改进VMD和提升小波在局部放电去噪中的应用

Application of Improved Variational Mode Decomposition and Lifting Wavelet Method in Partial Discharge Denoising

针对城市轨道电力电缆局部放电的噪声抑制问题,提出一种改进变分模态分解(VMD)结合提升小波分解的方法来提取纯净的局部放电信号。首先以能量差最小确定分解层数,运用VMD将含有噪声的局部放电信号分解到不同的模态分解中,并通过峭度指标去除掉周期窄带干扰分量和高频白噪声分量,保留局放特征分量;利用提升小波分解进一步去除局放特征分量中残留的白噪声,重构局放特征分量得到纯净的局放信号。将所提方法与提升小波分解算法、VMD算法进行比较分析,仿真结果表明,所提方法噪声抑制效果更好。因此,所提改进VMD结合提升小波法能够在城市轨道电缆局部放电去噪中更有效地应用。

Aiming at the noise suppression of partial discharge in urban rail power cables,an improved variational modal decomposition(VMD)combined with lifting wavelet decomposition was proposed to extract pure partial discharge signals.Firstly,determine the number of decomposition layers with the smallest energy difference,use VMD to decompose the noisy partial discharge signal into different modal decompositions,and remove the periodic narrow-band interference components and high-frequency white components through the kurtosis index,while retain the local discharge characteristic component;the lifting wavelet decomposition was used to further remove the residual white noise in the partial discharge characteristic component,and the partial discharge characteristic component was reconstructed to obtain a pure partial discharge signal.The proposed method was compared and analyzed with lifting wavelet decomposition algorithm and VMD algorithm.The simulation results show that the noise suppression effect of the proposed method is better.Therefore,the proposed improved VMD combined with the lifting wavelet method can obtain more effective applications in the denoising of urban rail cables partial discharge.