Robust PID sliding mode control of robot manipulators with online learning neural network

The paper presents a new adaptive control algorithm for robot motion tracking problem to overcome noises and large uncertainties using integral sliding surface with a neural network. The control quality has been improved compared to conventional PD type sliding surface. It removes chattering and increases the accuracy. The weights of the neural network are updated continuously online granting the approximation of uncertain nonlinearities in the robot dynamics. The stability of the overall system has been proved by Lyapunov direct method. Computer simulations are given to illustrate the robustness and applicability of the proposed method. © 2009 EUCA.