资源描述：
《adaptive neuro-fuzzy controler with genetic training for mobile robot control》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。
1、Int.J.ofComputers,Communications&Control,ISSN1841-9836,E-ISSN1841-9844Vol.VII(2012),No.1(March),pp.135-146AdaptiveNeuro-FuzzyControlerWithGeneticTrainingForMobileRobotControlO.Obe,I.DumitracheOlumideObe"Politehnica"UniversityofBucharest,RomaniaE-mail:oluobes@gmail.comIoan
2、Dumitrache"Politehnica"UniversityofBucharest,RomaniaE-mail:ioan.dumitrache@cncsis.ro,idumitrache@ics.pub.roAbstract:Inthispaper,weinvestigatetheuseofadaptivetechniquesintheoptimiza-tionofnavigationofKheperamobilerobotinanunstructuredanddynamicenvironment.Weoptimizetheperf
3、ormanceofoursimplifiedfuzzycontrollerusingneuralnetworkthatutilizesgeneticalgorithmlearning.Theadaptationofthesysteminvolvesthetuningofthecontrolrulestherebytrimmingthecontrolactions,andadjustingthefuzzycontrolleroutputgain.Werealisedanimprovedperformanceinouradaptiveneuro
4、-fuzzycontrollerwithgenetictrainingforvariousimplementedbehavioursontherobot.Keywords:Khepera,fuzzycontroller,neuro-fuzzycontroller,navigation,ge-neticalgorithm.1IntroductionNavigationandobstacleavoidanceareveryimportantissuesforthesuccessfuluseofanau-tonomousmobileroboti
5、nadynamicandunstructuredenvironment.Mobilerobotresearchersaimtobuildanautonomousandintelligentrobotwhichcanplanitsmotioninadynamicen-vironment.Asuccessfuluseofanautonomousmobilerobotdependsonitscontroller.Mobilerobotcontrolisdifficultastheyaresubjectedtonon-holonomic(non-in
6、tegrable)kinematiccon-straintsinvolvingthetimederivatesofconfigurationvariables[12]anddynamicconstraints.Bothanalyticallikepotentialfieldmethodaswellasgraph-basedtechniqueshavebeenusedtosolvethenavigationproblemsofrobotinvolvingbothstaticanddynamicobstacles.But,allsuchmetho
7、dsmaynotbesuitableforon-lineimplementationsduetotheirinherentcomputationalcomplexityandlimitations.Mobilerobotresearchershavecarriedoutvariousresearchesinthisdirectionusingvariousintelligenttechniquesmethodssuchasfuzzylogic,neuralnetworkandgeneticalgorithmandtheirdifferent
8、hybrids.Becauseofthenon-linearkinematicsoftherobot,theuncertaintyinsensorsreadings,andunstructur
