摘要
This paper researches identification and learning from adaptive neural network(NN) control of uncertain rigid-link electrically-driven(RLED) robot manipulators. By using input-to-state stability(ISS) adaptive NN controller and the small gain theorem, the closed-loop system is divided into two connecting subsystems, which proved to be exponential stability under the persistent excitation(PE) condition and the probable controller singularity problem is prevented. Meanwhile, accurate NN identification and learning control of the unknown system dynamics is realized along recurrent orbits in the control procedure.The learned knowledge that preserved as a series of constant neural weights can be recycled to reach stability and improve performance due to preventing the huge repeated training process of NNs. Simulation studies have shown the effectiveness of the presented control method.
This paper researches identification and learning from adaptive neural network(NN) control of uncertain rigid-link electrically-driven(RLED) robot manipulators. By using input-to-state stability(ISS) adaptive NN controller and the small gain theorem, the closed-loop system is divided into two connecting subsystems, which proved to be exponential stability under the persistent excitation(PE) condition and the probable controller singularity problem is prevented. Meanwhile, accurate NN identification and learning control of the unknown system dynamics is realized along recurrent orbits in the control procedure.The learned knowledge that preserved as a series of constant neural weights can be recycled to reach stability and improve performance due to preventing the huge repeated training process of NNs. Simulation studies have shown the effectiveness of the presented control method.
引文
[1]C.Y.Su,Y.Stepanenko,Redesign of hybrid adaptive/robust motion control of rigid-link electrically-driven robot manipulator,IEEE Trans.Robotics and Automation,14(4):651-655,1998.
[2]R.J.Wai.Tracking control based on neural network strategy for robot manipulators.Neuro computing,51:425-445,2003.
[3]B.J.Driessen,Adaptive global tracking for robots with unknown link and actuator dynamics,Int.J.Adapt.Control Signal Process,20:123-138,2006.
[4]S.N.Huang,K.K.Tan,T.H.Lee,Neural network control design for rigid-link electrically-driven robot,IMech E J.Syst.Control Eng,217:99-107,2003.
[5]C.M.Kwan,F.L.Lewis,D.M.Dawson,Robust neural-network control of rigid-link electrically-driven robots,IEEE Trans.Neural Network,9:581-588,1998.
[6]R.J.Wai,P.C.Chen,Robust neural-fuzzy-network control for robot manipulator including actuator dynamics,IEEE Trans.Ind.Electron.,53(4):1328-1349,2006.
[7]Y.C.Chang,H.M.Yen,M.F.Wu,An intelligent robust tracking control for electrically-driven robot system,Int.J.System Sci.,39:497-511,2008.
[8]C.Wang,D.J.Hill,Learning form neural control,IEEE Trans.Neural Netw.,17(1):130-146,2006.
[9]C.Wang,D.J.Hill,Deterministic Learning Theory for Identification.Recognition and Control,CRC Press,Boca Raton,FL,2009.
[10]T.F.Liu,C.Wang,D.J.Hill,Learning from neural control of nonlinear systems in normal form,Systems and Control Letters,58:633-638,2009.
[11]Y.X.Wu,M.Yang,C.Wang,Learning from output feedback adaptive neural control of robot,Control and Decision,27(11):1740-1744,2012.
[12]C.Wang,D.J.Hill,S.S.Ge,and G.R.Chen,An ISS-modular approach for adaptive neural control of pure-feedback systems,Automatica,42(5):723-731,2006.
[13]A.J.Kurdila,F.J.Narcowich,J.D.Ward,Persistency of excitation in identification using radial basis function approximants,ISIAMJ.Control Optim.,33(2):625-642,1995.
[14]E.D.Sontag,Y.Wang,New characterizations of input-to-state stability,IEEE Trans.Autom.Control,41(9):1283-1294,1996.
[15]S.Sastry,M.Bodson,Adaptive Control:Stability,Convergence,and Robustness,Prentice-Hall.Englewood Cliffs,NJ,1989.