基于ICA的轨道交通系统数据模拟方法  

Data simulation method of rail transit system based on ICA

作　　者：杨基宏 陈浩林 吴海平[3] 曹子杰 段铸 台永丰 师小龙 李燕飞[4] YANG Jihong;CHEN Haolin;WU Haiping;CAO Zijie;DUAN Zhu;TAI Yongfeng;SHI Xiaolong;LI Yanfei(CRRC Qingdao Sifang Locomotive Rolling Stock Co.,Ltd.,Qingdao 266111,China;School of Mathematics and Statistics,Central South University,Changsha 410083,China;School of Traffic and Transportation Engineering,Central South University,Changsha 410075,China;College of Mechanical and Electrical Engineeri ng,Hunan Agricultural University,Changsha 410128,China)

机构地区：[1]中车青岛四方机车车辆股份有限公司,山东青岛266111 [2]中南大学数学与统计学院,湖南长沙410083 [3]中南大学交通运输工程学院,湖南长沙410075 [4]湖南农业大学机电工程学院,湖南长沙410128

出　　处：《中国高新科技》2022年第6期67-68,共2页

基　　金：国家重点研发计划课题(2017YFB1201101)。

摘　　要：轨道交通系统是一个复杂巨系统,成本相关数据难以获取的问题阻碍着轨道交通系统的发展。文章通过独立成分分析(Independent Component Analysis,ICA)对轨道交通系统的成本影响要素进行去关联分析,利用核密度估计方法求得去关联后各成分的概率密度函数,最后利用概率密度函数随机采样,生成轨道交通系统模拟数据。通过案例分析,基于ICA的轨道交通系统模拟数据中,成本影响要素之间的皮尔逊相关系数与原始数据之差最大不超过6%,在误差允许的范围内,模拟数据与原始数据具有同等分布。The rail transit system is a complex giant system,and the difficulty of obtaining cost-related data hinders the development of the rail transit system.In this paper,the independent component analysis(ICA)is used to analyze the cost-influencing factors of the rail transit system.The kernel density estimation method is used to obtain the probability density function of each component after de-correlation,and finally the probability density function is used to randomly sample.Generate rail transit system simulation data.Through case analysis,in the simulation data of rail transit system based on ICA,the difference between the Pearson correlation coefficient between cost-influencing factors and the original data is no more than 6%.Within the allowable error range,the simulated data and the original data have the same distribution.

关 键 词：轨道交通系统 ICA 模拟数据 

分 类 号：U239[交通运输工程—道路与铁道工程]