Transformer Fault Position Recognition Based on Probability Support Vector Machine
Based on posterior probability of support vector machine,the analysis of dissolved gases in oil and the data of electrical tests are used co mprehensively to recognize intemal fault positions of the power transform er.This m ethod not only inherits the advantages of the support vector machine for sm all samples,strong generalization ability and so on,but al so provides the fault information of the power transformer by the form of probability.The output results not only provide the breaking down probability of the inner winding,tap changer and leads,iron core,structure and magnetic shielding body,insulating barrier and other parts,but also express the degree of credibility of conclusions.It adapts to the uncertainty characteristics of the fault positions.After analysis of examples,the validity of the model is verified.
transformer fault recognition fault position SVM posterior probability
Huang Song-bo Zhao Wei-min Zliang Tag Sima Li-ping Wang Bo Shu Nai-qiu
Foshan Power Supply Bureau,Guangdong Power Grid Corporation,Foshan,China School of Electrical Engineering,Wuhan University,Wuhan,China
国际会议
沈阳
英文
2020-2023
2012-09-26（万方平台首次上网日期，不代表论文的发表时间）