高压断路器振声联合故障诊断方法

发布时间：2018-01-05 21:01

本文关键词：高压断路器振声联合故障诊断方法　出处：《电工技术学报》2014年07期 　论文类型：期刊论文

【摘要】：针对现有的高压断路器机械故障诊断方法存在的不足,本文提出了一种新的高压断路器振声联合诊断机械故障的方法。此方法首先利用快速核独立分量分析(fast KICA)对采集到的声波信号进行盲源分离处理,并对处理后的声波信号和采集到的振动信号进行改进集合经验模式分解(EEMD)。其次,对分解后的每一个固有模态函数(IMF)求其二维谱熵,并以此二维谱熵矩阵的变换矩阵作为支持向量机的输入特征向量对断路器机械状态进行识别。最后实验表明,振声联合复合分析方法有效提高了高压断路器机械故障诊断的正确性和实用性。
[Abstract]:Aiming at the shortcomings of the existing mechanical fault diagnosis methods of high voltage circuit breakers. In this paper, a new method for mechanical fault diagnosis of high voltage circuit breakers by vibration and acoustic combination is proposed. This method firstly uses fast KICA (Fast Kernel Independent component Analysis). Blind source separation is performed on the collected acoustic signals. The processed acoustic signal and the collected vibration signal are decomposed by improved set empirical mode decomposition. Secondly, the two-dimensional spectral entropy of each inherent mode function after decomposition is obtained. The transformation matrix of the two-dimensional spectral entropy matrix is used as the input eigenvector of the support vector machine to identify the mechanical state of the circuit breaker. The accuracy and practicability of mechanical fault diagnosis of high voltage circuit breakers are improved effectively by combined vibration and acoustic analysis.
【作者单位】：华北电力大学保定校区电气工程学院;
【分类号】：TM561
【正文快照】： 1引言高压断路器是电力系统中最重要的设备之一[1,2],随着人工智能的发展,神经网络、支持向量机、人工免疫网络等新方法被引入故障诊断的领域,但支持向量机适于解决小样本、高维数、非线性等问题[3-5]。由于高压断路器运行现场环境复杂,采集到的信号若采用传统的时频分析方法,

【相似文献】