基于改进小波阈值的GIS局部放电在线监测白噪声干扰抑制研究

Research on White Noise Interference Suppression of GIS Partial Discharge On-line Monitoring Based on Improved Wavelet Threshold

进行气体绝缘金属封闭开关设备(GIS)局部放电在线监测时,为解决多态性局部放电脉冲信号受白噪声的干扰问题,文中在现有的小波阈值算法降噪的基础上,构造了任意阶可导的新阈值函数,克服了硬阈值函数的不连续性,改进了软阈值函数存在的恒定偏差问题。通过仿真构建白噪声干扰下的多态性局部放电信号,使用降噪后信号的信噪比及均方误差作为降噪评价指标,分析比较该降噪算法与传统的基于硬阈值与软阈值的小波降噪算法,发现该方法更能有效的抑制噪声干扰。最后通过现场模拟电晕放电模型实验,验证改进阈值算法适合应用到GIS局部放电在线监测系统中。

In order to solve the problem that polymorphic partial discharge pulse signal is interfered by white noise when on-line monitoring partial discharge in gas insulated switchgears(GIS),a new threshold function of arbitrary order is constructed based on the existing wavelet threshold algorithm. It overcomes the discontinuity of hard thresh-old function and the problem of constant deviation of soft threshold function is improved. By constructing the poly-morphic partial discharge signal under the interference of white noise,the signal noise ratio(SNR)and the mean square error(MSE)are used as the noise reduction evaluation index it is found that the method is more effective to suppress noise interference by comparing and analyzing the traditional wavelet denoising algorithm based on the hard threshold and the soft threshold. Finally the improved threshold algorithm is validated to be suitable for on-line monitoring system of partial discharge in GIS by field simulation corona discharge model experiment.