Abstract:

Various parking Apps have been introduced into our daily lives, which make real-time parking lot’s information become easier to be accessed. Providing real- time parking lot’s information also is a new way of dealing with parking problems. However, there has little theoretical research focusing on the mechanism of their impacts on the traffic system, as well as the quantitative analysis on them. In this paper, we take two kinds of realtime parking lot’s information as the subjects to study their impacts on the traffic system, which include real-time remained parking spaces and parking guidance information. Based on theories of point- queue model, learning behavior and Logit-based discrete traffic choice model, a simulation model for traffic mode choice is developed in the case of providing two kinds of real-time parking lot’s information. After that, three scenarios are designed and tested by simulation experiments. Then, the mechanism of their impacts is presented by analyzing simulation results. Finally, sensitivity analyses on serval parameters are carried out with the aim of finding effective ways to improve the efficiency of the traffic system.

Key words: intelligent transportation, simulation model, learning behavior, traffic mode choice, point-queue model

摘要：

停车App在交通出行中的广泛应用，使实时获取各类停车场信息更加便捷.然而，实时停车场信息在优化交通系统方面的作用机理尚无理论基础，其应用效果也缺乏量化分析. 本文将实时剩余停车位信息和停车诱导信息作为研究对象，基于点队列模型、学习理论和Logit选择模型，建立了实时停车场信息提供下的出行方式选择行为的仿真模型.然后，设计3种出行情景并结合仿真实验结果，对实时停车场信息提供影响出行者出行行为的机理进行分析，从而评估其对交通系统的优化效果.最后，基于对2类主要参数(学习因子和扩容因子)的敏感性分析，进一步讨论了提升实时停车场信息提供对交通系统优化作用的可行方式.

关键词: 智能交通, 仿真模型, 学习行为理论, 出行方式选择, 点队列模型