Design of GA-based Fuzzy Control for Ship Steering

A new hybrid control algorithm of autopilot for ship steering is proposed. Fuzzy control technology emulates the characteristics of ship's manual steering. And genetic algorithms are used to optimize the parameters of the fuzzy ship autopilot for their complicate coupling, and then realize the auto tradeoff among them. Since rule base is formulated in fuzzy inference, the optimization for fuzzy rule base can be achieved by optimizing fewer parameters. Simulation study shows that such an autopilot can smartly and steadily follow the reference course, and has small overshoot and good robustness.