时空可达性及其在交通数据分析中的一些应用研究
Space-Time Accessibility and Some Applications in Traffic Data Analysis

在评价交通服务系统中，时空可达性是重要考虑因素，基于时空可达性的交通数据分析，能为决策者在路段建设及公共交通路线设计方面提供合理意见，有助于提高出行者到活动地点的便利程度。本文首先针对城市基础交通网络，从时空棱柱理论框架中构建时空网络，考虑个体在起点及活动地点所受的时间约束及费用约束，构建基于时空可达性交通网络模型，并将该模型进行改进，应用于城市公共交通系统中。并针对模型特点，设计深度遍历算法，得到时空网络中所有的可行路径，从中选出最优解。为更好地适应大规模路网，又设计拉格朗日松弛算法，通过其耦合约束和难约束，将原问题分解为最短路问题及背包问题，并更新拉格朗日乘子协调各子问题。最后，以实际问题为例，验证了模型和算法的合理性。研究结果表明本文提出的模型能有效解决以个体出行需求为目标的交通网络设计问题。
In evaluating transportation system, space-time accessibility is an important factor. The transportation data analysis based on space-time accessibility can provide reasonable suggestions for policy-makers in road construction and public transportation route design, which will make it more convenient for travelers. We firstly focus on urban transportation network and build space-time network based on space-time prisms theories by considering the time constraint and cost constraint of traveler heading for destination, and finally building transportation network design models based on space-time-accessibility. Then the improved models are proposed and used to urban public transportation systems. Based on the models’ characteristics, we introduce the depth-first-search algorithm which can calculate all the possible routes in space-time networks and choose the best one from them. Largrangian relaxation algorithm is then designed to better adapt to large-scale road networks. And through the coupling constraints and hard constraints in the relaxation model, the original problem can be decomposed into the shortest-path problem and the knapsack problem, both of them can be solved by updating Largrangian multipliers. Finally, the rationality of the model and algorithm is verified by a practical example. The experimental results show that the proposed model can effectively solve the traffic network design problem aiming at individual travel demand.