燃气轮机故障诊断专家系统的设计与实现

Design and Fulfillment of Fault Diagnostic Expert System for Gas Generator

建立了一个适用于燃气轮机的故障诊断专家系统。介绍了该系统的结构、功能、自学能力和知识库的建立及推理机的实现。知识库采用产生式表示法，其组织结构分为三级：大范围经验知识、小范围经验知识和参考解决方案。系统采用双推理机机制，逆向精确推理用于教学和演示，能够查看和解释系统的推理过程；逆向不精确推理用于实际工作场合，提供具有较高可信度的推理结果。

This paper introduces the fault diagnostic expert system that includes system structure,function and self-study capacity for gas generator. The built of knowledge base and fulfillment of inference engine are mainly described. The structure of knowledge base is divided into three parts:knowledge in a large range, knowledge in a small range, and the reference-resolving program. The system uses twin inference engine mechanism. Converse precise inference is used in education and demonstraion,which can examine and explain inference process. Converse fuzzy-inference is applied to indusny production, which provides believable inference result.