A Fuzzy PID Autopilot for Ship Steering Based on Membership Function Optimized by an Improved Genetic Algorithm
Chengming Yang and Zaojian Zou
It is perceived that both the fuzzy rules and choice of membership functions greatly contribute to the control effect of fuzzy PID autopilot for ship steering. However, the parameters of membership functions are usually set subjectively. In this paper, an innovative framework for membership functions optimized in view of fuzzy linguistic variables is investigated. In particular, this scheme is explicitly proposed based on an improved genetic algorithm (GA) under the condition that fuzzy rules, quantization factor and scaling factor are determined, so that it could be more reasonable. Based on this scheme, simulation experiment is carried out using the Nomoto model for the ship “YULONG”. The results validate the performance of the proposed scheme in terms of course angle and rudder angle.
Keywords: Ship steering; PID; fuzzy control; membership function; genetic algorithm