Research On Dynamic Utilization Of Lanes In Urban Road Network In The Connected And Automated Driving Environment

The problems of congestion,energy consumption and emissions in urban transportation system mainly come from the imbalance between supply and demand.The lanes within the city’s road network can serve various purposes like accommodating vehicular traffic as well as providing parking options.Moreover,it can be utilized for non-traffic activities such as mobile stores and parklets when additional space is available.The advent of connected and autonomous vehicles holds promise in optimizing the utilization of lane resources,making it more efficient,reasonable,and effective.This study proposes to allocate lanes in the road network among different usage scenarios and lane directions.This study systematically focuses on the flexible utilization of lanes,aiming to enhance the overall performance of the transportation system.The main research content of this study includes:First,focusing on the dynamic traffic demands of vehicle passing,aiming at large differences in the quantity and spatial distribution of traffic demand at different periods of the day,the lane direction optimization problem under the connected and autonomous driving environment is studied in the first part.It is modeled as a mixed integer non-linear programming problem and the directions of general road segments and intersection approach lanes in the road network are optimized simultaneously,leading to coordination and consistency in lane arrangements,demand allocation,and intersection control.Furthermore,a two-layer solution algorithm that decouples lane arrangement optimization and network flow assignment is designed to address the largescale road network problem.Next,this study focuses on both the dynamic and static traffic demands of vehicle passing and parking.Taking advantage of the remote parking functionality of autonomous vehicles,the second part of this study investigates the problem of on-street parking planning in the context of connected and autonomous vehicles,and models it as a multi-stage mixed-integer linear programming problem.A solution algorithm is designed for large-scale road networks.By modeling the daily commute chain of users and vehicles,the study analyzes the characteristics of three types of traffic flows,namely user trips between homes and workplaces,vehicle access and leaving of the parking lots,and vehicle transfers between different parking lots.The optimization of parking lane locations,assignment of trip demands and parking demands and transfer flows are achieved simultaneously.Finally,in the third part of this paper,aiming at the tension of urban land caused by the gradual increase of urban road network density,it is proposed that lane resources can be used for various commercial and residential activities during off-peak hours.The activity space planning problem in the context of connected and autonomous vehicles is studied in this part.The result of reversible lane optimization is used as the initial lane setting before the planning of activity lanes.By drawing on the concept of average node degree in graph theory,I establish the indices and modeling methods of the time aggregation degree and space aggregation degree of activity roads.The activity space planning problem is modeled as a mixed integer bi-objective nonlinear programming problem.Further,an efficient solution algorithm based on tabu search is designed to solve this problem.The research of above three aspects comprehensively considers factors such as lane settings,modeling of intersection control methods,and network-level traffic flow assignment,providing planning and decision-making guidance for dynamic utilization of lanes from different perspectives.