摘要
地铁列车发车间隔优化是提高地铁服务水平的重要手段.文中通过分析地铁客流的时间分布特性,利用AP聚类算法对地铁客流数据进行聚类分析,将客流数据分为若干个子类;综合考虑列车满载程度及乘客舒适度,建立多目标发车间隔时间优化模型;采用基于NSGA-II的多目标遗传算法求解该问题的Pareto解集,综合考虑相邻时段的发车间隔的稳定性,确定各时间段的发车间隔.选取某地铁2号线为例,对该线路4类客流进行了列车发车间隔优化分析.算例分析表明,本模型在地铁列车行车间隔调度方面是合理可行的.
The optimization of trains headway time is an important strategy to improve the service level of subway. Through analysis the time distribution characteristics of subway passenger flow, the Affinity Propagation (AP) clustering algorithm is used to divide the passenger flow into several cluster subsets. Furthermore, a multi-objective optimization model of trains headway time is established considering the train full-load ratio and passenger comfort level. A Non-dominated Sorting Genetic Algorithm II (NSGA-II) is put forward to find this multi-objective model's Pareto solution set which can provide multiple options for operating managers. Then, the headway time is determined by considering the stability of the headway time. At last, Line No. 2 of a subway is chosen for optimization the trains headway time for four categories of passenger flow and the results illustrate the effectiveness of the optimization model and algorithm.
出处
《武汉理工大学学报(交通科学与工程版)》
2015年第6期1119-1124,共6页
Journal of Wuhan University of Technology(Transportation Science & Engineering)
基金
教育部人文社会科学研究项目(批准号:13XJC630017)
兰州交通大学青年基金项目(批准号:2014028)
甘肃省自然科学基金项目(批准号:148RJZA052)资助
关键词
地铁
发车间隔
多目标
遗传算法
subway
headway time
multi objective
genetic algorithm