红外激光光谱分析的变压器故障诊断  

Transformer fault diagnosis by infrared laser spectroscopy

作　　者：张鹏举[1] 刘建英[2] 

机构地区：[1]内蒙古师范大学计算机与信息工程学院,呼和浩特010022 [2]内蒙古机电职业技术学院电气工程系,呼和浩特010070

出　　处：《激光杂志》2016年第11期46-50,共5页Laser Journal

基　　金：内蒙古自治区高等学校科学研究项目(NJZC14391);内蒙古师范大学科研基金(2013ZRYB04)

摘　　要：利用红外激光光谱的稳态特征,提出一种基于红外激光光谱分析的变压器故障诊断方法。构建变压器在故障情况下的红外激光光谱特征采集模型和红外激光光谱信号分析模型,对采集的红外激光光谱信号进行抗干扰滤波处理,在红外激光光谱信号时频域中采用Hilbert-Huang时频特征提取方法进行故障特征提取,得到变压器故障特征信息的时间-频率联合分布Hilbert谱,实现对变压器故障红外激光光谱有效分析,对谱特征采用支持向量机分类器进行故障特征分类,实现故障的准确诊断。仿真实验分析表明,该方法进行变压器故障诊断的准确性较高,实时性较好,技术指标性能优越。A transformer fault diagnosis method based on infrared laser spectrum analysis is proposed by using the steady state characteristic of infrared laser spectrum. To build the analysis model of transformer fault conditions in the infrared laser spectral acquisition model and infrared laser spectral signal and to collect infrared laser spectroscopy signal of interference filter processing, in the infrared laser spectroscopy signal in frequency domain by using Hilbert Huang frequency characteristics provided method for fault feature extraction, time frequency transformer fault feature information of the joint distribution of the Hilbert spectrum is obtained, implementation of pressure sensor fault infrared laser spectral analysis, on spectral characteristics using a support vector machine classifier to classify the fault characteristics, realize fault accurate diagnosis. The simulation results show that this method is of high accuracy, good realtime performance and good performance of the transformer fault diagnosis.

关 键 词：红外激光 光谱分析 变压器 故障诊断 

分 类 号：TN27[电子电信—物理电子学]