基于规则法的水下避碰专家系统

针对复杂环境中船舶实现自动避碰问题，设计了一种基于规则法的智能型专家系统，即水下避碰专家系统．该系统具有多元分层知识库体系结构，采用框架式知识表示方式，实现了水下避碰知识的表示、存储，避免了传统专家系统的“窄台阶”问题．应用模糊理论，分别建立了针对水下静止目标和运动目标的碰撞危险度模型，实现了定性推理与定量计算的结合．给出了水下会遇态势分类模型，总结了水下常见的多种会遇态势．结合所建立的系统模型给出水下多目标避碰仿真示例．仿真实验结果表明，系统能对各种水下避碰情况及时给出正确的避让方案．

A collision avoidance expert system with rule knowledge for underwater vehicle

Aiming at the problem of automatic collision avoidance system for ships,an intelligent expert system,the collision avoidance expert intelligent system for underwater vehicle(CAESUV),is developed.The system has a multi-unit and layering knowledge base systematic structure and a frame class knowledge representation to represent and store all kinds of knowledge for underwater collision avoidance.It can overcome some shortcomings owned by traditional expert system.The system builds a set of models of judging collision risk for static target and moving target by conjoining qualitative analysis and quantitative analysis.The models of classifying encounter situation and all kinds of encounter situations underwater are given.For multi-objects encounter situation, the system puts forward a tactic to choose optimal collision avoidance scheme during the simulation test.The experiments show that the system can analyze and judge various conditions,and offer a reasonable scheme,which can settle the collision problem effectively.