基于多源大数据的城市交通微观-宏观行为规律研究

Understanding the spatial-temporal distribution of the urban travel activities and their evolution are of great importance for relieving traffic congestion and improving traffic efficiency of urban traffic system. This project investigates the urban travel behaviors from an interdisciplinary point, which employs the theories and methods of traffic and transportation science, geographical information science, systems science and complexity, mathematics and physics. Based on multi-source individual activity big data (such as mobile phone data, online individual activities’ data etc.) and urban geographical information, this project carries out a series of studies on urban travel behavior. The studies include: (1) to unveil the characteristics of individual behavior, build probabilistic models for individual travel trails, and implement algorithms for classification of individual trails; (2) to understand the spatial-temporal distribution and the evolution of urban traffic activities, build type-based trail-integration methods, and get type-based traffic flows; (3) to identify the bottle-necks of urban traffic system, build spatial-correlation models of type-based traffic flows, and discuss the microscopic behavior mechanism of the macroscopic traffic patterns. The results of this project will provide the theoretical foundations and big data modelling methods and applications for relieving urban traffic congestion, and deepening the understanding of the mechanism of micro-macro urban traffic behavior, and improving the level of urban traffic administration and decision.

深入理解城市交通出行活动的时空分布及其演化规律，对于缓解城市交通拥堵，提高城市交通系统的运行效率具有重要的理论和现实意义。本项目综合运用交通科学、地理信息科学、系统与复杂性科学、数学和物理学的理论和方法，基于多源个体活动大数据，结合城市地理信息，（1）建立个体出行轨迹概率分析模型，揭示城市交通个体出行行为特征；（2）实现个体出行行为类型识别算法，研究个体出行轨迹数据的交通流分类集成方法；（3）建立不同类型交通流的时空相关性计算模型，探讨城市宏观出行模式的微观形成机制，揭示不同出行类型交通流在城市交通拥堵瓶颈形成中的作用。本项目将加深人们对城市交通出行微观-宏观行为规律的理解，为缓解城市交通拥堵，改善城市交通运行状况，提高城市综合交通管理与决策水平，提供理论依据和大数据建模与应用方法支持。

深入理解城市交通出行活动的时空分布及其演化规律，对于缓解城市交通拥堵，提高城市交通系统的运行效率具有重要的理论和现实意义。本项目综合运用交通科学、地理信息科学、系统与复杂性科学、数学和物理学的理论和方法，基于多源个体活动大数据，结合城市地理信息，（1）建立个体出行轨迹概率分析模型，揭示城市交通个体出行行为特征；（2）实现个体出行行为类型识别算法，研究个体出行轨迹数据的交通流分类集成方法；（3）建立不同类型交通流的时空相关性计算模型，探讨城市宏观出行模式的微观形成机制，揭示不同出行类型交通流在城市交通拥堵瓶颈形成中的作用。本项目将加深人们对城市交通出行微观-宏观行为规律的理解，为缓解城市交通拥堵，改善城市交通运行状况，提高城市综合交通管理与决策水平，提供理论依据和大数据建模与应用方法支持。