数据驱动的共享单车停放区规划方法研究

从宏观和微观两个角度研究共享单车停放区规划方法，宏观层面解决停放区选址定容问 题，微观层面进一步优化每个停放区内部的停车位布局问题。宏观层面利用DBSCAN对共享单 车停/取点进行聚类分析得到停/取需求点；然后考虑共享单车使用者路径选择、停/取点选择和停 放区服务能力约束，建立混合整数线性规划模型优化停放区选址及容量。微观层面充分考虑行 人、自行车与汽车之间的影响，其中行人和自行车采用改进社会力模型进行仿真，并考虑共享单 车使用者最优停放区选择；汽车采用智能驾驶模型进行仿真。停车位布局通过仿真通行效率及 混合交通流风险进行评价，并基于代理模型进行优化。该方法应用于北京市崇文门地铁站周边 共享单车停放区规划，验证了方法的有效性，该方法可提高混合交通流通行效率，降低交通风险。

Transportation Network Systems Engineering

Networking is the natural attribute of transportation systems. There are always inherent requirements of systems networking for construction and management of transportation infrastructure, the operation and service, as well as the administrative agencies and policies of transportation systems. In this paper, the concept of transportation network systems engineering (TNSE) is defined firstly, the state of the art of TNSE both in scientific research and practice are introduced from aspects of complex network analysis, networking operation management, and networking control & guidance. The existing scientific problems of TNSE are also summarized. The key scientific problem of complex network analysis for transportation systems is to explore the dynamic coupling and matching relationship between different transportation network topologies and the spatial-temporal evolution pattern of traffic dynamics, network capacity and reliability. The key scientific problem of networking operation management is how to realize the trade- off or Pareto optimality among multi- agents with benefits heterogeneity (governments, operating agencies, passengers) under the complex network structured transportation infrastructure. For the future development of transportation network systems engineering, we think that it is necessary to strengthen the research on the theory and methodology of integrated transportation network systems engineering, which should focus on scientific problems such as structural evolution mechanism of integrated transportation network, supply and demand balance mechanism of urban transportation network, complex structural characteristics and dynamic process of multi-layer integrated transportation network, as well as diversity and predictability of travel behavior, etc. Furthermore, the covering scopes of transportation network systems engineering also need to be fundamentally upgraded due to the dramatic change of transportation system accompanying with increasing of new emerging technologies. The flexible transportation network with nodes of self- driving, self- organizing, and self- decision- making ability and its shared operation mechanism will be an important research field of transportation network systems engineering in the future.