ICEEMDAN-小波阈值法在谐波信号中的应用

Application of ICEEMDAN-Wavelet Thresholding Method in Harmonic Signals

为了在噪声环境下检测电力系统谐波信号,引入一种基于改进自适应噪声集合经验模态分解(ICEEMDAN)与小波阈值降噪的电力系统谐波信号检测新方法。该方法首先通过ICEEMDAN对含噪谐波信号分解,获取若干内在模态分量(IMF),计算每个IMF与原谐波信号的相关系数来确定含噪模态分量;然后使用改进的小波阈值函数对含噪分量进行去噪处理,通过希尔伯特变换(HT)获取谐波信号的幅频信息;最后将去噪后的IMF与优势IMF进行重构得到去噪后的信号。算例分析表明:ICEEMDAN和改进的小波阈值联合检测使得电力系统稳态谐波的平均误差为3.11%,暂态谐波的平均误差为3.64%,信噪比提高且去噪后的信号能够更真实地反映电力系统扰动信号的特征。

In order to conveniently detect power system harmonic signals in noisy environments,a new method for power system harmonic signal detection based on improved complete ensemble empirical mode decomposition with adaptive noise(ICEEMDAN)with wavelet threshold noise reduction was introduced.Firstly,the noisy harmonic signal was decomposed by ICEEMDAN to obtain several intrinsic mode components(IMFs),and the noisy mode components were determined by calculating,the correlation coefficients between each IMF and the original harmonic signals.Then,the modified wavelet threshold function was used to de-noise the noisy components,and the amplitude-frequency information of the harmonic signal was obtained by Hilbert transform(HT).Finally the denoised IMFs and the dominant IMFs were reconstructed to obtain the denoised signals.Example analysis shows that the combination of ICEEMDAN and the improved wavelet threshold method has an average error of 3.11%on Steady-state harmonics in power systems and 3.64%on transient harmonics,the signal-to-noise ratio has been improved,and the denoised signals are able to realistically reflect the characteristics of the harmonic signals of the power system.