基于群智感知的城市出行需求匹配与共享交通调度优化研究

Building efficient and high-quality urban transportation systems to improve the satisfaction of citizens' travel demand has been the emphasis of the supply-side reform in urban transportation. The existing public transportation systems often suffer from overcrowding at peak times due to transportation resource limitation, while the low occupancy of the fast-growing private transportation modes (e.g. cars) have led to traffic congestion and air pollution. With the recent emergence of crowdsensing technologies, Shared Transportation Modes (STMs), such as ride sharing systems, have become increasing popular, providing new potentials for dynamic scheduling of transportation resources according to travel demand, which can eventually improve the satisfaction of travel demand. However, due to the fact that travel demand is highly dynamic and thus difficult to predict, satisfaction of travel demand is expensive to collect and thus difficult to evaluate, and STM scheduling is restricted by resources and thus difficult to optimize, the current travel demand matching and STM scheduling schemes are usually sub-optimal. In this project, based on crowdsensing theories and technologies, we propose to design effective STM resource scheduling and travel demand matching strategies, aiming at improving the satisfaction of travel demand. The main research issues include: 1) travel demand modeling and prediction based on heterogeneous contextual factors, 2) low-cost, multi-view travel demand satisfaction evaluation, and 3) STM scheduling and travel demand matching optimization under different resource constraints. This project will benefit STM resource scheduling and management, and provide decision strategies for the urban transportation system development.

建设高效优质的城市综合交通体系，提升大众出行需求满意度，是城市交通供给侧改革的重要目标。目前的公共交通系统运力有限，高峰期出行体验较差，而私人交通工具的低乘载率和过度增长则带来道路拥堵和环境污染。随着群智感知技术的发展，催生了基于互联网的共享交通模式，其运力资源灵活度高、可调度性强，可匹配出行需求以提升出行满意度。然而，由于出行需求高度动态难预测、获取出行需求满意度成本高难评价、共享交通匹配和调度受资源约束难优化，使得共享交通调度未能优化匹配出行需求。本项目基于群智感知理论和方法，以提升出行需求满意度为目标，探索出行需求匹配与共享交通运力调度的优化策略。研究内容包括：1）复杂时空情境下的出行需求建模和预测方法；2）多视角低成本的出行需求满意度评价方法；3）资源约束条件下的需求匹配和共享交通调度优化方法。本项目成果将提升共享交通的调度和管理水平，为城市综合交通体系的发展提供决策依据。

本项目基于人机协作的群智感知理论和方法，以提升城市出行需求满意度为目标，探索出行需求匹配与共享交通运力调度的优化策略，主要研究复杂时空情境下的出行需求建模和预测方法，多视角低成本的出行需求满意度评价方法，以及资源约束条件下的需求匹配和共享交通调度优化方法。研究内容包括：1）复杂时空情境下的出行需求建模和预测方法；2）多视角低成本的出行需求满意度评价方法；3）资源约束条件下的需求匹配和共享交通调度优化方法。本项目成果将提升共享交通的调度和管理水平，为城市综合交通体系的发展提供决策依据。本项目在论文发表、人才培养、国际交流合作方面取得了阶段性的成果，累计发表国际期刊会议论文15篇（包括3篇CCF-A类文章），申请发明专利10项，授权4项；研究成果获福建省科技进步一等奖1项（排序9）；协助指导博士研究生2名，培养硕士研究生18名；主办CCF智能感知与城市计算前沿论坛1次，参加国际会议或受邀访问国外大学3次。项目研究成果均已上线至空间群智感知应用展示平台，并获批产学研转化项目3项，累计项目经费40万元。.