Expert System For Multi-Ships Collision Avoidance Based AIS

This thesis universalize the application of computer and the development and the application of the technology of Expert System and Artificial Intelligent and the advantage of AIS to avoid the collision accidents of ships efficiently at sea, this thesis designed and implemented a kind of Expert System for Collision Avoidance based AIS. The main research work include several parts, which will be represented as below:(l)By analyzing the advantage of AIS data from that of ARPA, based on the collection and study of research essays published domestically and internationally and the documents of IMO on the AIS technology. This thesis advanced the assume that AIS can be exerted in intelligent navigation.(2)According to the navigation rules' understanding and analyzing, and by the navigation experience and navigation samples' collecting and trimming, we put forth and build a tree layering knowledge base systematic structure, and implement the knowledge base management. According to features of different knowledge, I build many kinds of sub-KB, such as collision avoidance rules' KB, assistance collision avoidance KB and navigation experience KB etc. I adopt multifarious knowledge representation, such as: frame knowledge representation, and production knowledge representation.(3)1 creat a new model for Expert System for Collision Avoidance, also build a multi-inference system, which based on forward chain and backward chain, and analog inference, and nonmonotonic inference and uncertain inference. At the same time, we develop each reasoning algorithm.(4)The design and implementation of Expert System for Collision Avoidance Based AIS takes advantage of AIS sufficiently, and it is mainly based on Expert System. It is a kind of Expert System for collision with intelligence. This system can analyze the encounter of two ships at sea, and it can offer a reasonable precept to avoid collision, which settle the collision problem effectively and ensure the navigation safety.