基于专家系统开发工具的汽轮机振动故障多征兆模糊诊断方法  

Multiple-symptom fuzzy fault diagnosis for steam turbine vibration based on CLIPS

作　　者：王正英[1] 王太勇[1] 胥永刚[2] 李强[1] 

机构地区：[1]天津大学机械工程学院,天津300072 [2]北京工业大学机电学院北京市先进制造技术重点实验室,北京100022

出　　处：《中国电力》2008年第4期67-70,共4页Electric Power

基　　金：国家自然科学基金资助项目(50475117);高校博士点基金资助项目(20060056016)

摘　　要：提出一种基于专家系统开发工具(CLIPS)的汽轮机振动故障多征兆模糊诊断方法,研究了多征兆模糊规则的诊断知识表示及规则结论的可信度组合,使用Visual C++和CLIPS开发了火电厂汽轮机振动故障诊断专家系统。该系统能根据汽轮机振动现象进行故障诊断,先通过频谱振幅信息进行初步诊断,再利用多征兆自动诊断和人工模糊诊断相结合的方法,选择出现的故障征兆,确定各征兆的发生程度,进一步缩小故障范围。诊断实例验证了该系统的可靠性。A multiple-symptom fuzzy fault diagnosis method based on CLIPS was proposed to diagnose the steam turbine vibration. The expression of multiple symptom fuzzy diagnosis knowledge and subordinate degree of rules were researched. On the platform of Visual C ＋＋ and expert system tool CLIPS, the steam turbine vibration fault diagnosis expert system was developed. This system can perform the fault diagnosis based on the vibration of steam turbine. That is performing preliminary diagnosis according to the spectral amplitude information, and using the combined diagnosing method to choose the symptoms, determine the probability of faults and further reduce the fault coverage. The diagnosis instances verified the fault diagnosing method.

关 键 词：汽轮机振动 故障诊断 多征兆 模糊诊断 CLIPS 

分 类 号：TK268.1[动力工程及工程热物理—动力机械及工程]