The application of fuzzy neural network in ship course control system

When ship docks at harbor or other sea area, both the robustness and dynamic properties of course control system are required strictly. Because the model parameters are relative to the speed and load of the ship, it is difficult to design a good controller based on the ship model. This paper combines internal model control with fuzzy neural network to design a course system. First, it designs a zero steady-state error controller which ensures robustness of the course control system relay on inner model. For the zero steady-state error controller, it is necessary to reduce the dynamic properties of the control system to ensure the robustness. Thus, this paper uses fuzzy neural network to adjust the pole sites of the closed control system based on the ship course and course change rate. It designs a new structure of fuzzy neural network which uses neural network to represent the fuzzy rules. According to expert experiences, it also gives out the weights computation method of fuzzy neural adjuster. At last, the hybrid course control system is applied in a actual ship and the course response curves indicate that the course control system possess good robustness and dynamic properties.