高铁牵引供电系统故障的Logistic关联分析  

Logistic correlation analysis of high speed railway traction power supply system fault

作　　者：程宏波 李新原[2] 王勋 辛建波[1] 

机构地区：[1]国网江西省电力公司电力科学研究院,江西南昌330096 [2]华东交通大学电气与自动化学院,江西南昌330013

出　　处：《铁道科学与工程学报》2017年第10期2065-2071,共7页Journal of Railway Science and Engineering

基　　金：江西省重点研发计划项目(20161BBH80033);江西省自然科学基金资助项目(20143ACB21019);江西省博士后科研择优资助项目(2016KY36)

摘　　要：为尽可能减少高铁牵引供电系统故障发生的次数、降低故障的严重程度,以牵引供电系统故障严重程度为因变量,从牵引供电系统内部结构和外部环境等因素中拟定候选自变量,采用后向逐步选择法判断候选自变量与因变量之间的相关性;利用累积Logistic模型建立高铁牵引供电系统故障严重程度与影响因素之间的关联分析模型,从成比例检验、拟合优度检验和预测准确度检验3个方面对模型进行检验。利用某供电段近几年的供电系统故障统计数据进行分析,分析结果表明:系统服役年限、维修次数、雷击、风速与牵引供电系统故障严重程度显著相关,分析结果可为减少牵引供电系统故障次数、降低牵引供电系统故障危害提供措施依据。In order to minimize the number of high-speed railway traction power supply system fault times and toreduce the severity of the fault, this paper used the traction power supply system fault severity as a dependentvariable. The candidate variables were formulated from traction power internal structure and external environment.The Backward Stepwise Selection was adapted to make sure the relativity between dependent variable andcandidate variables. By using the cumulative, the association model between high-speed rail traction power faultserious degree and effect factors was set up. The model was then tested in three aspects of the proportion test,goodness of fit test and prediction accuracy test. The analysis results from a power supply system power supplyfailure statistics recent years show that： service life of system, maintenance frequency, lightning strikes, and windspeed have significant relationship to power supply system fault serious degree. The results can provide basicmeasures to minimize traction power supply system fault times and reduce the damage of fault.

关 键 词：牵引供电系统 故障 LOGISTIC模型 关联分析 后向逐步选择法 

分 类 号：TM922.3[电气工程—电力电子与电力传动]