RBF-Network-Based Predictive Ship Course Control

The control of the ship's course determines the correctness of navigation, ensures the correct destination, and plays a significant role in the navigation of the ship. Based on RBF neural network and ship mathematical model, a predictive controller using Euclidean distance to calculate center vector for network learning is designed to predict and adjust the three parameters of PID in real time and online, so as to obtain the optimal PID parameters faster. By supervised learning ensuring the course to meet the requirements and the stable navigation of the ship. Then the simulation experiment is carried out to verify the accuracy of the designed controller in Simulink environment of MATLAB software. Compared with the traditional PID controller, this method of this paper reduces the overshoot of the system and the setting time, and improves the accuracy of the ship's course automatic control simultaneously. In addition, the simulation results of ship course control simulation demonstrate the better adaptability, robustness and anti-interference ability of the intelligent PID control strategy.