基于专家系统的原理，模拟专家的思维活动，集众专家的领域知识，根据应用与管理系统问题特征，建立了故障智能诊断专家系统模型.该模型的知识表示使用了产生式规则形式，推理机采用了规则推理和不精确推理相结合的方法.在不确定性知识处理上，应用确定性理论建立了不精确推理模型.该模型通过人机交互功能及自学习功能，可自动获取和完善知识库中的知识，具有很好的实时性.

Based on the theory of expert systems, simulating thoughts process of experts, gathering domain knowledge of all experts and according to the problem characteristics of application and management system, an expert systems model of malfunction intelligent diagnose has been built up. In the system, the method of knowledge representation is genetic rule, the inference engine combines uncertain inference and rule inference. In order to deal with the uncertain knowledge, an uncertain inference model based on confirmation theory is established. This model can auto-gets and modify knowledge in knowledge base, achieve a better real-time characteristics, through interacting between human and machine as well as self-learning function.