基于改进小波包阈值的GIS局部放电信号降噪方法研究

特高频法是GIS设备在状态检测中最常用的方法之一，但在应用过程中传感器所采集的特高频信号通常会携带较多噪声，会干扰检测的准备性，因此对信号进行有效的噪声抑制是局部放电检测的关键技术。本文基于小波包去噪理论，提出了一种综合考虑平滑度和边缘特征的改进阈值函数，以实现对GIS局部放电特高频模拟信号的降噪。通过信噪比（SNR）、均方误差（MSE）和波形相似数（NCC）三种参数对方法的去噪效果进行评估，结果表明，改进小波包阈值的信号降噪方法具有良好的噪声抑制功能，应用后可提高GIS局部放电采集信号的初始数据质量。

Research on Noise Reduction Method of GIS Partial Discharge Signal Based on Improved Wavelet Packet Thresholding

Ultrahigh frequency (UHF) method is one of the most commonly used monitoring method for GIS equipment. However, the UHF signal collected by the sensor in the application process usually carries much noise, which will interfere the accuracy of the detection. Effective noise suppression of the signal is therefore a key technique for partial discharge detection in GIS. Based on wavelet packet theory, this paper proposes an improved threshold function that integrates smoothness and edge features, to achieve effective noise reduction of partial discharge UHF analog signals. After evaluating the proposed method by three parameters, namely signal-to-noise ratio (SNR), mean square error (MSE) and waveform similarity number (NCC), the results show that the signal noise reduction method with improved wavelet packet threshold has good noise suppression, improving the initial data quality of the GIS partial discharge acquisition signal in application.