基于小波包分解与GG模糊聚类的转辙机退化阶段划分研究

针对道岔转换设备退化特征提取以及性能退化阶段准确划分的问题,提出一种基于小波包分解与GG聚类的退化阶段划分方法.首先对采集的ZYJ7转辙机工作功率曲线进行小波包分解,获得表征道岔转辙机状态的特征向量,构建退化性能指标;其次采用GG模糊聚类方法对道岔转辙机性能退化状态进行阶段划分,识别不同的退化状态;最后选用分类系数、平...

Research on Division of Degradation Stage of Turnout Equipment Based on Wavelet Packet Decomposition and GG Fuzzy Clustering

Aiming at the problem of feature extraction of turnout equipment degradation and accurate division of performance degradation stage,a degradation stage division method based on wavelet packet decomposition and GG clustering was proposed.Firstly,wavelet packet was used to decompose the collected working power curve of the ZYJ7 switch machine to obtain the characteristic vector representing the state of the switch machine and construct the degradation performance index.Secondly,GG fuzzy clustering method was used to divide the performance degradation state of the switch machine to identify different degradation states,and the classification coefficient and average fuzzy entropy were used to synthesize the clustering effect and to compare with GK clustering and fuzzy C-means.The results show that GG clustering effect is obviously better than fuzzy C-means and GK clustering effect,which can effectively realize the accurate division of performance degradation stage of turnout equipment,and provide theoretical support for condition based maintenance on the spot.