电工电气杂志社

基于VGG16图像特征提取和SVM的电能质量扰动分类

来源：电工电气发布时间：2023-07-27 14:27 浏览次数：361

基于VGG16图像特征提取和SVM的电能质量扰动分类

童占北¹，钟建伟¹，李祯维²，吴建军²，李家俊²

(1 湖北民族大学智能科学与工程学院，湖北恩施 445000；

2 国网湖北省电力有限公司恩施供电公司，湖北恩施 445000)

摘要：针对传统电能质量扰动分类方法需人工选取特征量，易受人为经验干扰的问题，提出一种基于 VGG16 图像特征提取和支持向量机 (SVM) 结合的电能质量扰动分类方法。通过格拉姆角场 (GAF) 将电能质量扰动信号映射到极坐标系中，生成格拉姆矩阵，并转换为二维扰动图像；利用 VGG16 网络自动提取图像特征的特点，将扰动图像输入 VGG16 网络中进行提取；将提取的特征数据作为 SVM 分类器的输入，并引入十折交叉验证对SVM 进行训练，以提升分类器的性能，最后对扰动信号进行准确分类。仿真结果表明，该方法对于电能质量扰动的分类具有较高的准确率。

关键词： 电能质量；扰动分类；格拉姆角场；VGG16 网络；支持向量机；十折交叉验证

中图分类号：TM712 文献标识码：A 文章编号：1007-3175(2023)07-0007-07

Power Quality Disturbance Classification Based on

VGG16 Image Feature Extraction and SVM

TONG Zhan-bei¹, ZHONG Jian-wei¹, LI Zhen-wei², WU Jian-jun², LI Jia-jun²

（1 College of Intelligent Systems Science and Engineering, Hubei Minzu University, Enshi 445000, China;

2 State Grid Hubei Electric Power Co., Ltd. Enshi Power Supply Company, Enshi 445000, China）

Abstract: Traditional power quality disturbance classification methods need to manually select feature quantities, which are susceptible to human experience interference. Hence, the paper proposes a power quality disturbance classification method based on the combination of VGG16 image feature extraction and Support Vector Machine (SVM). It first maps power quality disturbance signals to the polar coordinate system through Gramian Angular Field (GAF) to generate the Gramian matrix which is transformed into a two-dimensional disturbance image. Second, the characteristic of VGG16 network to automatically extract image features is used to input disturbed images into VGG16 network for extraction. Third, the extracted feature data is used as the input of SVM classifier, ten-fold cross-validation is introduced to train SVM to improve the performance of the classifier, and then disturbance signals are classified in an accurate way. The simulation results show that this method has higher accuracy of power quality disturbances classification.

Key words: power quality; disturbance classification; Gramian angular field; VGG16 network; support vector machine; ten-fold cross-validation

参考文献

[1] WANG K, XI Y.A new method of power quality disturbance classification based on deep belief network[J].Journal of Physics Conference Series，2021，1827(1)：012021.

[2] HUANG J, QU H, LI X.Classification for hybrid power quality disturbance based on STFT and its spectral kurtosis[J].Power System Technology，2016，40(10)：3184-3191.

[3] 徐佳雄，张明，王阳，等. 基于改进 Hilbert-Huang 变换的电能质量扰动定位与分类[J]. 现代电力，2021，38(4)：362-369.

[4] 程志友，杨猛. 基于二维离散余弦 S 变换的电能质量扰动类型识别[J] . 电力系统保护与控制，2021，49(17)：85-92.

[5] 徐艳春，樊士荣，谭超，等. 基于改进 EWT-CMPE 的高渗透率主动配电网电能质量扰动检测与分类[J]. 电网技术，2020，44(10)：3991-4000.

[6] 杨剑锋，姜爽，石戈戈. 基于分段改进 S 变换的复合电能质量扰动识别[J] . 电力系统保护与控制，2019，47(9)：64-71.

[7] SUNDARAM P K, GIRISHKUMAR G.Power Quality Disturbance Classification Based on Kalman Filter and Adaptive Neural Fuzzy Inference System (ANFIS)[C]//International Conference on Robotics and Artificial Intelligence，2021.

[8] BHAGAT A, NIMKAR S, DONGRE K, et al.Power Quality Disturbance Detection and Classification Using Artificial Neural Network Based Wavelet[J].International Journal of Computational Intelligence Research，2017，13(8)：2043-2064.

[9] THIRUMALA K , PAL S , JAIN T , et al . A classification method for multiple power quality disturbances using EWT based adaptive filtering and multiclass SVM[J].Neurocomputing，2019，334 ：265-274.

[10] MAHELA O P, SHAIK A G, KHAN B, et al.Recognition of Complex Power Quality Disturbances Using S-Transform Based Ruled Decision Tree[J].IEEE Access，2020，8：173530-173547.

[11] DAWOOD Z, BABULAL C K.Power quality disturbance classification based on efficient adaptive Arrhenius artificial bee colony feature selection[J].International Transactions on Electrical Energy Systems，2021，31(5)：e12868.

[12] MALIK H, ALMUTAIRI A, ALOTAIBI M A.Power quality disturbance analysis using data-driven EMD-SVM hybrid approach[J].Journal of Intelligent and Fuzzy Systems，2021，42(6)：1-10.

[13] 曹梦舟，张艳. 基于卷积- 长短期记忆网络的电能质量扰动分类[J] . 电力系统保护与控制，2020，48(2)：86-92.

[14] WANG Z, OATES T.Imaging time-series to improve classification and imputation[C]//Proceedings of the 24th International Conference on Artificial Intelligence，2015：3939-3945.

[15] SHUKLA J, PANIGRAHI B K, Ray P K.Power quality disturbances classification based on Gramian angular summation field method and convolutional neural networks[J].International Transactions on Electrical Energy Systems，2021，31(12)：e13222.

[16] SIMONYAN K , ZISSERMAN A . Very deep convolutional networks for large-scale image recognition[EB/OL].(2015-04-10)[2023-03-10].http://arxiv.org/abs/1409.1556.

[17] 邵凯旋，何怡刚，汪磊. 基于多尺度熵分析与改进 SVM 的变压器故障识别[J] . 电子测量与仪器学报，2022，36(6)：161-168.

[18] MAMAT N, HAMZAH F M, JAAFAR O.Hybrid Support Vector Regression Model and K-Fold Cross Validation for Water Quality Index Prediction in Langat River, Malaysia[EB/OL].(2021-02-15)[2023-03-10].http://www.biorxiv.org/content/10.1101/2021.02.15.431242v1.abstract.

[19] IEEE Power & Energy Society.IEEE recommended practice for monitoring electric power quality:IEEE Std 1159-2019[S].Washington：IEEE，2019：34-79．

返回索引列表

Article retrieval

文章检索

基于VGG16图像特征提取和SVM的电能质量扰动分类