LEARNNING-BASED CONTROL OF ROBOT TRAJECTORY

A learnning control method based on the principle of PID regulator for robot repeated trajectory control ispresented in this paper.The stability criterion of the learnning control method for linear time-varying systemis established and an iterative learnning method of the nonlinear robot system is deduced in this paper.Theproposed learnning scheme is structurally simple and can be used in Real-Time control.Finally,simulationson the PUMA 560 robot trajectory control show that the actual motion trajectories of the manipulator con-verge to the desired ones as the learnning number increases.