基于知识库的雷达辐射源识别专家系统

从分析雷达辐射源识别的特点出发，提出了基于知识库的雷达辐射源识别专家系统。讨论了系统总体结构模型和面向对象模型。根据系统要求建立了数据库；积累案例形成案例库；获取专家经验及领域知识建立了规则库；依据知识层次结构建立了模型库，这四库一体化形成专家系统。介绍了基于案例、基于规则和基于模型的推理方法，比较了各种推理方法的优、缺点，并提出了基于案例、规则和模型的混合推理。最后给出了系统的功能和实现方法。应用证明基于知识库的雷达辐射源识别专家系统可以得到与专家同样的结果。

Expert System of Radar Radiating-source Recognizing based on Knowledge Library

An expert system of radar radiation-source recognizing based on knowledge library is proposed upon the analysis of the characteristics of radar radiating-source recognizing. The system structure and object-oriented model are discussed in this paper. Data bases (DB) are set up according to system demand, case bases (CB) are established on accumulated cases, rule bases (RB) are established after accepting expert experience, knowledge of this field, and model bases (MB) built according to the structure of knowledge. The four bases are combined in the expert system. The reasoning method based on case, rule and model are described. Comparing with each other, the advantage and disadvantage of these reasoning methods are obtained. The mixed reasoning based on case, rule and model is sussested to overcome their shortcomings. The functions of system and way of realization are described in the paper. It is proved by usage that the expert system of radar radiation-source recognizing based on knowledge base can achieve the same reasoning result as the expert.