基于变分模态分解的电能质量扰动检测新方法

摘要

In order to extract the characteristics of disturbance signal more accurately,a new method based on VMD(Variational Mode Decomposition) is proposed to detect power quality disturbance.The proposed method is composed of VMD and HT(Hilbert Transform).Firstly,Fourier transform of disturbance signal is used to determine the preset decomposition scale of VMD.Then,the disturbance signal is decomposed into the sum of a series of AM-FM functions by VMD.Finally,the instantaneous amplitude and instantaneous frequency of AM-FM signal are obtained by HT and then the characteristics of disturbance signal are determined.Compared with Hilbert-Huang transform and local mean decomposition method,VMD method can analyze the disturbance signals with different time sets,handle complex disturbances and odd harmonics with similar frequency,has no mode confusion and can obtain the instantaneous amplitude and instantaneous frequency more accurately.Analysis results of simulation signals and voltage signals when switching capacitor bank verify the feasibility and effectiveness of the proposed method.%为了更加准确地提取扰动信号特征,提出了基于变分模态分解(VMD)的电能质量扰动检测新方法.该方法由VMD和希尔伯特变换(HT)2个部分组成.首先,对扰动信号进行傅里叶变换以确定VMD的预设分解尺度;然后,利用VMD将扰动信号分解为系列调幅-调频函数之和;最后,对每个调幅-调频函数进行HT,求取瞬时幅值和瞬时频率,进而确定扰动信号特征.较之希尔伯特-黄变换和局部均值分解方法,VMD方法不仅可分析不同时间支集的扰动信号,处理复合扰动和频率相近的奇数次谐波,也不存在模态混叠,获取的瞬时幅值和瞬时频率更加准确.仿真信号和变电站电容器组投入时的电压信号分析结果证明了所提方法的可行性和有效性.