基于t-SNE和ECOC-ISSA-SVM的变压器故障诊断

打开文本图片集

中图分类号：TM407 文献标志码：A 文章编号：1008-0562（2025）05-0606-08

Transformer fault diagnosis based on t⋅ -SNEand ECOC-ISSA-SVM

LIUMeng'，ZHAO Chenxiao'， ZHU Qiaobo'，LILiang'，YAO Xu2 ，LI Xin* ， ZHAO Ming？ （1.State GridJibei Electric Power CompanyLimited，Beijing 10oo89，China;2.Beijing Kedong PowerControl Systems CompanyLimited，Beijing10oo89，China;3.Faculty ofElectrical and Control Engineering，Liaoning Technical University，Huludao 1251o5， China）

Abstract：To solve the problemsof insufficient hyperparameter optimization and multi-classfication 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 -SNE）. Error correction output codes （ECOC） are introduced，and the improved sparow search algorithm （ISSA） was combined with Chebyshev chaotic mappng and Cauchy-Gaussian variational strategy to optimize the hyperparameters of SVM and handle multi-clasification problems.The research results show that the diagnostic accuracy，recallrate，specificityandF1 value of the ECOC-ISSA-SVM（tSNE）model are 95.6% ， 97.8% ， 99.6% and 97.8% respectively. Compared with the traditional model，the improvement effect of each indexis significant.The diagnostic time isshortened to1l msand the diagnostic efficiency issignificantly improved.Theresearch conclusion provides technical support for the intelligent operation and maintenance of power equipment.

Keywords： fault diagnosis;transformer; dissolved gas inoil; support vector machine; sparrow search algorithm; t⋅ -SNE dimensionality reduction; error correction output codes

0引言电力变压器是电网核心设备，具有电压调控与能量传输功能，但长期运行易发生故障，对故障类型进行精准诊断是保障系统稳定运行的关键环节[。（剩余13093字）