An Identification Method of Switchgear Insulation Defect Based on Partial discharge Statistical Characteristics and the Radar Map

Inhalt:
Switchgear is the key equipment in the distribution network, and fast and efficient accurate identification of partial discharge for switchgear is critical for the early warning of insulation failure. In this paper, a feature characteristics extraction and a simple and intuitive identification method is proposed by experimentally simulating the partial discharges of typical defects of switchgear equipment. Firstly, different discharge models are built in the laboratory and UHF sensors are used to obtain the partial discharge signals. Subsequently, 6 partial discharge statistical feature characteristics are calculated by analyzing the characteristics of the 3-dimensional partial discharge patterns. The feasibility of defect type identification based on 6 feature characteristics is verified by using K-means clustering. On this basis, the 6 partial discharge statistical feature characteristics of different defects are plotted in the radar map, and different typical defects are identified visually and clearly by the shape and area presented by the feature characteristics in the radar map. The results show that the selected statistical features: the maximum magnitude of discharge, the average magnitude of discharge, the total amount of discharge magnitude, the total number of discharge, the ratio of positive and negative half-period total discharge magnitude and the mean value of the standard deviation of discharge magnitude can well reflect the discharge characteristics of different defects, and the accurate identification of fault types can be realized by drawing the radar map of different defects' features