改善EMD端点效应的煤矿供电系统谐波检测方法

Harmonic Detection Method of Coal Mine Power Supply System to Improve EMD End Effect

为了改善经验模态分解(EMD)在谐波检测时产生的端点效应,提出了波形自适应延拓的改进型EMD算法。以信号端点附近波形的极大值和极小值斜率为基准,寻找信号内具有相同斜率的一组极大值和极小值,为了使寻找到的目标波形与端点处的波形更加相似,引入波形匹配度函数,用其数值大小判断两个波形相似程度。根据端点位置不同,把寻找到的目标波形前一个或后一个波形的极大值和极小值向端点处延拓。在EMD分解过程里用分段三次Hermite插值函数替代三次样条插值函数,消除了包络线过冲和欠冲问题。通过测量稳态谐波与暂态谐波的实验,验证了此算法能高效遏制EMD端点效应,显著提升了分解精度。且测出的谐波幅值和频率与快速傅里叶(FFT)相比更精确,符合电力系统谐波分析要求。

In order to improve the endpoint effect of empirical mode decomposition(EMD)in harmonic detection,an improved EMD algorithm based on waveform adaptive continuation is proposed.The method takes the maximum and minimum slopes of the waveform near the signal endpoint as the benchmark,and finds a group of maximum and minimum values with the same slope in the signal.In order to find the target waveform more similar to the waveform at the endpoint,the waveform matching degree function is introduced,and its numerical value is used to judge the similarity degree of two waveforms.Depending on the position of the end point,the maximum value and the minimum value of the waveform before or after the target waveform found are extended to the end point.In the EMD decomposition process,the cubic spline interpolation function is replaced by the piecewise cubic Hermite interpolation function,which eliminates the overshoot and undershoot of the envelope.Through the experiments of measuring steady-state harmonics and transient harmonics,it is verified that the algorithm can effectively suppress the EMD end effect and significantly improve the decomposition accuracy.And the measured harmonic amplitude and frequency are more accurate compared to Fast Fourier Transform(FFT)and meets the requirements of power system harmonic analysis.