Ship course steering predictive control based on RBF neural network

Abstract

Because the ship steering control is uncertain, nonlinear and time-varying. A predictive control algorithm based on RBF neural network is adopted to the ship steering control. Recursive k-means clustering algorithm and recursive least squares algorithm are used to adjust the RBF neural network. And clonal selection algorithm is used in predictive control algorithm to ensure the global optimal solution. The simulation results show that the predictive control algorithm based on RBF neural network possesses good control performance and strong robustness.