To solve the problems of insufficient hyperparameter optimization and multi-classification performance of support vector machine (SVM) in the fault diagnosis of power transformers, the nonlinear dimension reduction of 26-dimensional dissolved gas analysis(DGA)data is carried out by using the t-distributed stochastic neighbor embedding (t-SNE). Error correction output codes (ECOC) are introduced, and the improved sparrow search algorithm (ISSA) was combined with Chebyshev chaotic mapping and Cauchy-Gaussian variational strategy to optimize the hyperparameters of SVM and handle multi-classification problems. The research results show that the diagnostic accuracy, recall rate, specificity and F1 value of the ECOC-ISSA-SVM(t-SNE) model are 95.6%, 97.8%, 99.6% and 97.8% respectively. Compared with the traditional model, the improvement effect of each index is significant. The diagnostic time is shortened to 11 ms and the diagnostic efficiency is significantly improved. The research conclusion provides technical support for the intelligent operation and maintenance of power equipment.
ECOC采用二值编码策略,结合机器学习方法与纠错输出编码,将n类问题分解为n(n-1)/2个二分类子任务,可在传输过程中实现模型的多类识别和纠错功能[26]。采用ECOC策略下的OVO(one versus one)编码来构建多类模型,将6类故障诊断转化为15个二元分类问题,见表7。其中,1表示正样本问题;-1表示负样本问题;0表示忽略该问题。在进行故障诊断时,分类器Ci (i=1,2,…,15)首先生成测试样本的预测编码,再通过汉明距离匹配预设故障编码库,确定距离最小的编码组合,最终输出该编码组合所对应的故障类别。
LIBin, BAIXiangxu.Research on transformer fault diagnosis of LSSVM optimized by multi-strategy improved COA algorithm[J].Advanced Technology of Electrical Engineering and Energy,2025,44(4):112-119.
XIEGuomin, QIXiaoliang.Research on transformer fault diagnosis based on multi-strategy fusion and improved AO-optimized SVM[J].Control Engineering of China,2024,31(11):2000-2009.
ZHAOXin, WANGDongli, PENGHong,et al.Transformer fault diagnosis based on a multi-strategy improved dung beetle optimizer[J].Power System Protection and Control,2024,52(6):120-130.
ZHAOChao, ZHANGXun, WANGMian,et al.Gas concentration prediction of power transformer test and monitoring data based on improved random forest[J].Electrical Measurement & Instrumentation,2024,61(11):205-210.
YANGXu, ZHOUWen, CHENGLin,et al.Research on multi-source information fusion fault diagnosis of transformer based on KPCA-SVM[J].High Voltage Apparatus,2025,61(2):54-62.
WANGYingjie, CAOTienan.Transformer fault diagnosis method based on improved Elman neural network[J].Journal of Shenyang University of Technology,2021,43(3):254-258.
LIUDidi, WANGYang, LIUHuiqian,et al.POA-SVM transformer fault diagnosis based on ADASYN balanced data set[J].Power System and Clean Energy,2023,39(8):36-44.
ZHOUXiaohua, FENGYuchen, HUXuchu,et al.Transformer fault diagnosis based on SVM optimized by bald eagle search algorithm[J]. Southern Power System Technology,2023,17(6):99-106, 116.
[17]
WUY H, SUNX B, ZHANGY,et al.A power transformer fault diagnosis method-based hybrid improved seagull optimization algorithm and support vector machine[J].IEEE Access,2021,10:17268-17286.
[18]
LIJ F,HAI C, FENGZ,et al.A transformer fault diagnosis method based on parameters optimization of hybrid kernel extreme learning machine[J].IEEE Access,2021,9:126891-126902.
[19]
DINGC, DINGQ C, FENGL,et al.Prediction model of dissolved gas in transformer oil based on VMD-SMA-LSSVM[J].IEEJ Transactions on Electrical and Electronic Engineering,2022,17(10):1432-1440.
LIYunhao, XIANRichang, ZHANGHaiqiang,et al.Fault diagnosis for power transformers based on improved grey wolf algorithm coupled with least squares support vector machine[J].Power System Technology,2023,47(4):1470-1478.
[22]
LIUY, WUG N, GUOY J,et al.A new noncontact detection method for assessing the aging state of composite insulators[J].IEEE Transactions on Industrial Informatics,2024,20(4):6802-6813.
ZHOUXuan, WUWeili.Transformer fault diagnosis method based on improved SMOTE unbalanced sample processing and IHPO-DBN[J]. Power System Protection and Control,2024,52(11):21-30.
[25]
SHANGH K, LIUZ D, WEIY L,et al.A novel fault diagnosis method for a power transformer based on multi-scale approximate entropy and optimized convolutional networks[J].Entropy,2024, 26(3):186.
XUHuiling, LIUSheng, LIAndong.A heterogeneous guided whale optimization algorithm based on forward-reverse local exploitation and the golden sine algorithm[J].Computer Engineering & Science,2024,46(6):1128-1140.
YUShuhao, SUShoubao.Research and application of chaotic glowworm swarm optimization algorithm[J].Journal of Frontiers of Computer Science and Technology,2014,8(3):352-358.
LIANGChenglong, CHENZhihuan.Improved salp swarm algorithm based on hybrid strategy[J].Control and Decision,2024,39(8): 2541-2550.
[36]
LIK W, LIS H, HUANGZ C,et al.Grey Wolf Optimization algorithm based on Cauchy-Gaussian mutation and improved search strategy[J].Scientific Reports,2022,12:18961.
FENGXin, YANGXiong, ZENGYuhao.Robot path planning based on IGJO-DWA algorithm in multiple scenarios[J].Transducer and Microsystem Technologies,2024,43(11):131-134, 138.
LIHai, TIANZhong, QIANJun.Hydrometeor classification for radar based on ECOC-balanced random forest[J].Systems Engineering and Electronics,2024,46(5):1599-1606.