城市轨道交通车站客流仿真中的事件驱动模型

Event-driven simulation model for passenger flow in urban mass transit station

利用多智能体(Multi-agent)仿真技术对轨道交通车站设计方案、客流组织方案进行评估是一种科学有效的方法.在对"客流事件"、"服务流程"及"乘客流"进行定义的基础上,利用三种"元事件"以组合的方式构建了七种"客流事件",从而实现了一种能够描述客流在车站内的复杂行为的统一模型.大量仿真实验表明,此模型不但能够显著降低建立仿真项目的工作量,并且更加真实地反映了乘客在车站内的行为习惯.此外,模型还具有良好的可扩展性,易推广到其它场景,如大型体育比赛(奥运会、世界杯)、大型展览会(世博会)等条件下的观众疏散、交通组织等领域.

It＇s an effective way to compare alternative designs of UTM Stations（Urban mass transit）, as well as to evaluate passenger-flow management with multi-agent simulation technique.The terms of ＂Passenger Flow＂,＂Passenger-flow Event＂,and ＂Service Process＂ are defined.Based on these definitions, a unified model,which could simulate complex behaviors of passengers,is developed with the idea of using 3 kinds of ＂Atomic Events＂ to build 7 kinds of ＂Passenger-flow Events＂,in the way of combination. Large number of simulation tests proved that：this model not only can reduce the workloads of designing simulation scenarios,but also simulate the behaviors and habits of passengers in a realistic way.Besides, characteristic by openness and expansibility,this model can be easily applied to other situations,such as egress analysis and flow management during Olympics and Expo,etc.