Towards An Efficient Traffic System via CAVs: Demand Management and Real-time Traffic Control

With the dramatically rising traffic congestion issue, people are suffering the loss of working hours, increase in traffic accidents and pollution around the world. In Canada, drivers in the major cities were estimated to lose over 50 working hours per year in congestion during 2022, and in the US, it has cost more than 300 billion for the government to tackle the congestion issue. In recent decades, the unprecedented development in connected and automated vehicles (CAVs), coupled with the advancement of 5G and mobile edge computing (MEC), has brought profound changes to deal with the traffic congestion challenges. By enabling timely data exchange between vehicles and infrastructures, CAVs provide new possibilities for better demand management, more efficient and practical real-time traffic control. In light of the anticipated emergence of CAVs, the research in this dissertation aims to improve the traffic system by developing intelligent traffic control strategies, real-time vehicle guidance, and appropriate demand management measures. The overarching goal is to enhance traffic efficiency to reduce congestion and improve traffic mobility, especially for the urban arterials considering the existence of traffic signals. To achieve this goal, the whole research is structured into four key components: The first part focuses on developing a joint dynamic route guidance and signal control (DRG-SC) model for urban arterial traffic networks, which serves as a robust linkage that effectively connects demand modeling with traffic control. In this model, the real-time location and velocity data of CAVs as well as the signal timing plan of intersections will be utilized to capture the interaction between signal control and vehicle routing. The vehicle routing plan will be optimized by considering the signal delays at each intersection, and the signal timing plans will be updated based on the real-time traffic volume resulting from routing. The joint model utilizes a closed-loop control framework, which is more effective than open-loop control and can significantly reduce travel time. As a continuation of the first part of the research and considering that generating solutions from such a centralized model is computationally intensive, the second part of the research presented a distributed dynamic route guidance algorithm that utilizes local intersections’ information only but generates globally optimized results for the whole traffic network with the support of the MEC technology. The algorithm is derived from the backpressure routing control and the result suggested that the control effectiveness was much better than the dynamic shortest path (DSP) while close to the dynamic system optimal (DSO) traffic assignment. More importantly, the algorithm was verified to be effective in reducing communication and computation cost. In the third part, grounded on the prior research of traffic demand modeling in the first part, two strategies were proposed to better manage the CAV dedicated lane (CAV-DL) in a mixed traffic environment to improve capacity utilization. The CAV-DL is designed to physically separate the CAVs and Human Driving Vehicles (HDVs) to maximize the benefit of CAVs. However, the CAV-DL may be underutilized especially when the penetration rate of CAVs is low. To address this issue, in the first strategy, a dynamic right-of-way allocation method is adopted to allow HDVs to use the CAV-DLs when the lanes are relatively vacant. The second strategy designs tolling policies based on economic theory to explore the best demand distribution and further balance the travel time on the general lanes and CAV-DLs. Both methods were proven to be effective in better-utilizing road capacity and improving traffic mobility. The last part of the research focused on using CAV technology to promote electric vehicles (EVs), and a traffic environment with connected and automated electric vehicles (CAEVs) is assumed. In alignment with dual-carbon policies, encompassing carbon neutrality and carbon peaking, the promotion of EVs stands out as a prominent and transformative trend in the future of transportation. Following the similar closed-loop control logic developed in the first part, the work in this part tries to reach a trade-off between the energy consumption and the total travel time of CAEVs. By simultaneously optimizing the trajectory of the vehicles as well as the signal timing plans, the result effectively shows how CAV technology can help improve the travel and energy efficiency of EVs. Overall, the research presented in the dissertation covers real-time traffic control and urban arterial demand management in the CAV environment. The models and algorithms presented herein can effectively improve traffic efficiency. The results in this dissertation contribute to the CAV-related studies methodologically which provide insights into realizing a sustainable and efficient traffic system in the near future.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.