Adaptive Neural Fuzzy Control for Robot Manipulator Friction and Disturbance Compensator

This paper presents an application of an adaptive neuro-fuzzy controller for compensating friction and disturbance effects on robot manipulators. The frictions and disturbances are important parts of the dynamic system of a robot manipulator. However, they are highly nonlinear and not easily modeled. Feedback linearization control combined with PD type fuzzy control and adaptive neural network control are proposed to control robot manipulators with unmodeled frictions. The feedback linearization control is designed to control the trajectory of the robot manipulator while the PD type fuzzy control is added as a parallel controller to control frictions and disturbances. This fuzzy control ensures that good tracking control is maintained even if some modeling error, disturbance, noises exist. The neural network can also be trained with experimental data for the frictions and disturbances. Simulations show that the joint positions are well controlled under wide variation of operation conditions and existences of uncertainties.