A New Method of Intelligent Controller Design and its Application in Ship Course Control

摘要

Based on the Reference Trajectory Setting (RST) and Noise Adding (NA) technologies, a new intelligent controller design method was proposed. According to the method, there are four steps in controller design process. Firstly, a certain precision model of the plant and the disturbances should be acquired. Secondly, RST was used to plan the control objective. Thirdly, the Fuzzy-Neural Network (FNN) was chosen as the controller structure to be designed. Fourthly, in the dynamic and uncertainty environment, the FNN''s parameters were searched and optimized automatically by using one certain evolutionary algorithm. At last, the new method was used to design the ship nonlinear course system''s controller, and the simulations of course track and keep control were done in the classical ocean environment. The results show that the controller designed has good robustness performance, and the system has fast transition process and high dynamic and stable precision. And the new method''s feasibility and validity are proved successfully.