中国电机工程学报Issue(6)：1541-1548,8.DOI:10.13334/j.0258-8013.pcsee.2015.06.031
基于混沌理论和K-means聚类的有载分接开关机械状态监测
Mechanical Condition Monitoring of On-load Tap Changers Based on Chaos Theory and K-means Clustering Method
摘要
Abstract
ABSTRACT:Reliability of on-load tap changer (OLTC) is essential to the secure and stable operation of power system. Since the vibration signals during the operation of OLTC were closely related to its mechanical performance, the measured vibration signals were reconstructed in the phase space with the calculated embedding dimension and delay time based on the chaotic feature of vibration signal. Then the K-means clustering analysis method was applied to recognize the signal feature according to the distribution pattern of the reconstructed vibration signal. Consequently, the mechanical performance of OLTC could be analyzed successfully. Calculated results of measured vibration signals have shown that the chaotic feature of vibration signals during the operation of OLTC is further verified due to the positive of the largest Lyapunov exponent obtained by the improved Wolf algorithm. The fault of OLTCs can be found clearly according to the variations of displacement vector of cluster centers calculated by K-means method. The obtained results can provide theoretical basis for the condition monitoring of OLTC in the viewpoint of chaos dynamics.关键词
K-means聚类/最大李雅普诺夫指数/有载分接开关/相空间重构/振动信号Key words
K-means clustering/largest Lyapunov exponent/on-load tap-changer/phase space reconstruction/vibration signal分类
信息技术与安全科学引用本文复制引用
周翔,王丰华,傅坚,林嘉杨,金之俭..基于混沌理论和K-means聚类的有载分接开关机械状态监测[J].中国电机工程学报,2015,(6):1541-1548,8.基金项目
国家自然科学基金项目(51207090)。 （51207090）
