现代功率谱估计在绝缘子污秽诊断中的应用

Application of Modern Spectral Estimation in Diagnosis of Discharge of Polluted Insulator

针对现有绝缘子在线监测系统基于泄漏电流时域特征分析的诊断和预警很难分析其表面的污秽放电状况的问题。提出了绝缘子泄漏电流信号的功率谱分析法。它采用现代功率谱估计中的AR参数模型在MATLAB中运用Burg算法估计不同污秽状况下的绝缘子泄漏电流数据,并结合泄漏电流时域特征进行了联合分析。结果表明,不同污秽状况下绝缘子泄漏电流的功率谱特性有显著区别,将现代功率谱估计与时域特征分析相结合综合分析可以更好地掌握绝缘子表面污秽放电状况并做出较准确的预警。

Existing online monitoring system of insulator leakage current can only analyze the features in time domain and provide diagnosis results and alarms. However, more detailed analysis of discharge condition on polluted insulator surface cannot be gained in this way. In this paper, the power spectral analysis of insulator leakage current is stated in order to find more features to reveal pollution condition on insulators. The AR parametric model, classified as modern spectral estimation, is adopted. Based on Burg algorithm in MATLAB environment, the spectral estimation of insulator leakage current under different pollution condition is implemented. Moreover, the synthetic analysis associated with the features in time domain has also been carried out. As a result, the spectral features of insulator leakage current under different pollution condition is distinguished each other apparently. The synthetic analysis of insulator leakage current, combined modem spectral estimation with time domain features, will obtain more detailed information of discharge condition on oolluted insulator surface, and then provide alarm more exactly.