摘要
为了改善变压器故障类型识别效果,针对变压器故障识别问题,本文构建了一种结合了最小二乘支持向量机(LSSVM)的强大预测能力与白冠鸡算法(COA)的优化特性的独特模型。通过应用COA算法对LSSVM的参数进行优化,开发了一种高效的变压器故障识别方案。用COA算法确定了LSSVM参数最优值,解决了LSSVM参数选择问题,在此基础上构建了COA-LSSVM分类模型。算例分析结果表明,COA-LSSVM模型的分类精度高达98.57%,显著高于其他两种对比模型,针对变压器故障识别能力,进行COA-LSSVM模型有效性和可行性的识别。
In order to improve the recognition effect of transformer fault types,this paper designs a transformer fault recognition model based on the cool optimization algorithm(COA)optimized least squares support vector machine(LSSVM).The COA algorithm was used to determine the optimal value of LSSVM parameters,solving the problem of LSSVM parameter selection.Based on this,a COA-LSSVM classification model was constructed.The analysis results of the example show that the classification accuracy of the COA-LSSVM model is as high as 98.57%,significantly higher than the other two comparison models,verifying the feasibility and effectiveness of the COA-LSSVM model in transformer fault recognition.
作者
彭克民
PENG Kemin(Guangdong Lucheng Engineering Consulting Co.,LTD.,Zhongshan 528400,Guangdong,China)
出处
《电气传动自动化》
2024年第6期58-62,共5页
Electric Drive Automation
关键词
变压器
故障识别
最小二乘支持向量机
白冠鸡算法
Transformer
Fault identification
Least squares support vector machine
Coot optimization algorithm
