摘要
变压器是电力供应的重要组成部分,油中溶解气体测试方法通常用于配电变压器的质量检测,从而诊断其故障。文中提出了一种改进的优化算法,利用溶解气体分析(disslved gases analysis,DGA)对电力变压器质量检测从而进行故障预测。首先,从各种来源收集来自变压器DGA的数据,并选择最佳混合特征集作为模型的输入。其次,对于特征选择,采用核主成分分析(kernel principal component analysis,KPCA),为了优化最小二乘支持向量机(least squares support vector machine,LSSVM)模型并构建质量检测诊断模型,提出了一种结合混沌阿基米德优化算法(chaotic archimedes optimization algorithm,CAOA)和LSSVM的混合优化算法。最后,将所提CAOA、其他模型及建议的模型进行对比,测试结果表明所提CAOA在检测变压器故障方面比其他传统方法更准确。
Transformer is an important component of power supply,and the testing method of dissolved gas in oil is usually used for quality inspection of distribution transformer so to diagnose its faults.In this paper,an improved optimization algorithm is proposed and the dissolved gases analysis(DGA)is used for quality detection of power transformer and thus for fault prediction.Firstly,the DGA data from of transformer is collected from various sources and the best mixed feature set is selected as the input to the model.For feature selection,the kernel principal component analysis(KPCA)is used.For optimizing the least squares support vector machine(LSSVM)model and constructing a quality detection diagnostic model,a hybrid optimization algorithm combining chaotic archimedes optimization algorithm(CAOA)and LSSVM is proposed,and the proposed CAOA and other models as well as the suggested model are compared.The test results show that the proposed CAOA is more accurate than other traditional methods in detecting the fault of transformer.
作者
田霖
张达
刘振
吴宏波
鄢晶
TIAN Lin;ZHANG Da;LIU Zhen;WU Hongbo;YAN Jing(State Grid Hebei Electric Power Co.,Ltd.,Shijiazhuang 050000,China;State Grid Hebei Marketing Service Center,Shijiazhuang 050000,China;Wuhan University,Wuhan 430000,China)
出处
《高压电器》
北大核心
2025年第6期131-137,共7页
High Voltage Apparatus
基金
国网河北省电力有限公司科技项目(KJ2021-018)。
关键词
优化算法
诊断
电力变压器
测试
检验
algorithm optimization
diagnosis
power transformer
testing
inspection
