Development of Micro-simulation Model for Urban Goods Movements

因此，本研究的目的是建立一个关注货运行为层面的模型。该模型考虑了货运代理的个体行为及其在供应链中的相互作用，能够真实地反映货运的运动机制。所提出的模型是对传统的四步法的修正，因为它单独考虑了每个货运代理的行为。这项研究利用微观模拟作为建模的一种方法，考虑了每个公司的单独行为。模型结构由商品生产与消费、商品配送、商品流向车流转化、交通分配四个阶段组成，模型考虑了货运代理人的个体行为，既适用于静态货运系统，也适用于动态货运系统。静态模型得出了每个起点和目的地的平均卡车出行次数，而动态模型则考虑了货运运动的动态过程，该过程随着消费趋势的变化而变化。此外，动态多智能体货运模型还可以反映经济中的几种现象(如牛鞭效应)和货运需求的季节变化。应用该模型对东京都市区的城市货运进行了分析。本研究讨论了模型开发和验证所涉及的问题，包括概念框架、数学公式、估计结果和验证结果。此外，该模型还通过应用于评估两个拟议的场景来测试其能力。结果表明，该模型适用于捕捉货运代理对出行模式变化的响应行为。

The objective of this study is therefore to develop a model that focuses on the behavioral level of freight transportation. The proposed model considers the individual behavior of freight agents and their interactions in supply chain resulting in the model's capability to reflect the real mechanism of freight movement. The proposed model is a modification of the traditional four-step approach in that it considers the behavior of each freight agent individually. The study utilizes micro-simulation as an approach to the modeling, considering the behavior of each firm individually. The model structure consists of four stages : commodity production and consumption, commodity distribution, conversion of commodity flows to truck flows, and traffic assignment.Since the model considers the individual behavior of freight agent, the model can be applied to both static and dynamic of freight transportation system. Static model results the average number of truck trips of each origin and destination; on the other hand, dynamic model concerned the dynamic process of freight movement that varies according to trend in consumption. In addition, several phenomena in economic (such as bull-whip effect) and seasonal variation in freight traffic demand can be captured by the dynamic multi-agent freight model. The proposed model has been applied to analyze the urban freight transportation in the Tokyo Metropolitan Area. This study discusses the issues involved in model development and validation including conceptual framework, mathematical formulations, estimated results, and validation results. In addition, the model was also tested for its capability by applying to evaluate two proposed scenarios. The results show that the model is applicable to capture the responsive behaviors of freight agents that accompany the travel pattern changes.