INVERSE KINEMATIC NEURO-CONTROL OF ROBOTIC SYSTEMS

The emergence of the theory of dynamic neural computing has made it po ssible to develop neural learning and adaptive schemes that can be use d to obtain feasible solutions to complex control problems. such as in verse kinematic control for robotic systems. In this paper, such a neu ral learning scheme using a multilayered dynamic neural network (MDNN) is proposed. The basic dynamic computing element of the MDNN is a dyn amic neural unit (DNU), developed in this paper. The learning and adap tive capabilities of the dynamic neural unit can be used for developin g complex dynamic structures. In this paper, the DNU has been used for developing a multilayered dynamic neural network for the inverse kine matic control of a two-linked robot. The validity of the proposed sche me is demonstrated through computer simulation studies. (C) 1998 Elsev ier Science Ltd. All rights reserved.