Queue Dissipation-Based Max Pressure Signal Control Using Vehicle Re-Identification Data

Traffic congestion is a persistent urban challenge, and signal control strategies play a crucial role in enhancing network efficiency. Traditional max pressure-based control systems prioritize optimizing throughput based on known queue lengths but often overlook the impact of intersection delays and real-world operational constraints. This study aims to develop a decentralized max-pressure-based traffic signal control system that integrates both queue length and delay effects within the max-pressure framework by introducing a novel performance metric called queue dissipation time into the control strategy. The proposed cyclic-based max pressure algorithm relies on sample travel time data obtained from probe data sources. It considers pre-defined phase sequences to enhance real-world applicability. When evaluated in a simulation environment with a realistic 20% vehicle re-identification penetration rate, the proposed approach achieved queue and delay performance comparable to control strategies relying on complete vehicle arrival data. Furthermore, implementing dynamic cycle length led to a 23% reduction in delay compared to the fixed-cycle max pressure method, inferring the benefits of adaptive timing adjustments. By incorporating queue dissipation time and considering the constraints of real-world signal operations, the proposed control system enhances the practicality of max pressure-based traffic signal strategies. This study provides a foundation for implementing decentralized adaptive traffic signal control in urban networks, particularly in environments with limited vehicle data availability. The algorithm exhibited reliable performance across various levels of data availability, delivering comparable results under both low (20%) and high (50%) penetration rate scenarios. This consistency highlights its adaptability and operational feasibility, without requiring excessive traffic data or extensive infrastructure.