基于多智能体的停车选择行为仿真研究

城市中心区停车供需矛盾日益突出，研究停车选择行为并评价停车管理手段具有重要意义．基于前景理论建立驾驶员停车选择行为，构造战略层智能体及战术层智能体，利用多智能体仿真软件Starlogo对停车选择行为进行仿真实验，通过设定不同的情景评价停车管理政策和措施的效果．仿真实验的结果表明在调查区域内调整停车费率可以优化停车系统的资源利用水平，而通过设置路内停车场可以进一步缓解停车供需矛盾．

Simulation Study of Parking Choice Behavior Based on Multi-agent

The contradiction between parking supply and demand in central business districts becomes increasingly prominent, research on parking choice behavior and the evaluation of parking manage- ment is important. Base on prospect theory, a new parking choice model is proposed. As the parking choice model is constructed, the tactical layer agent and strategic layer agent are established. A simu- lation of parking choice behavior is implemented in Starlogo. The effect of implementing parking poli- cies and measures is evaluated in various scenarios. The results of the simulation indicate that the parking system＇s level of resource utilization can be optimized by adjusting parking rates, and the con- tradiction of parking supply and demand can be mitigated by the setting of curb parking.