基于免疫机理的足球机器人协作控制研究-模式识别与智能系统专业论文.docx

该算法能够适用于动态复杂的对抗型比赛环境中足球机器人的协作控制，改善多机器人冲突的混乱局面，增强足球机器人策略行为的有效性，增加机器人在比赛过程中的协作进攻次数，提高足球机器人系统的智能协作性。实现了论文的设计目标，使我校足球机器人团队在近年的全国大赛上取得了优异的成绩。关键词：多机器人协作；人工免疫算法；足球机器人；体系结构研究类型：应用研究Subject:SoccerRobotCooperativeControlResearchBasedonImmuneMechanismSpecialty:PatternRecognitionandIntelligentSystemName:KangQian(Signature)Instructor:HouYuanbin(Signature)ABSTRACTRobotcooperativecontrolisoneofthecoreissuesofmulti-robotsystemsresearch,whichrelatestomanyaspectssuchastherobotarchitecture,perception,communication,algorithms,collaborativemechanismsandsoon.Asatypicalmulti-robotsystem,thesoccerrobotsystemprovidesagoodexperimentalplatformfortheintelligencetheorystudiesandavarietyofintegratedapplication.ThesubjectisbasedonthesoccerrobotsystemresearchplatformofXianUniversityofScienceandTechnology.Aimingatthedynamicenvironmentofsoccerrobotcompetition,therobottrackingaccuracyisincreasedbyimprovingthetargettrackingalgorithms,andtheautonomouscollaborationofthesoccerrobotisadvancedthroughusingimmunemechanismandartificialimmunealgorithm,tosolvethecongestionandresourcewasteproblemswhichoftenhappensintheactualcourseofthegame,andimprovetheintelligenceofthesystemfinally.Thesoccerrobotsystem,hardwarestructurecompositionandworkingprincipleareintroducedandanalyzedinthepaper.Focusingontheproblemsofmobilerobottargettracking,thegaussiansumparticlefilter(GSPF)improvedalgorithmisprovidedforsoccerrobottargettracking,parallelextendedkalmanfilteringisreplacedbyparallelgaussianparticlefilterofgaussiansumfilteringintheGSPF,andfullconsiderationisgiventotheinfluenceofthecurrentobservedquantitytothestatedistribution,whichimprovesthetargettrackingprecisionofsystem.Thesimulationresultsshowthatthisalgorithmcaneffectivelyapplytosoccerrobottargettrackinginthedynamicandcomplexenvironment.Accordingtotherobotarchitecture,thebasisofthecollaborativebehaviorforrobotsoccer,consideringtheincentiveandtheinhibitionbetweenantigenanddifferentantibody,andtheimmunemechanismsuchasthecommunicationbetweendifferentimmunecellswitheachothe