Adaptive Traffic Signal Control: Game-Theoretic Decentralized vs. Centralized Perimeter Control

Area of Science:

• Traffic Engineering
• Transportation Science
• Control Systems

Background:

• Traffic congestion remains a major challenge in urban networks, impacting travel times, delays, and emissions.
• Existing traffic signal control strategies, including adaptive and perimeter gating control, have limitations in managing complex urban traffic dynamics.
• Decentralized control approaches offer potential for improved traffic flow and network performance.

Purpose of the Study:

• To compare the operational effectiveness of a novel decentralized Nash bargaining (DNB) traffic signal controller against state-of-the-art adaptive and perimeter gating control systems.
• To evaluate the DNB controller's ability to prevent and disperse traffic congestion within a protected urban network.
• To assess the impact of the DNB controller on key performance indicators such as travel time, delay, and emissions.

Main Methods:

• Implementation and assessment of DNB, adaptive, and gating traffic signal controllers within the INTEGRATION microscopic simulation software.
• Application of perimeter control (gating) based on the network fundamental diagram (NFD) on the borders of a protected urban network (PN).
• Comparative analysis of controller performance on a grid network, focusing on congestion prevention and network-wide efficiency.

Main Results:

• The DNB controller effectively prevents congestion within the protected network and enhances overall network performance, even without specific design for perimeter control issues.
• The DNB controller demonstrated superior performance compared to both gating and non-gating controllers.
• Statistically significant reductions were observed: 21-41% in average travel time, 40-55% in total delay, and 12-20% in emissions/fuel consumption.

Conclusions:

• The developed decentralized Nash bargaining traffic signal controller offers significant advantages over current state-of-the-art centralized and decentralized traffic control systems.
• DNB controller implementation leads to substantial improvements in traffic flow efficiency, reduced travel times, and lower environmental impact in urban settings.
• The findings support the adoption of DNB controllers for advanced urban traffic management, addressing congestion and enhancing network performance.