Podporin S. Development of methods of intellectual ship course steering

The thesis addresses the issue of raising the efficiency of ship course steering. Intelligent control techniques such as genetic algorithms, fuzzy logic, and artificial neural networks are used to deal with the problem. Genetic methods are proposed for optimization of tuning procedure for the ship's autopilot control algorithm. A new control algorithm for course steering based on fuzzy inference machine with nonlinear membership functions is proposed and studied. A new technique of finding near optimal parameters of ship steering algorithm with help of artificial neural network is proposed and studied. An alternative adaptive method of ship course steering based on neural control techniques is proposed and studied. The latter was shown to be able to compensate for changing ship's dynamics and function as a self-tuning adaptive algorithm. Imitative modeling of real refrigerating ship's behavior was undertaken in order to assess and compare performance of proposed intelligent control techniques. The results obtained allowed to make conclusion that such techniques are fully applicable for the task and generally may perform better than traditional PID-algorithm.