Asymptotic tracking by a reinforcement learning-based adaptive critic controller

(整期优先)网络出版时间:2011-03-13
/ 1
Adaptivecritic(AC)basedcontrollersaretypicallydiscreteand/oryieldauniformlyultimatelyboundedstabilityresultbecauseofthepresenceofdisturbancesandunknownapproximationerrors.Acontinuous-timeACcontrollerisdevelopedthatyieldsasymptotictrackingofaclassofuncertainnonlinearsystemswithboundeddisturbances.TheproposedAC-basedcontrollerconsistsoftwoneuralnetworks(NNs)-anactionNN,alsocalledtheactor,whichapproximatestheplantdynamicsandgeneratesappropriatecon...