Fuzzy RBF Neural Network Control for Unmanned Surface Vehicle

Abstract

For the problem of the heading control of USV, the intelligent control method is achieved by fuzzy RBF neural network. Considering the uncertainty of the USV motion system, the fuzzy system with universal approximation performance is used to fuzzily approximate the uncertainties and external disturbances in the USV motion model. To further enhance the fuzzy system approximation, fuzzy rules were optimized online by RBF neural network with fast learning ability. The intelligent control method proposed realizes continuous and stable tracking of USV heading through simulation.