基于改进CEEMDAN在电能质量复合扰动去噪中的应用

Application of power quality composite disturbance denoising based on improved CEEMDAN

为提高噪声环境下电能质量复合扰动识别精度,提出一种基于改进自适应噪声完备经验模态分解(CEEMDAN)去噪算法。首先通过CEEMDAN方法将含噪信号分解为若干本征模态函数(IMF);然后将改进兰氏距离与多重分形去趋势波动分析(MFDFA)结合,把若干IMF分量分为信号IMF分量、噪声和信号混叠IMF分量、噪声IMF分量。对于混叠IMF分量、噪声IMF分量分别采用改进奇异谱分析(SSA)、小波阈值(WT)去噪;最后,将经去噪处理的IMF分量与信号IMF分量进行重构。实验表明:与对比算法相比,含噪扰动经新算法去噪后,信噪比显著提高,去噪效果良好。

In order to improve the recognition accuracy of power quality composite disturbances in noisy environments,a denoising algorithm based on the improved complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)is proposed.The noise⁃containing signal is decomposed into several intrinsic modal functions(IMFs)by the CEEMDAN method,and then the improved Lance and Williams distance is combined with multifractal detrended fluctuation analysis(MFDFA)to classify several IMF components into a signal IMF component,a noise and signal aliasing IMF component,and a noise IMF component.Improved singular spectrum analysis(SSA)and wavelet thresholding(WT)are used to denoise the aliased IMF component and the noise IMF component,respectively.The denoised IMF components are reconstructed with the signal IMF component.The experiments show that after the noise⁃containing perturbation is denoised with the proposed algorithm,the signal⁃to⁃noise ratio(SNR)is improved significantly,and the denoising effect is good in comparison with the contrast algorithm.