Abstract
This paper develops a chattering-free neuro-wavelet sliding mode control (NWSC) strategy to resolve the ship trajectory tracking problem. The proposed control strategy consists of a neuro-wavelet controller and a smooth compensator. The neuro-wavelet controller using a wavelet neural network (WNN) is the principal tracking controller based on sliding mode control (SMC) method; and the parameters of WNN are on-line tuned by the derived adaptation laws in the Lyapunov stability sense. The smooth compensator is designed to recover the approximation error, so the robust tracking performance can be guaranteed. A comparison between conventional SMC and the proposed chattering-free NWSC is made in the simulation. Moreover, the superiority of the NWSC system is indicated in comparison with the different learning rates and attenuation level. From the simulation results, it is shown that the proposed NWSC system not only achieves satisfactory control performance, but also eliminates the chattering phenomena of the control effort.