Metaheuristic approach for designing robust traffic signal timings to effectively serve varying traffic demand
Chaitrali Shirke, Nasser R. Sabar, Edward Chung, Ashish Bhaskar
Abstract
Traffic demands at intersections vary across various periods of a day and from day to day. Generally, fixed time traffic signals are designed considering the average traffic flows across multiple days over a predetermined time interval. This approach overlooks the day to day variability in traffic demand, leading to inefficient and unreliable signal control performance. A signal plan should be robust such that it is less sensitive to demand variations and can maintain near-optimal performance during varying traffic demand. To address this need, the paper presents a new offline scenario-based framework, named Metaheuristic Robust plan Approach (MHRA), that identifies a robust plan for fixed time signals. MHRA includes a heuristic that considers optimum signal plan for various demand scenarios and corresponding costs to find a robust solution. The numerical experiments are performed using realistic traffic demand scenarios on an arterial corridor to verify the MHRA framework. The outcomes concluded that the framework produces a robust signal plan that outperforms a nominal signal plan based on average traffic demand and maintains stable performance under varying demand. Benchmarking MHRA with other scenario-based approaches proposed in the literature such as mean-variance optimization and conditional value at risk minimization confirms better efficiency for MHRA.