Abstract：A new approach for power quality disturbance identification based on local mean decomposition (LMD) energy entropy and GK fuzzy clustering is introduced. LMD energy entropy has localized features and can represent the complexity of disturbance signals, quantizing disturbance characteristic. GK fuzzy clustering can process data with irregular distribution, so as to identify various disturbance signals. Non-stationary disturbance signals are decomposed by LMD, obtaining a number of product function (PF) components with physical meaning, which is screened out by Shannon entropy feature selection methods to obtain the energy entropies and construct the eigenvectors. The constructed eigenvectors are further put into GK classifier to recognize different identification types. The experiment demonstrated that the method is able to identify disturbance signals accurately, with better anti-noise performance.
［8］Smith J S. The local mean decomposition and its application to EEG perception data［J］. Journal of the Royal Society Interface,2005,2(5):443-454.
［9］李亮. 基于EEMD样本熵和模糊聚类的轴承故障诊断方法研究［D］. 秦皇岛:燕山大学,2013.
［10］程建,陆奎. 基于MATLAB的GK模糊聚类算法程序设计［J］.安徽理工大学学报:自然科学版,2006,26(1):38-40.
［11］张淑清,孙国秀,李亮,等. 基于LMD近似熵和FCM聚类的机械故障诊断研究［J］. 仪器仪表学报,2013,34(3):714-720.
［12］卢雪峰,王增平,徐岩. 应用模糊贴近度原理识别励磁涌流的新方法［J］. 高电压技术,2008,34(1):154-157.
［13］Lei Y, He Z, Zi Y, et al. New clustering algorithm-based fault diagnosis using compensation distance evaluation technique［J］. Mechanical Systems and Signal Processing,2008,22(2):419-435.
2016, Vol. 37 
