![]() | Only 14 pages are availabe for public view |
Abstract The theory of fuzzy logic provides a mathematical framework to capture the , .uncertainties associated with human cognitiye processes, such as thinking and reasoning. On the other hand, artificial neural networks replicate, on a small scale, some of the computational operations observed in biological learning and adaptation. The intent of this thesis is to use a fuzzy-neural approach to extract the fuzzy rules to control the direction and steering of a ship with inertia. Th~ rules are extracted directly from operator-generated actions. utilizing the learning capabilities of the neural , . network. The structure of the network represents the rules. The use of the trained networks to build a fuzzy controller for the steering of the ship is discussed. The controller is required to avoid collision with another object and then steer the ship to reach a certain destination. Simulation results for the motion of the controlled ship show improvements over existing systems in terms of the reduction in the number of obtained rules. The adapted rule acquisition procedure is simple and automatic, hence. it decreases the dependency on human expertise to a very small level. |