DATA AUGMENTATION FOR AN INTEGRATED ACITIVITY-TRAVEL DEMAND AND MULTIMODAL TRAFFIC MICROSIMULATION MODEL

拟议的研究将开发一个活动-旅行需求和多模式交通流的综合微观模拟模型，该模型可以在空间分解的单个包裹级别（邮政编码）上对城市交通系统进行分分钟的模拟。这些模型需要大量的数据，这一问题将通过开发数据增强方法来解决，这些方法将使用基于智能手机的应用程序收集小规模但详细和准确的活动旅行数据，并将其与区域家庭旅行调查数据融合在一起。研究区域将是大多伦多和汉密尔顿地区（GTHA），因为该地区已经建立了一个家庭旅行调查数据收集计划和一个完整的多模式交通微观模拟模型原型。本研究的主要目标是：1)通过融合多数据源开发新的数据增强方法；2)基于融合数据估计和校准一套一致、协调的概率行为模型；3)开发一个集成的多式联运出行需求模拟系统；4)将集成的模拟系统应用于交通政策分析。第一个主要目标可以进一步阐述如下：(a)开发一种基于智能手机的自动活动-旅行数据收集工具，能够准确收集详细的时空运动信息。这将减少调查回复的负担，并创造一种真正普及的旅行数据收集方法。(b)制订数据扩充程序，以小规模详细数据补充区域家庭旅行调查数据。这将通过填补缺失和更多的分类信息来丰富大样本家庭旅行调查。第二个和第三个长期目标可以进一步阐述如下：(c)制定一个微观模拟框架，在考虑到一天内交通流量的时间变化的多式联运（公路和公共交通）运输网络内无缝地整合人们对活动-旅行的需求。(d)利用扩大的活动-旅行数据集估计和验证这一建模框架所有组成部分的计量经济模型。组件计量选择模型包括全面的活动生成和调度模型，其中包含基于动态包裹的活动位置选择模型和基于行程的模式选择模型。模式将以家庭为基础，并将考虑到家庭内部的资源和任务分配。所有这些目标都有短期和长期的影响。在短期内，将为GTHA开发一个微观模拟框架，并估计组件模型。从长远来看，建模框架将演变为区域尺度规划和政策分析的标准运输需求建模系统。区域尺度的动态旅客旅行需求模型在加拿大将是独一无二的，可以作为模板，为加拿大其他地区开发类似的模型。该模型将有助于评估加拿大各级政府正在考虑的各种交通政策和基础设施改善方案。这些措施包括道路收费、替代网络基础设施更新方案（例如，建设地铁/地铁救济线，用新的高速公路取代旧的高架高速公路或高速公路和交通要素的组合）以及城市地区的替代土地使用章程，以平衡城市系统并尽量减少社会排斥。该模型还将能够准确评估各种出行需求管理政策，例如拼车激励、灵活办公时间选择、远程办公选择等。

The proposed research will develop an integrated microsimulation model of activity-travel demand and multimodal traffic flow that can allow the minute-by-minute simulation of an urban transportation system at the individual parcel level (postal code) of spatial disaggregation. Such models have extensive data requirements, and this issue will be addressed by developing data augmentation methods that will collect small-scale but detailed and accurate activity-travel data using smartphone-based applications and fusing it with regional household travel survey data. The study area will be the Greater Toronto and Hamilton Area (GTHA) as the region already has an established household travel survey data collection program and a full prototype multimodal traffic microsimulation model. The major objectives of the research are to: 1)develop new methods for data augmentation by fusing multiple sources of data, 2)estimate and calibrate a consistent, coordinated set of probabilistic behavioural models based on the fused data, 3)develop an integrated multimodal travel demand simulation system, and 4)apply the integrated simulation system to the analysis of transportation policy. The first major objective can be further elaborated as follows: (a)Developing a smartphone-based automated activity-travel data collection tool that can accurately collect detailed spatio-temporal movement information. This will reduce the survey response burden and create a truly pervasive travel data collection method.(b)Developing a data augmentation process to supplement regional household travel survey data with small-scale detailed data. This will enrich the large-sample household travel survey by filling in missing and more disaggregate information.The second and third long-term objectives can be further elaborated as follows:(c)Developing a microsimulation framework that seamlessly integrates peoples' demand for activity-travel within a multimodal (road and public transport) transportation network considering temporal variations of traffic flow in the network within a day.(d)Estimating and validating econometric models for all components of this modelling framework by using augmented activity-travel datasets. Component econometric choice models include fully comprehensive activity generation and scheduling models containing dynamic parcel-based activity location choice models and tour-based mode choice models. Models will be household-based and will consider intra-household resource and task allocations.All of these objectives have both short- and long-term implications. In the short-term, a microsimulation framework will be developed and component models will be estimated for the GTHA. In the long-term, the modelling framework will evolve into a standard transportation demand modelling system for regional-scale planning and policy analysis. The regional-scale dynamic passenger travel demand model will be unique in Canada and can be used as a template to develop similar models for other regions in Canada. The model will be useful in evaluating a variety of transportation policies and infrastructure improvement options that are under consideration by various levels of government in Canada. These include road pricing, alternative network infrasture renewal options (e.g., building subway/metro relief lines, replacing old elevated expressways with new expressways or combinations of highways and transit elements) and alternative land use bylaws in urban areas to balance urban systems and minimize social exclusion. The model will also be able to accurately evaluate various travel demand management policies, e.g. carpooling incentives, flexible office hour option, telecommuting options, etc.