基于经验模态分解能量分布和灰色相似关联度的污秽绝缘子放电识别

Contaminated Insulator Discharge Recognition Method Based on Empirical Mode Decomposition Energy Distribution and Grey Similar Incidence

在深入研究污秽绝缘子放电声发射信号的基础上,提出通过对污秽绝缘子放电声信号进行经验模态分解和灰色相似关联度的方法实现放电程度的识别。针对不同放电模式下的声信号进行经验模态分解得到本征模态分量,计算电晕放电、局部放电和电弧放电3种放电模式的声发射信号各阶本征模态分量能量分布。根据前几阶本征模态分量的能量分布构造特征矢量,计算声发射信号与不同放电模式的特征矢量的灰色相似关联度判断绝缘子所处的放电阶段,实现绝缘子外绝缘状态的监测。实例计算结果表明该方法的有效性。

On the basis of in-depth study of acoustic emission signals emitted from contaminated insulator discharge, a new method is proposed to achieve contaminated insulator discharge recognition by empirical mode decomposition and grey similar incidence. The intrinsic mode functions of sound signals in different discharge modes is got by empirical mode decomposition and the energy distribution of acoustic emission signals intrinsic mode components in three discharge modes which include the corona discharge, partial discharge and arc discharge is calculated. The feature vector is constructed according to the energy distribution of the first few intrinsic mode components containing the most dominant information. The grey similar incidence of different acoustic emission signals and feature vector in different discharge modes is calculated to identify the discharge modes of the insulator, which realizes the monitoring of the external insulation condition of insulator. Practical calculation results show the effectiveness of this method.