摘要
针对凝汽器故障诊断中的复杂性和不确定性,对凝汽器故障征兆集与故障论域进行改进。基于信息论中模糊交互测度(fuzzy cross entropy method,FCEM)的概念,提出一种广义的距离测度,计算凝汽器典型故障模糊模式与未知故障模糊模式之间的差异程度。通过对典型故障集的细化和扩充,提高了故障诊断的准确率。最后,将该方法用于600和300MW火力发电厂汽轮机组凝汽器故障诊断中,结果表明该方法准确率达95%以上,且易于工程实现,证明了该方法的有效性。
Due to complexity and uncertainty of condenser fault diagnosis, symptoms and typical faults of condenser were modified. Based on the information theory, the fuzzy cross entropy method (FCEM) was introduced to measure the distance between the known and unknown fuzzy fault models. The accuracy of fault diagnosis was improved through the refinement and expansion of symptoms and typical faults. With the application on the 600 and 300 MW power plant simulator, the results show that the accuracy rate of recognition is 95% and easy to achieve in practice. The result also proves that the new method is reasonable and feasible.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2009年第20期6-11,共6页
Proceedings of the CSEE
基金
国家重点基础研究发展计划项目(973项目)(2007CB206904)
吉林省科技发展计划项目(20070529)~~
关键词
电站仿真机
凝汽器
故障诊断
模糊交互熵算法
征兆集
power plant simulator
condenser
fault diagnosis
fuzzy cross entropy method
symptoms