卡尔曼滤波算法在变压器绕组脉冲频率响应法中的应用  

Application of calman filtering algorithm in pulse frequency response analysis of transformer winding

作　　者：王文山[1] 田源[2] 周峰[1] 赵雪骞[1] 程槐号 胡焕[2] 李昊鸾 

机构地区：[1]国网北京市电力公司 [2]华北电力大学河北省输变电设备安全防御重点实验室

出　　处：《电子世界》2016年第22期199-200,F0003,共3页Electronics World

摘　　要：如果将纳秒级脉冲源应用于变压器绕组变形故障诊断,可以将检测频段提高至10MHz,甚至更高,从而提高FRA的检测灵敏度。但是,现场的噪声干扰会影响到脉冲频率响应曲线的形状。为了准确获得关于绕组变形状态的判断,本文提出了基于卡尔曼滤波算法的脉冲信号处理方法。本文将直接对信号进行FFT得到的绕组频率响应曲线,与对信号先进行卡尔曼滤波再作FFT得到的绕组频率响应曲线进行了对比。试验结果表明,本文提出的脉冲信号处理方法具有可行性及较好的效果。If the nanosecond pulse source is applied to the transformer winding deformation fault diagnosis,the detection frequency band can be increased to 20 MHz,or even higher,so as to improve the detection sensitivity of FRA.However,the noise interference will affect the shape of the pulse frequency response curve.In order to obtain the accurate judgment of the winding deformation,a pulse signal processing method based on Calman fi ltering algorithm is proposed in this paper.In this paper,the frequency response curves of the windings obtained directly from the FFT result of the original signal are compared with the frequency response curves of the windings obtained from the FFT result of the signal fi ltered by Calman fi lter.The experimental results show that the pulse signal processing method presented in this paper is feasible and effective.

关 键 词：电力变压器 绕组变形 噪声干扰 卡尔曼滤波 

分 类 号：TN953[电子电信—信号与信息处理]