大型机电设备分布式故障诊断专家系统的求解策略

A Problem-Solving Strategy for Distributed Fault Diagnostic Expert System on a Large Mechanic-Electric Equipment

基于分布式人工智能，针对大型机电设备的故障诊断，提出分布式诊断专家系统的体系结构。指出任何大型机电设备在结构上和功能上都具有分布性和层次性，诊断任务可按结构、空间、时间等３方面进行分解；分布式故障诊断系统中各系统（视为结点）间采用同僚式组织结构，结点内部的体系结构分为控制层和领域层；各结点之间采用信件传递通信方式，而在结点内部各模块采用共享存贮器方式；由于信件采用框架结构，且假设也采用框架结构，使结点之间的沟通与合作大为简化，并论述诊断任务的分解，提出基于知识的合作策略。

The architecture for distributed fault diagnostic expert system on a large mechanic-electric equipment is proposed based on distributed artificial intelligence. Based on the properties of distribution and hierarchy in the structure and function of a large mechanic-electric equipment, the diagnostic task is decomposed in the field of structure,space and time. The architecture between agents is the system takes the form of companionship, and the architecture of every agent is divided into the control lay and domain lay. The structure of communicative letters and the structure of fault hypotheses take the same form of frame,so that the cooperation between agents becomes very simple. Meanwhile, the strategy of cooperation based on knowledge is presented.