基于自适应多元变分模态分解的电力系统低频振荡模态辨识

针对电力系统多元非线性信号模态辨识困难的问题,提出一种自适应多元变分模态分解方法(自适应MVMD)对多元低频振荡信号进行辨识.自适应MVMD法通过对构建的多元约束变分模型迭代求解获得最优分离模态集合,避免了噪声扰动下的模态混叠及虚假模态等问题.首先通过最大复原近似度确定分离模态数K,然后利用自适应多元变分模态分解法对多元信号进行辨识以获得模态集合,对各信号中同频模态分类提取,并利用Hilbert变换以及傅里叶变换频谱分布对振荡参数进行辨识.测试算例及仿真算例证明了该方法的有效性,与经验模态分解法对比结果显示自适应MVMD法对含噪声信号辨识能力更强.

Low-frequency Oscillation Mode Identification of Power System Based on Adaptive Multivariate Variational Mode Decomposition

Aiming at the difficulty of multi-linear signal modal identification in power system,an adaptive multivariate variational mode decomposition(adaptive MVMD)method is proposed to identify multi-oscillation signals.The adaptive MVMD algorithm obtains the optimal set of separated modes by iteratively solving the constructed multivariate constrained variational model,avoiding the problems of modal aliasing and false mode under noise disturbance.Firstly,the separation modal number K is determined by the maximum restoration approximation degree,then the multivariate signal is identified by the adaptive multivariate variational mode decomposition method to obtain the separated modal set,the same frequency signal in each channel is classified and extracted,and the Hilbert transform is utilized.And the Fourier transform spectrum distribution identifies the oscillation parameters.The test examples and simulation examples demonstrate the effectiveness of the proposed method.Compared with the empirical mode decomposition method,the adaptive MVMD method has stronger ability to identify noise-containing signals.