摘要
提出一种基于模糊C均值聚类(FCM)、隐马尔可夫模型(HMM)和支持向量机(SVM)相结合的电力电子故障诊断方法.采用FCM方法对故障信号进行模糊聚类,提取故障特征;根据隐马尔可夫模型进行动态过程建模;根据支持向量机进行模式分类;基于HMM-SVM混合的故障诊断模型实现了对机车变流器电路中晶闸管断路故障的诊断.实验结果分析表明,该方法能准确地对电力电子电路进行诊断和故障元定位,诊断精度高,具有很好的实用价值.
Based on Fuzzy C-mean clustering（FCM）,Hidden Markov Model（HMM） and Support Vector Machine（SVM）,a mixed diagnostic model is presented for power electronic circuit faults diagnosis in this paper.FCM is applied to fuzzy cluster for fault signals and to extract the fault features,HMM is applied to deal with continuous dynamic signals and to calculate the matching degree,and SVM is applied to classify fault models and to diagnose faults.The power electronic circuit faults diagnosis in locomotive convertors is implemented with the FCM-HMM-SVM based mixed diagnostic model.Experimental results show that the proposed method can detect and locate faults with high precision.
出处
《电力科学与技术学报》
CAS
2010年第2期61-67,共7页
Journal of Electric Power Science And Technology
基金
福建省自然科学基金(A0710003)
福建省教育厅科学基金(JB06045)
关键词
故障诊断
电力电子电路
模糊C均值
隐马尔可夫模型
支持向量机
fault diagnosis
power electronic circuit
fuzzy C-mean clustering
discrete hidden markov model
support vector machine
