Generalized predictive control based on self-recurrent wavelet neural network for stable path tracking of mobile robots: adaptive learning rates approach

Abstract

In this paper, a generalized predictive control (GPC) method based on self-recurrent wavelet neural network (SRWNN) is proposed for stable path tracking of mobile robots. Since the SRWNN has a self-recurrent mother wavelet layer, it can well attract the complex nonlinear system although the SRWNN has less mother wavelet nodes than the wavelet neural network. Thus, the SRWNN is used as a model identifier for approximating on-line the states of the mobile robot. In our control system, since the control inputs, as well as the parameters of the SRWNN identifier are trained by the gradient descent method with the adaptive learning rates (ALRs), the optimal learning rates which are suitable for the time-varying trajectory of the mobile robot can be found rapidly. The ALRs for training the parameters of the SRWNN identifier and those for learning the control inputs are derived from the discrete Lyapunov stability theorem, which are used to guarantee the convergence of the GPC system. Finally, simulation results are provided to demonstrate the effectiveness of the proposed control strategy.