Research on a method for GIS partial discharge pattern recognition based on polar coordinate map

作者：Jun Xue, Xiaolan Zhang, Weidong Qi, Guoqiang Huang, Bo Niu, Jing Wang · 年份：2016 · DOI：10.1109/ichve.2016.7800595 · 被引用次数：5 · 研究领域：Advanced Algorithms and Applications、Embedded Systems and FPGA Design、Evaluation Methods in Various Fields

This paper proposes a brand new method for drawing partial discharge fault maps based on polar coordinate for insulation defects of GIS. As different from a traditional plane coordinate map, the polar coordinate map can realize head-tail connection of the power frequency signal phase, so as to better observe the phase distribution of partial discharge pulses with the cycle, especially the symmetric characteristics of both phase distribution and amplitude distribution. Adaptive classification is used to classify partial discharge pulses into a number of classes, and the number, characteristics and comparison of these classes can contribute to better analysis on partial discharge. The phase distribution characteristics of each class with concentrated partial discharge can be clearly observed through the given median value and quartile value of phase. The amplitude distribution characteristics of each centralized PD class can be clearly observed through the given median value and 75% fractile of amplitude. Therefore, this method can overcome the difficulty in observing the distribution characteristics of each class of partial discharge pulse in a plane coordinate system and can effectively improve the efficiency and accuracy of GIS partial discharge pattern recognition.