Development of a Multi-scale, Integrated Microsimulation Model of Urban Systems

NSERC Discovery Grant建议支持达尔豪西大学的一项长期研究计划，该计划旨在开发一个集成的城市建模系统，为土地利用和交通模型的多尺度集成提供理论和经验基础。长期的家庭位置选择会影响短期旅行选择，反之亦然。有必要更好地了解这些行为联系，并为运输工程师开发决策支持工具，以促进土地利用和运输系统的综合规划。拟议的研究将集中在两个主题领域：1)开发微观计量经济模型，探索家庭决策过程如何与交通系统提供的机会相互作用；2)建立基于代理的微观模拟工具，用于预测和政策情景测试。通过利用CFI赞助的研究基础设施，拟议的计划将开发一个微观模拟工具包，模拟城市系统的演变，从住宅选址到日常出行的车辆分配，从生命阶段的过渡到活动议程的调整，从每年的位置模式到每秒的交通模拟。具体地说，拟议的研究将提供先进的分析方法来建立联系模型，例如车队纵向所有权与每日旅行车辆分配之间的联系。基于微观模型中确定的关系，这项研究将开发一个原型模拟工具，使用加拿大哈利法克斯100%的合成人口来预测30年(2006-2036)期间的家庭位置、车辆利用和出行模式。拟议的大规模微观模拟工具展示的粒度和系统交互作用将使测试较新的政策干预措施成为可能，例如，碳定价战略对住宅地点选择或日常旅行议程的影响。拟议的研究计划将在一个研究实验室培养两名博士生、两名硕士和一名博士后研究员，该实验室吸引土木工程和规划学科的研究生。协作、多学科的研究环境将促进开发综合城市模型的创新方法，并将其应用于评估加拿大社区指导可持续城市发展所需的复杂交通和土地使用政策。

This NSERC Discovery Grant proposal supports a long-term research program at Dalhousie University in developing an integrated urban modelling system that will provide a theoretical and empirical foundation for multi-scale integration of land use and transport models. Long-term household location choices influence short-term travel choices and vice versa. There is a need to better understand these behavioural linkages and develop decision-support tools for transportation engineers to facilitate integrated planning for land use and transportation system. The proposed research will focus on two thematic areas: 1) developing micro-level econometric models that explore how household decision processes interact with the opportunities that the transportation system offers, and 2) building agent-based, microsimulation tools for forecasting and policy scenario testing. By utilizing CFI-sponsored research infrastructure, the proposed program will develop a microsimulation toolkit that simulates the evolution of urban systems, from residential location choice to vehicle allocation for daily trips, from life-stage transitions to adjustments in activity agenda and from year-by-year location patterns to second-by-second traffic simulation. Specifically, the proposed research will offer advanced analytical methods to develop link models, for instance linkages between longitudinal vehicle fleet ownership and vehicle allocation for daily travel tours. Based on the relationships identified in micro-models, the research will develop a prototype simulation tool using a 100% synthetic population of Halifax, Canada to predict household location, vehicle utilization and travel patterns for a thirty-year period (2006-2036). The granularity and systems interactions that the proposed large-scale microsimulation tool exhibits would enable testing of newer policy interventions, for instance the impacts of carbon pricing strategies on residential location choices or daily travel agendas. The proposed research program will train two PhD students, two Masters students and a post-doctoral fellow in a research laboratory that draws graduate students from the civil engineering and planning disciplines. The collaborative, multidisciplinary research environment will foster innovative approaches to the development of integrated urban models and their application for evaluating complex transport and land use policies that Canadian communities need to guide sustainable urban development.