基于经验模态分解的水轮发电机组局部放电信号提取

Extracting partial discharge signals of hydroelectric generating set based on empirical mode decomposition

提出一种基于经验模态分解(EMD)的水轮发电机组局部放电信号提取方法。对实际局部放电信号,在频域内降低其干扰的幅值后得到中间信号,然后对中间信号进行经验模态分解,得到包含特征频率的固有模态函数(IMF),接着对所得到的固有模态函数分量局部重构,从而提取出局部放电信号。通过实例证明了该方法的有效性。

This paper investigates a new method of extracting partial discharge signals of hydroelectric generating set based on empirical mode decomposition （EMD）. Considering the factual partial di can be produced by reducing the amplitude of noise in frequency region, which is scharge signals, the mid-signals decomposed with EMD, and the intrinsic mode functions which contain the specific frequency can be obtained, then the partial discharge signals can be extracted by partly restructuring IMF. The validity of this method has been proved by an example.