Sensing and Direction in Locomotion Learning with a Random
Karl Hedman, David Persson, Per Skoglund, Dan Wiklund, Krister Wolfi, Peter C. Nordin · 1998
KarlHedman DavidPersson PerSkoglund DanWiklund KristerWolfi PeterNordinComplexSystemsGroup,DepartmentofPhysicalResourceTheoryChalmersUniversityofTechnologySE-41296G˜oteborg,SwedenE-mail: fwolfi,[email protected]flrstinstanceinsensinganddirectionwithalearningRandomMorphologyrobot. UsingGP,itlearnstolocomoteitselfindifierentdirectionsandbylettingdifierentsolutionsmastertherobotindifierentsituationsitcanthusfollowanarbitrarypath.The Random Morphology robot [Dittrich et al,1998] is composed of seven standard ofi-the-shelfR/C servo motors that are interconnected arbitrar-ily in a two dimensional plane. The robot alsohas a proximity sensor onboard. To control theservo motors and process sensory data, we usethe EyeBot MK3