Implementation of neural network sliding-mode controller for DD robot

Abstract

The experimental development of a trajectory tracking neural network controller based on the theory of continuous sliding-mode controllers is shown in the paper. The neural network control law was verified on a real direct drive 3 DOF PUMA mechanism. The new neural network sliding-mode controller was successfully tested for trajectory tracking sudden changes in the manipulator dynamics (load). The comparision between the neural network sliding mode controller, a computer torque method controller and a continuous sliding mode controller with PI-estimator for sudden load changes on the real robot mechanism is shown.