| 基于小波变换和灰度-梯度共生矩阵的局部放电特征提取及识别 |
| |
| 引用本文： | 杨攀烁,贾文阁,刘森,李吉生,张平,李旭,李彬,安国庆,安琪,韩晓慧.基于小波变换和灰度-梯度共生矩阵的局部放电特征提取及识别[J].科学技术与工程,2023,23(27):11673-11680. |
| |
| 作者姓名： | 杨攀烁 贾文阁 刘森 李吉生 张平 李旭 李彬 安国庆 安琪 韩晓慧 |
| |
| 作者单位： | 河北科技大学电气工程学院;中广核工程有限公司;大亚湾核电运营管理有限公司;保定天威新域科技发展有限公司 |
| |
| 基金项目： | 河北省重点研发计划项目(20312101D) |
| |
| 摘 要： | 为了能够充分利用局部放电(Partial Discharge, PD)信号中包含的特征信息，提高变压器内部局部放电类型的识别率，文中提出了一种基于小波变换(Wavelet Transform)和灰度-梯度共生矩阵(Gray-GradientCo-occurrence Matrix,GGCM)算法的局部放电类型识别方法。结合变压器内部结构特点，设计四种局部放电缺陷类型，在实验室搭建变压器局部放电实验检测平台，通过脉冲电流法采集局部放电高频电流信号。运用小波变换对非平稳信号处理时的灵活性对局部放电信号脉冲构建时频谱图；然后结合GGCM算法提取时频谱图的15维纹理特征组成特征向量；将特征向量输入到支持向量机(Support Vector Machine,SVM)分类器进行模式识别。结果表明，小波变换和GGCM算法结合的识别方法能够有效地对不同局部放电缺陷类型进行识别。
|
| 关 键 词： | 局部放电 故障诊断 小波变换 灰度梯度共生矩阵 支持向量机 模式识别 |
| 收稿时间： | 2022/12/26 0:00:00 |
| 修稿时间： | 2023/7/14 0:00:00 |
Feature extraction and recognition of partial discharge based on wavelet transform and GGCM |
| |
| Yang Panshuo,Jia Wenge,Liu Sen,Li Jisheng,Zhang Ping,Li Xu,Li Bin,An Guoqing,An Qi,Han Xiaohui.Feature extraction and recognition of partial discharge based on wavelet transform and GGCM[J].Science Technology and Engineering,2023,23(27):11673-11680. |
| |
| Authors: | Yang Panshuo Jia Wenge Liu Sen Li Jisheng Zhang Ping Li Xu Li Bin An Guoqing An Qi Han Xiaohui |
| |
| Affiliation: | School of Electrical Engineering, Hebei University of Science and Technology |
| |
| Abstract: | In order to make full use of the characteristic information contained in the partial discharge (PD) signal and improve the recognition rate of the partial discharge type in the transformer, this paper proposes a partial discharge type recognition method based on wavelet transform (WT) and Gray-Gradient Co-occurrence Matrix (GGCM) algorithm. According to the internal structure characteristics of the transformer, four types of partial discharge defects are designed, and the transformer partial discharge experimental detection platform is built in the laboratory, and the high-frequency partial discharge current signal is collected by pulse current method. The flexibility of wavelet transform in non-stationary signal processing is used to construct time-frequency spectrum of partial discharge signal pulse; Then, combined with the GGCM algorithm, the 15-dimensional texture features of the time-frequency spectrum are extracted to form the feature vector; Input the feature vector into the support vector machine (SVM) classifier for pattern recognition. The results show that the recognition method combining wavelet transform and GGCM algorithm can effectively identify different types of partial discharge defects. |
| |
| Keywords: | partial discharge fault diagnosis wavelet transform gray gradient co-occurrence matrix support vector machine pattern recognition |
|
| 点击此处可从《科学技术与工程》浏览原始摘要信息 |
|
点击此处可从《科学技术与工程》下载全文 |
|
