故障诊断系统中深浅知识集成表示方法的研究  

INTEGRATED MODEL OF SHALLOW AND DEEP KNOWLEDGE IN FAULT DIAGNOSIS SYSTEMS

作　　者：王道平[1] 马少平[1] 

机构地区：[1]清华大学计算机系智能技术与系统国家重点实验室,北京100084

出　　处：《小型微型计算机系统》2001年第9期1025-1028,共4页Journal of Chinese Computer Systems

基　　金：国家自然科学基金资助项目 (批准号 79770 0 2 5 )

摘　　要：传统的故障智能诊断系统大都只是基于专家的经验 .但是 ,由于现代设备的复杂性 ,使其往往得不到准确的结果 ,这时必须借助于与诊断对象有关的原理性知识 .本文基于设备的层次性理论 ,提出了一种深浅知识集成表示模型 ,并采用面向对象表示和产生式规则相结合的方法予以实现 .深浅知识的有效结合 ,不仅可以提高故障诊断的准确性 。Traditionally, the intelligent systems for fault diagnosis are only based on expert experience called shallow knowledge. However, they sometimes lead to dissatisfactory conclusions as a result of the complexity of modern equipment. To obtain better results, the knowledge about the principle of equipment called deep knowledge appears in these systems. In this paper, based on the hiberarchy of equiment, we first present an integrated model that combines the two kinds of knowledge. Then, the model is realized by using object oriented programme and production rules. By using shallow and deep knowledge synthetically, the veracity and the efficiency of systems are much improved.

关 键 词：故障诊断系统 知识表示 深知识 浅知识 专家系统 人工智能 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置] TP18[自动化与计算机技术—控制科学与工程]