地铁自动检票机故障分析及可靠性研究  

Failure Analysis and Reliability Research of Metro Automatic Gate Machine

作　　者：徐余明[1] 黎家靖 张宁[2] 石先明[1] 刘利平[1] 胡祖翰[1] XU Yuming;LI Jiajing;ZHANG Ning;SHI Xianming;LIU Liping;HU Zuhan(China Railway SI YU AN Survey and Design Group Co.,Ltd,Wuhan 430063,China;Research Center of ITS,Southeast University,Nanjing 210018,China)

机构地区：[1]中铁第四勘察设计院集团有限公司,湖北武汉430063 [2]东南大学智能运输系统研究中心,江苏南京210018

出　　处：《大连交通大学学报》2023年第2期101-107,共7页Journal of Dalian Jiaotong University

基　　金：国家重点研发计划资助项目(2020YFB1600700、H202120807)。

摘　　要：考虑到地铁自动检票机组成复杂且故障形式多样,对其进行故障分析并以二参数威布尔分布为基础,提出一种基于混合威布尔分布的设备可靠性评估模型。为提高模型拟合精度,基于误差平方和最小思想构建非线性最小二乘参数优化估计模型并使用粒子群算法(PSO)进行最优参数求解。以南京地铁油坊桥车站自动检票机实际故障数据为例,进行实例验证。结果表明,基于PSO算法的混合威布尔分布可靠性评估模型优于传统单威布尔分布,其均方根误差、平均绝对百分比误差、皮尔逊相关系数均为最优。It is an important prerequisite of analyzing equipment reliability for an appropriate reliability evaluation model.The fault of metro automatic gate machine with its the complex composition and various fault forms are analyzed,and an equipment reliability evaluation model is proposed based on mixed weibull distribution.A nonlinear least squares parameter optimization estimation model is constructed to improve the model fitting accuracy based on the idea of minimum sum of errors squares and using particle swarm optimization to solve the optimal parameters.The AGM actual fault data of Youfangqiao station of Nanjing Metro is adopted.The results show that the reliability evaluation model based on mixed weibull distribution and PSO algorithm has obvious advantages than traditional single weibull distribution and mixed weibull distribution.It has the best grades of root mean square error,average absolute percentage error and Pearson correlation coefficient.

关 键 词：自动检票机 故障分析 可靠性评估 混合威布尔分布 参数优化估计 粒子群算法 

分 类 号：U293.22[交通运输工程—交通运输规划与管理]