交通运输系统工程与信息 ›› 2015, Vol. 15 ›› Issue (4): 181-186.

• 系统工程理论与方法 • 上一篇    下一篇

不确定需求下轨道交通网络的鲁棒性优化

孙杨*1,刘星材2,景春光1,宋瑞2,聂婷婷1   

  1. 1. 交通运输部科学研究院,北京100029;2. 北京交通大学交通运输学院,北京100044
  • 收稿日期:2015-03-16 修回日期:2015-04-19 出版日期:2015-08-25 发布日期:2015-08-25
  • 作者简介:孙杨(1983-),男,黑龙江哈尔滨人,博士.
  • 基金资助:

    国家重点基础研究发展计划资助课题(2012CB725403)

Robust Optimization for Rail Transit Network under Uncertainty Demand

SUN Yang1,LIU Xing-cai2,JING Chun-guang1,SONG Rui2,NIE Ting-ting1   

  1. 1. China Academy of Transportation Sciences, Beijing 100029, China; 2. School of Traffic and Transportation, Beijing Jiaotong University, Beijing 100044, China
  • Received:2015-03-16 Revised:2015-04-19 Online:2015-08-25 Published:2015-08-25

摘要:

轨道交通网络设计是轨道交通规划的重点,本文研究不确定需求下轨道交通网络设计的鲁棒性优化问题.提出了不确定需求下轨道交通网络鲁棒性的概念.针对不确定需求可以被预测和不可以被预测的两种情况,分别建立了scenario 模型、minmax模型,这两个模型在优化目标中均综合考虑了最小化轨道交通线路总长度、最小化乘客总出行距离、最小化乘客总换乘次数,并基于遗传算法给出了这两个模型的求解算法.scenario 模型权衡网络的服务水平与网络对于不确定需求的抗干扰能力;minmax 模型侧重于网络在最坏情况下仍然能够保持较好的服务性能.最后,给出算例,验证了提出模型与算法的有效性.

关键词: 城市交通, 网络设计, 遗传算法, 轨道交通网络, 鲁棒性优化

Abstract:

Rail transit network design is the key of rail transit planning. The rail transit network design problem under uncertainty demand is studied. The robustness of rail network under uncertainty demand is defined. According to two assumptions respectively that uncertainty demand could be predicted, could not be predicted, scenario model and minmax model are proposed, whose objectives are to minimize total length of rail transit lines, minimize total travel distances of passenger, and minimize total transfer times of passenger. Two methods based on Genetic Algorithm are developed to solve the two proposed models respectively. Scenario model makes a tradeoff between the network performance and the anti-jamming ability to uncertainty demand. Minmax model makes the network perform better for the worst-case scenario. At last, a numerical example is given. The validity of the proposed models and algorithms are demonstrated by numerical results.

Key words: urban traffic, network design, genetic algorithm, rail transit network, robust optimization

中图分类号: