Adaptive neural network control of coordinated manipulators

In this article, adaptive neural network control of coordinated manipulator s is considered in an effort to eliminate the time-consuming and error pron e dynamic modeling process which is necessary for the implementation of con ventional adaptive control. After a concise dynamic model in the object coo rdinate space is developed for the coordinated manipulators, an adaptive ne ural network controller is presented by combining the techniques of neural network parameterization, adaptive control, and sliding mode control. It ca n be shown that the motion tracking errors converge to zero asymptotically whereas the internal force tracking error remains bounded and can be made a rbitrarily small. Numerical simulations are conducted to show the effective ness of the proposed method, (C) 1999 John Wiley & Sons, Inc.