微型电脑应用2025，Vol.41Issue(2)：234-239,6.
基于瞬时能量特征排列熵的局部放电位置辨识
Partial Discharge Location Identification Based on Permutation Entropy of Instantaneous Energy Features
王喆 1杨聪 1沈润夏 1李淑倩 1屈炳建1
作者信息
- 1. 国网四川省电力公司凉山供电公司,四川,凉山 615000
- 折叠
摘要
Abstract
Partial discharge(PD)positioning is a reliable and necessary technique that can predict the upcoming occurrence of transformer windings early.PD sources are affected by insulation,and PD measured at different locations exhibits different en-ergy characteristics.In this paper,a method for locating PD inside windings based on energy characteristics is proposed.In or-der to reveal the multi-scale eigen characteristics of the PD signal,the intrinsic mode function(IMF)is extracted using inte-grated ensemble empirical mode decomposition(EEMD).The instantaneous energy distribution values of the first few IMF are calculated to capture PD energy,and a fast principal component analysis(F-PCA)based on a fixed-point algorithm is used for characteristic dimensionality reduction.An adaptive density based spatial clustering of applications with nearest neighbor graph(ADBSCAN-NNG)algorithm is proposed to identify local high-density energy components,and use a high-frequency current sensor(HFCT)to verify the feasibility of the proposed algorithm on the vacuum-cast transformer winding.After verification,the positioning results show that the algorithm can realize the positioning of the PD signal of the transformer winding,and the complexity of the algorithm is greatly reduced,which effectively improves the positioning accuracy and speed,and has certain engineering practice guiding significance.关键词
变压器/局部放电/瞬时能量特征/排列熵/最近邻图算法Key words
transformer/partial discharge/instantaneous energy feature/permutation entropy/nearest neighbor graph algo-rithm分类
动力与电气工程引用本文复制引用
王喆,杨聪,沈润夏,李淑倩,屈炳建..基于瞬时能量特征排列熵的局部放电位置辨识[J].微型电脑应用,2025,41(2):234-239,6.
