出租车空车搜索多主体仿真模型

Multi-Agent Simulation Model for Vacant Taxi Searching Behavior

出租车运营中空车与乘客的相互搜索是一个动态、随机的复杂交互过程,解析方法难以刻画这类行为.采用多主体技术设计城市出租车运营仿真架构,模拟出租车与乘客在路网上相互搜索的动态过程.出租车主体采用慎思结构实现,乘客决策依据随机效用理论,路网分解为道路、交叉口和站场三类Agent.模型能够描述出租车运营中常见的路边随机上下车行为,并考虑了出租车运营的动态交通环境,可以为解决我国出租车问题提供理论支持.

Vacant taxi behavior possesses the dynamic and randomness features.It is difficult to model vacant taxi behavior through mathematical methods.Therefore,we propose a simulation framework of taxi operation system based on multi-agent technology,so that the dynamic process taxis and passengers searching for each other can be recurred.Taxi agents are illustrated through deliberation method.Passengers＇ decision making is based on random utility theory.Road network is described with segment,intersection and station agents.The model can describe the commonly observed behavior that taxis may pick up passengers anywhere along roadsides,and the dynamic traffic environment of taxi operation system is taken into consideration.These are helpful for solving the problems of taxi industry.