论文部分内容阅读
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。
Based on the steady state characteristics of infrared laser spectroscopy, a method of transformer fault diagnosis based on infrared laser spectroscopy is proposed. The infrared laser spectrum characteristic acquisition model and the infrared laser spectrum signal analysis model of the transformer in the fault condition are established to deal with the anti-interference filtering of the collected infrared laser spectrum signal and the Hilbert-Huang time-frequency characteristic in the time-frequency domain of the infrared laser spectrum signal Extraction methods for fault feature extraction, the time-frequency distribution Hilbert spectrum of transformer fault feature information is obtained to effectively analyze the infrared laser spectrum of the transformer fault, the feature of the spectrum is classified by the support vector machine classifier to realize the accurate diagnosis of the fault . Simulation analysis shows that this method has higher accuracy, better real-time performance and superior performance of technical indicators.