Adaptive friction observer and sliding mode controller development with RFNN for nonlinear friction compensation

In this paper, the adaptive friction compensation schemes are developed to provide much enhanced position tracking performance against nonlinear dynamic friction. The adaptive friction parameter observer possessing a simple structure and to be easy to implementation into controller is first studied to estimate the friction parameters. The process of the uncertainty approximation using the RFNN technique is considered to enhance the positioning performance. Suppressing additional unknown friction uncertainty by the RFNN, the favorable position tracking result can be achieved via some simulation and experiment to the rotary servo mechanical system.