高铁限速运行时分的小数据驱动建模与仿真  

Small-data Driven Modeling and Simulation of High-speed Train Running Time Under Limited Speeds

作　　者：徐鹏 冯国奇[1] 代学武 崔东亮 魏七龙 李宝旭 李建明 Xu Peng;Feng Guoqi;Dai Xuewu;Cui Dongliang;Wei Qilong;Li Baoxu;Li Jianming(School of business administration,Northeastern University,Shenyang 110169,China;Laboratory of Synthetical Automation for Process Industries,Northeastern University,Shenyang 110819,China;Signal&Communication Research Institute,China Academy of Railway Sciences,Beijing 100081,China;China Railway Shenyang Group Co.,Ltd,Shenyang 110000,China)

机构地区：[1]东北大学工商管理学院,辽宁沈阳110169 [2]东北大学流程工业综合自动化国家重点实验室,辽宁沈阳110819 [3]中国铁道科学研究院通信信号研究所,北京100081 [4]中国铁路沈阳局集团有限公司,辽宁沈阳110000

出　　处：《系统仿真学报》2021年第8期1892-1904,共13页Journal of System Simulation

基　　金：国家自然科学基金(61790574,U1834211,61773111);辽宁省自然科学基金(2020-MS-093);中央高校基本科研业务专项资金(N2008001);国家铁路智能运输系统工程技术研究中心开放课题(RITS2018KF03)。

摘　　要：为实现高铁列车群调度优化算法的数据支撑及可行性评估,提出一种机理模型和小数据驱动相结合的建模方法。构建限速条件下列车运行的参数化分段拟合模型,减少待辨识参数的个数;设计粒子群和最小二乘法相结合的参数辨识算法避免拟合不当的情况;得到临时限速下列车运行的"位置-时间-速度"特征曲面。在仿真平台上对所建立模型进行试验,对各种限速条件下列车运行时分进行准确快速的仿真,验证了模型及算法的有效性和可用性。In order to provide data support and evaluate the feasibility of high-speed train group scheduling optimization algorithm,a method of combining the mechanism model with the small-data drive is proposed.The train segment fitting model under speed limit is constructed and parameterized to reduce the number of parameters to be identified:In order to avoid the improper fitting,a parameter fitting algorithm based on the particle swarm optimization and the least square is proposed."Location-Time-Speed"model for temporary speed limits together are proposed.The model is demonstrated on the simulation platform,and the train running time is simulated accurately and quickly under various speed limit conditions,and the effectiveness and availability of the model and method are verified.

关 键 词：列车运行时分 临时限速 小数据驱动 参数辨识 仿真平台 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]