改進的免疫遺傳算法在基于神經網(wǎng)絡的多機器人協(xié)作搬運中的應用.pdf

1、中南大學碩士學位論文改進的免疫遺傳算法在基于神經網(wǎng)絡的多機器人協(xié)作搬運中的應用姓名：張穎申請學位級別：碩士專業(yè)：控制科學與工程指導教師：譚冠政20080528ABSTRACTThemechanismofthemultiagentsystem’ScontrolhasbeenaninterestingresearchprojectinthefieldsofroboticsandartificialintelligenceItincorpor

2、atesmanyimportantsubjects，suchasintelligentcontrol，artificiallife，evolutionarycomputationrobotprogrammingtechnologyetcWththedevelopmentofresearchonmulti—agentsystem，ithasbecomeatrendtoconducttheoptimizationdesignandperfo

3、rmancesimulationofmultiagentsystemresearchbasedontheartificial1ifeaswellasevolutionaryalgorithmsInthispaperwehavedevelopedaninnovativecontrolstrategybasedontheevolvableartificialneuralnetworkandtheartificialimmunealgorit

4、hmwithelitismforthecooperationbehavesoftheMultiRobotSystemsUsingthissystemweexecutedperformancesimulationandparametricdesignofthecooperativetransportbehaviorforMulti—RobotSystemswithsingleprey碭einnovativeImmune—GeneticAl

5、gorithmwithelitistgivesthenewdefinitionsofparametersforalgorithmnissystemusedaBackPropagationNeuralNetworkwiththreelayerstoprocesstheperceptiveinformationofeveryrobotandmadeadecisionforitsactionTheIGAEWasusedtooptimizeth

6、econnectionweightvaluesofthisneu／alnetworkwhichmadetheperformanceoftheneuralnetworktobeevolvedcontinuouslyandfinallyabehaviordecisionmakingsystemwithgoodperformancecouldbeobtainedTheresultsofsimulationexperimentsindicate

7、thatthemobilerobotscankeepstabletransportofthepreywhichprovesthegooddecisionmakingabilityoftheIGAEANNbehaviordecisionmakingsystemduringthetransportKEYWORDS：‘Multi—RobotSystems；ImmuneGeneticAlgorithmwithElitist(IGAE)；Arti