摘要
目前我国高速铁路在CTCS2/3级运行条件下,基本达到3min追踪间隔时间,为进一步缩短列车追踪间隔时间,基于线路坡度参数,对列车制动距离进行多阶段划分,通过建立列车混合优化模型和基于动态规划的多阶段决策模型,采用COADP算法(自适应动态规划协同优化算法)优化列车制动距离,得到各阶段的最优决策序列,以及列车制动距离的最优目标函数,对高速铁路列车追踪间隔时间进行优化,并对ADP算法(自适应动态规划算法)和COPADP算法的优化结果进行了仿真对比,结果表明COADP算法不仅有效避免了ADP算法的"维数灾"问题,而且对追踪间隔时间的优化作用更为明显,提升了高速铁路的通过能力。
At present,under the condition of CTCS2/3 operation,the tracking interval of high-speed railway in our country basically reaches 3 min.In order to further shorten the tracking interval time,the train braking distance is divided into several stages.By establishing a hybrid optimization model and a multi-stage decision model based on dynamic programming,COADP algorithm(Cooperative Optimization algorithm for Adaptive Dynamic Programming),based on the line slope parameter,is used to optimize the train braking distance,then the optimal decision sequence of each stage and the optimal objective function of train braking distance are obtained and the track interval time of high-speed railway train is optimized.By comparing the simulations of ADP(Adaptive Dynamic Programming)algorithm and COPADP algorithm,the results show that COADP algorithm not only avoids the"dimension disaster"problem of ADP algorithm effectively,but also optimizes the tracking interval time more obviously,which improves the capacity of high-speed railway.
作者
郑云水
高生霖
束展逸
ZHENG Yun-shui;GAO Sheng-lin;SHU Zhan-yi(Rail Transit Electrical Automation Engineering Laboratory,Lanzhou Jiaotong University,Lanzhou 730070,China;School of Automation & Electrical Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China)
出处
《测控技术》
2019年第6期120-125,共6页
Measurement & Control Technology
基金
国家自然科学基金(61763023)
关键词
高速铁路
制动距离
追踪间隔时间
动态规划
high-speed railway
braking distance
tracking interval time
dynamic programming
