Centralized Dynamic Vehicle Routing with Macroscopic Fundamental Diagram

Optimizing urban mobility through dynamic vehicle routing and centralized control mechanisms is a complex task with a high computational cost. Macroscopic fundamental diagram (MFD) provides a practical tool for traffic state control at a macroscopic level, aiming to achieve system optimum. This paper introduces a centralized dynamic vehicle routing system based on MFD for connected vehicles. Its objective is to minimize travel time delays for individual vehicles and the entire city. The system divides the city into multiple regions and calibrates the MFD for each region. Connected vehicles monitor the macroscopic traffic state in real time and adjust the weight of link travel time accordingly. This enables the system to dynamically search for the most efficient routes, maintain region traffic close to capacity, maximize throughput, and alleviate network congestion. The system's performance is evaluated on both a synthetic grid network and a real-world city network, demonstrating significant mobility benefits. Results show a potential doubling of region capacity and up to a 60% reduction in travel time delays. Additionally, the paper analyzes the influence of connected vehicle market penetration rates and network congestion levels on mobility benefits using microscopic traffic simulations.