城市轨道交通网络化列车开行方案优化方法  

Optimization Method for the Networking Train Operation Plan of Urban Rail Transit

作　　者：王永亮[1] 张星臣[1] 徐彬[1] 解晓灵[1] 

机构地区：[1]北京交通大学城市交通复杂系统理论与技术教育部重点实验室,北京100044

出　　处：《中国铁道科学》2012年第5期120-126,共7页China Railway Science

基　　金：国家自然科学基金资助项目(71131001;70971010)

摘　　要：通过引入备选集,建立双层规划模型,将网络化列车开行方案这一复杂的混合交通网络设计问题简化成0-1规划问题;采用混合遗传模拟退火算法对城市轨道交通网络化列车开行方案优化问题进行求解。上层模型以城市轨道交通网络化列车开行方案涉及的运营单位和乘客综合费用最小化为目标,以交路必须覆盖整个运营网络、限制区间交路数、满足发车间隔时间限制和区间客流量小于列车运力为约束条件;下层模型为用于城市轨道交通网络客流分配的随机用户均衡问题优化模型,它充分考虑了乘客选择出行时路径感知费用与实际费用之间的差异。算例结果表明,采用基于备选集的双层规划模型和混合遗传模拟退火算法对城市轨道交通网络化列车开行方案进行优化是可行和有效的,而且求解效率高。Networking train operation plan for urban rail transit, which was a complicated mixed traffic network design problem, was simplified into 0-1 programming problem by introducing potential line set and establishing bi-level programming model, and hybrid genetic simulated annealing algorithm was designed to find its solution. The objective of the upper-level model was to minimize the comprehensive cost of both operators and passengers involved in the networking train operation plan for urban rail transit. The constraints included routing covered whole operation network, routing number of limit section, satisfying the time limit for departure interval, and passenger flow in section less than the transport capacity of train. The lower-level model, namely, the random user equilibrium problem optimization model, which fully considering the difference between the perceived cost and actual cost of passenger in the choice of travel path, was used for passenger assignment in urban rail transit network. Numerical example results reveal that it is feasible and effective to adopt the bi-level programming model based on potential line set together with hybrid genetic simulated annealing algorithm to optimize the networking train operation plan of urban rail transit. Besides, the efficiency for working out solution is high.

关 键 词：列车开行方案 备选集 双层规划模型 遗传算法 模拟退火算法 城市轨道交通 

分 类 号：U239.5[交通运输工程—道路与铁道工程] U292.4