基于时频流形的水轮发电机组局放信号特征提取方法

发布时间：2017-12-27 17:13

本文关键词：基于时频流形的水轮发电机组局放信号特征提取方法　出处：《水力发电学报》2016年09期 　论文类型：期刊论文

【摘要】：准确地提取水轮发电机机组局部放电信号特征,对于发电机绝缘在线监测具有重要意义。为此,本文提出了基于时频流形的发电机局部放电信号特征提取方法。首先通过相空间重构算法,将局部放电(PD)时域信号转换为多个子序列,并分别求其时频分布,构建PD信号的时频流形。然后利用局部线性嵌入算法(LLE)将高维数据映射到低维空间,提取PD信号在低维空间的特征参数。最后,通过K-最近邻分类器(KNNC)的故障诊断模型实现发电机组不同局部放电的模式识别,其故障识别率高于95%。
[Abstract]:The accurate extraction of the characteristics of the partial discharge signal of the turbine generator set is of great significance to the on-line monitoring of the insulation of the generator. In this paper, a feature extraction method for local discharge signal of generator based on time-frequency manifold is proposed. First, phase space reconstruction algorithm is applied to transform the partial discharge (PD) time domain signal into many sub sequences, and the time-frequency distribution is obtained respectively, and the time-frequency manifold of PD signal is constructed. Then, the local linear embedding algorithm (LLE) is used to map the high dimensional data to the low dimensional space and extract the characteristic parameters of the PD signal in the low dimensional space. Finally, the fault diagnosis model of the K- nearest neighbor classifier (KNNC) is used to realize the pattern recognition of the different partial discharge of the generator set, and the fault recognition rate is higher than 95%.
【作者单位】：西安理工大学水利水电学院;
【基金】：国家自然科学基金(51279161)
【分类号】：TM312
【正文快照】： 0引言发电机作为电力系统的主电源设备,其的安全性与可靠性关乎整个系统的运行安全。在实际运行中,发电机主绝缘体长期受电应力、电磁力、机械力及化学腐蚀的因素影响,进而导致其防晕层破坏、断裂、电树叶形成及绝缘碳化等的现象产生,威胁机组运行。据统计资料显示[1-2]:在发

【相似文献】