann.OperationsRes., 269 (2018a),693–707.

31. V.Singh,S.P.Yadav,Modelingandoptimizationofmulti-objectiveprogrammingproblemsin intuitionisticfuzzyenvironment:Optimistic,pessimisticandmixedapproaches, ExpertSyst.Appl., 102 (2018b),143–157.

32. F.Smarandache,Aunifyingfieldinlogics:Neutrosophiclogic.In Philosophy,AmericanResearch Press,(1999),1–141.

33. A.M.Tarabia,M.A.Kassem,N.M.El-Badry,Amodifiedapproachforsolvingafuzzymultiobjectiveprogrammingproblem,In AppliedInformatics, 4 (2017),SpringerOpen.

34. S.L.Tilahun,Feasibilityreductionapproachforhierarchicaldecisionmakingwithmultiple objectives, OperationsResearchPerspectives, 6 (2019),100093.

35. L.Zadeh,Fuzzysets, Inf.Control, 8 (1965),338–353.

36. M.Zangiabadi,H.R.Maleki,Fuzzygoalprogrammingtechniquetosolvemultiobjective transportationproblemswithsomenon-linearmembershipfunctions,2013.

37. S.Zeng,Y.Hu,T.Balezentis,D.Streimikiene,Amulti-criteriasustainablesupplierselection frameworkbasedonneutrosophicfuzzydataandentropyweighting, Sustain.Dev., 28 (2020), 1431–1440.

38. M.Zheng,Y.Yi,Z.Wang,T.Liao,Efficientsolutionconceptsandtheirapplicationinuncertain multiobjectiveprogramming, Appl.SoftComput., 56 (2017),557–569.

39. H.J.Zimmermann,Fuzzyprogrammingandlinearprogrammingwithseveralobjectivefunctions, FuzzySetsSyst., 1 (1978),45–55.

© 2021theAuthor(s),licenseeAIMSPress.This isanopenaccessarticledistributedunderthe termsoftheCreativeCommonsAttributionLicense (http://creativecommons.org/licenses/by/4.0)

AIMSMathematics

Volume6,Issue5,4556–4580.

4580
View publication stats

Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.