考虑乘客群体分异性的城轨网络客流分布时空全景推演研究

城轨网络客流分布时空全景推演是面向智慧城轨需求管理实施主动调控组织的关键支撑。传统网络客流分布估计方法在时效性和准确性方面的不足，是新一代智慧型城轨建设的主要瓶颈。本研究利用人工智能等新兴信息技术，依托多个不同规模城市轨道交通系统的多源时空大数据，围绕“乘客群体辨识-时空轨迹重构-实时分布推演-综合实证评估”研究主线，对快速精确的线网客流分布时空全景推演方法展开研究。本研究重点聚焦以下内容：基于乘客时空出行活动特征研究典型乘客群体辨识方法，解析出行需求时空分布中的乘客群体分异性特征；探索能够刻画乘客复杂出行行为的出行路径推断方法，融合多源完备信息构建乘客-列车关联推算序列决策模型；剖析群体融合OD客流动态演化规律，通过准确捕捉OD客流和网络客流分布之间的映射关系建立客流空间传播快速估计方法。项目研究是人工智能等新兴信息技术在智慧城轨建设领域的有益探索，对推动大数据赋能交通发展具有积极作用。

The spatial-temporal panoramic deduction of passenger flow distribution on urban rail transit networks is critical to active control for smart urban rail transit management. Due to the lack of timeliness and accuracy of traditional network passenger flow distribution estimation methods, it is the main bottleneck to constructing the new generation of intelligent urban rail transit. Based on the emerging information technology such as artificial intelligence and the multi-source spatiotemporal big data collected from several urban rail transit systems with different scales, the proposed research focuses on passenger group identification, spatiotemporal trajectory reconstruction, real-time distribution deduction, and comprehensive empirical evaluation. This project aims to advance the method for fast and accurate spatiotemporal panoramic deduction of line network passenger flow distribution. More specifically, the objectives of this proposed project are as follows: (1) to study the identification method of typical passenger groups based on the characteristics of passengers’ spatiotemporal travel activities; (2) to analyze the group-stratified heterogeneity in the spatiotemporal evolution of travel demand; (3) to explore the passenger route inferring method which can describe the complex travel process of passengers; (4) to construct the sequence decision model for the association between passenger-train by fusing multi-source complete information; (5) to analyze the dynamic evolution law of group-convergence OD passenger flow, and establish the spatial propagation estimating method of network passenger flow by capturing the mapping relationship between dynamic OD flow and network passenger flow distribution. This research is an exploration of emerging information technologies such as artificial intelligence in the field of smart urban rail transit construction, which plays a positive role in promoting the development of big data enabled transportation.