基于多智能体的地铁列车运行调整方法  

Method for Subway Operation Adjustment Based on Multi-Agent

作　　者：路飞[1] 宋沐民[1] 田国会[1] 

机构地区：[1]山东大学控制科学与工程学院,山东济南250061

出　　处：《中国铁道科学》2007年第1期123-126,共4页China Railway Science

基　　金：国家自然科学基金资助项目(60104009)

摘　　要：针对移动闭塞条件下地铁列车运行的特点,将降低列车群的总晚点时间和提高相邻列车对客流的吸纳水平作为性能指标,建立地铁列车运行调整模型。应用Agent、多Agent协作技术,构建地铁列车Agent,设计地铁列车运行调整算法,给出基于规则的地铁列车运行调整控制策略。根据列车群晚点程度的不同,列车运行调整分为全局调整和局部调整。全局调整采用偏移计划运行图方法,局部调整采用滚动优化方法,通过多次调整达到恢复计划运行图的目的,并分别给出相应的调整方法和调整步骤。以北京1号线为背景进行仿真测试,仿真结果表明:基于多Agent协作技术的地铁列车运行调整方法是合理的,并具有动态的扩展性。After the analysis of movement of subway under moving block system, the problem of operation adjustment is studied. The model of the problem is constructed. The objectives are decreasing the total delay time and increasing the absorption to passengers of the successive trains. The technology of Agent and multi-Agent is adopted in the construction of train Agent and the design of operation adjustment. The control strategy based on rules for the operation adjustment is proposed. Different ways of global adjustment and local adjustment are taken according the delay degree of the trains. The objective of global adjustment is to decrease the amount of the trains that need to be adjusted through offsetting the planning table. The way of local adjustment is the roll optimization, its objective is to resume the planning table. The simulation taking the No I subway line in Beijing as background shows the Multi-Agent technology is suitable to the subway operation adjustment, it has the ability to extend dynamically.

关 键 词：列车运行调整 多Agent控制策略 移动闭塞 地铁列车 

分 类 号：U231.92[交通运输工程—道路与铁道工程] TP391[自动化与计算机技术—计算机应用技术]