基于波形相似度检测的配电网故障原因分析

Cause analysis of distribution network failure based on waveform similarity detection

配电网故障录波元件的波形与网络故障信息判断密切相关,需要判断与原始施加波形的相似度,将高准确度波形用于故障判断。提出了一种基于波形相似度检测的配电网故障原因分析方法。首先分析了电能质量扰动数据的特点和偏差特性,说明了扰动相似度;然后利用基于欧几里得距离的最邻近分类方法说明了波形预处理、相似度检测、语境相似度和联合相似度的计算原理,并提出了基于波形相似度检测的分析流程;最后针对相似度检测的实际算例进行了说明。

The waveform of the distribution network fault recorder is closely related to the judgment of the network fault information.It is necessary to judge the similarity to the original applied waveform and use the highprecision waveform for fault detection.A method for analyzing the cause of a distribution network failure based on waveform similarity detection is proposed.First,the characteristics and deviation characteristics of power quality disturbance data are analyzed,and the disturbance similarity is explained.Using the nearest neighbor classifcation method based on Euclidean distance,the calculation principles of waveform preprocessing,similarity detection,context similarity,and joint similarity are explained.An analysis process based on waveform similarity detection is proposed.Finally,the actual calculation example of similarity detection is explained.