Scenario-Based Nash Bargaining Control (NBC): A Cooperative Game Theoretic Approach for Resilient Urban Traffic Signal Optimization

Efficient traffic signal control in urban grid networks requires a system that can adapt robustly to highly variable demand patterns, ranging from free-flow conditions to heavy congestion. While decentralized game-theoretic approaches have shown promise, they often optimize locally, leading to suboptimal network-wide performance known as the "Price of Anarchy." This paper proposes a practical and resilient Scenario-Based Nash Bargaining Control (NBC) system that frames the entire multi-intersection grid as a single cooperative game. By treating the network as a "Grand Coalition," the controller maximizes a Global Nash Product, explicitly coordinating spatial interactions to achieve network-wide Pareto efficiency. A key innovation is the introduction of an "Adaptive Disagreement Point," a dynamic reference mechanism that ensures continuous, responsive control authority across all traffic regimes, including oversaturation, without relying on static capacity limits. The methodology was rigorously validated using SUMO microsimulation on a ۲ ×۲ grid under diverse and challenging scenarios: time-varying demand, structural imbalance, platoon progression, and high-density saturation. Quantitative results demonstrate that the proposed NBC delivers consistent, significant improvements over Fixed-Time control, reducing average waiting times by up to ۷۱.۶% and total CO ۲ emissions by ۲۱.۵% while maintaining network stability under stress.

کلیدواژه ها:

Traffic Signal Control ، Cooperative Game Theory ، Global Optimization ، Nash Bargaining ، Urban Grid Resilience