交通运输系统工程与信息

交通运输系统工程与信息 ›› 2025, Vol. 25 ›› Issue (6): 153-164.DOI: 10.16097/j.cnki.1009-6744.2025.06.014

• 系统工程理论与方法 • 上一篇    下一篇

网络整体效能导向的城轨线路运能协同优化配置

张龙豪，徐瑞华* ，单奕嘉，蓝阳泽   

  1. 同济大学，上海市轨道交通结构耐久与系统安全重点实验室，上海201804
  • 收稿日期:2025-08-06 修回日期:2025-10-10 接受日期:2025-10-16 出版日期:2025-12-25 发布日期:2025-12-24
  • 作者简介:张龙豪(1998—)，男，四川乐山人，博士生。
  • 基金资助:
    国家自然科学基金 (72171174)。

Network Performance-oriented Collaborative Optimization of Urban Rail Transit Lines' Capacity Allocation

ZHANG Longhao, XU Ruihua*, SHAN Yijia, LAN Yangze   

  1. Shanghai Key Laboratory of Rail Infrastructure Durability and System Safety, Tongji University, Shanghai 201804, China
  • Received:2025-08-06 Revised:2025-10-10 Accepted:2025-10-16 Online:2025-12-25 Published:2025-12-24
  • Supported by:
    National Natural Science Foundation of China (72171174)。

PDF

摘要： 在城市轨道交通网络运行图调整优化中，全网同步优化面临建模难度大和计算复杂性高的问题。本文基于按线调整的现实需求，提出一种线路间列车开行方案运能配置协同优化策略。首先，从客流转移竞合协同与网络效能贡献协同两方面，解析线路开行方案运能配置与网络整体效能的相互作用机理；其次，构建运输效率与服务水平双维度网络效能评价指标体系，依托自研多智能体线网仿真系统与理想解排序法实现效能量化评估；最后，提出按线编制模式下的线路开行方案运能配置协同优化方法及决策模型。基于某市真实路网数据，通过363组方案的仿真与评估验证策略有效性。结果表明：本文提出的以网络效能为优化目标，按照线路重要度顺序，采用“逐线调整+固定其他线路开行方案”迭代机制的运能逐线协同优化方法，在按线编制模式下能生成最优运能配置方案，其优化精度与全网启发式搜索等效，可作为全网同步优化的时间节省替代方法，使案例网络效能从初始0.519提升至0.588，增幅度达13.4%，为大规模城市轨道交通网络开行方案协同优化提供兼顾优化质量与计算效率的解决策略。

关键词: 城市交通, 网络运能配置, 多智能体仿真, 列车开行方案, 网络效能, 评价指标体系

Abstract: Network-wide synchronous optimization of urban rail transit timetables faces challenges in modeling complexity and high computational demand. To address the practical need for line-by-line adjustment, this paper proposes an inter-line collaborative optimization strategy for capacity allocation in line planning. First, the interaction mechanisms between capacity allocation and network performance are analyzed from the perspectives of passenger flow redistribution synergy and network performance contribution synergy. Next, a dual-dimensional evaluation index system (transport efficiency and service level) is established, with quantitative performance assessment achieved through a self-developed multi-agent network simulation system and the TOPSIS method. Finally, a collaborative optimization model and corresponding algorithms under the line-by-line planning framework are proposed. Validation using real network data involves 363 simulation experiments. Results demonstrate that the proposed method—which targets network performance, follows line importance order, and iteratively adjusts individual lines while keeping other line planning fixed—generates optimal capacity allocation, achieving accuracy comparable to that of network-wide synchronous heuristic search. As a time-efficient alternative to network synchronous optimization, the method improves network performance from 0.519 to 0.588, a 13.4% increase, thereby providing a solution balancing quality and efficiency in large-scale urban rail transit networks.

Key words: urban traffic, network capacity allocation, multi-agent simulation, line planning, network performance, evaluation index system

中图分类号: