Model-Free adaptive neural fuzzy feed forward torque control for nonlinear parallel mechanism

Many modern and intelligent control methods had been developed for nonlinear systems in order to get better motion accuracy and dynamic performance for parallel robot. This paper aims to propose a nouvelle model-free adaptive neural fuzzy feed forward torque control for parallel mechanism. The advantage of this kind model-free control is that it uses the information directly from the nonlinear dynamics, without knowing the robot physical parameters and complex models. The neural fuzzy inference system for the model-free adaptive neural fuzzy feed forward control is learning from the robot dynamical data base (joint angular displacement, velocity, acceleration and torque) generated from a PID control system. It is believed that the model-free control is simple, flexible and robust. Results from numerical simulation on a 4-bar planar parallel mechanism show the effectiveness and satisfactory of the proposed control.