摘要
为了能够解决传统人工经验方法准确率低、智能算法中存在的适用性不强等问题,针对支持向量机受到惩罚系数和核函数的敏感性,提出一种变压器诊断模型,该模型利用三对比值作为特征输入,利用灰狼优化算法优化支持向量机的惩罚系数和核函数,并利用优化后的参数模型去对变压器进行故障诊断。实验结果表明,支持向量机模型预测的准确率为90%,而基于灰狼优化算法优化支持向量机的变压器故障诊断模型的准确率达到100%,预测准确率比支持向量机的高,有效地实现了对变压器故障的诊断,给变压器故障诊断提供了一种可借鉴的方法。
In order to solve the problems of low accuracy of traditional manual experience method and weak applicability of intelligent algorithm,a transformer diagnosis model is proposed for the sensitivity of penalty coefficient and kernel function of support vector machine.The model uses three-pair ratios as feature input,and uses grey wolf optimization algorithm to optimize the penalty coefficient and kernel function of support vector machine,and uses the optimized parameter model to diagnose the fault of transformer.The experimental results show that the prediction accuracy of the support vector machine model is 90%,while the accuracy of the transformer fault diagnosis model based on the grey wolf optimization algorithm to optimize the support vector machine reaches 100%.The prediction accuracy is higher than that of the support vector machine,which effectively realizes the diagnosis of transformer faults and provides a reference method for transformer fault diagnosis.
作者
何闰丰
黄莺
HE Runfeng;HUANG Ying(Liuzhou Railway Vocational Technical College,Liuzhou 545616,China)
出处
《红水河》
2022年第1期84-88,共5页
Hongshui River
基金
2018年广西高校高水平创新团队及卓越学者计划资助项目。
关键词
变压器
故障诊断
三比值法
灰狼优化算法
支持向量机
transformer
fault diagnosis
three-ratio method
grey wolf optimization algorithm
support vector machine
