Neural network-based adaptive sliding mode control method for tracking of a nonholonomic wheeled mobile robot with unknown wheel slips, model uncertainties, and unknown bounded external disturbances

Abstract

This article represents a neural network-based adaptive sliding mode control (ASMC) method for tracking of a nonholonomic wheeled mobile robot (WMR) subject to unknown wheel slips, model uncertainties and unknown bounded disturbances. Self-recurrent wavelet neural networks (SRWNN) are employed in order to approximate unknown nonlinear functions due to the unknown wheel slips, model uncertainties, and unknown bounded external disturbances. By doing this, their harmful effects are compensated effectively. Thanks to this control method, a desired tracking performance of the closed-loop control system is achieved where the position tracking errors converge to an arbitrarily small neighborhood of the origin, regardless their initial values. According to Lyapunov theory and LaSalle extension, the stability of the whole closed-loop system is guaranteed to achieve the desired tracking performance. It is unnecessary to preliminarily offline train the neural network weights since they are easily initiated. Online tuning algorithms are established and used, for training the weights. Computer simulations are implemented to demonstrate the validity and efficiency of this proposed control method.