Optimising Traffic Flow Through Synchronised Signal Configurations Along Urban Corridors

Traffic signal optimisation is crucial for reducing congestion and improving traffic flow, and a large body of work is being devoted to designing and developing corresponding techniques. However, there are significant challenges to be faced when deploying generated traffic signal plans, as they must align with the constraints and features of the specific real-world infrastructures where they will be implemented. To tackle this, this work introduces a configuration optimisation methodology that can meet the specific requirements of the urban traffic control (UTC) architecture widely used in the United Kingdom. A mathematical model is designed to maximise total throughput along the corridor by dynamically adjusting configurations of each junction to deal with expected traffic conditions. A Rolling Horizon (RH) approach combined with an adaptive large neighbourhood search (ALNS) method is employed to solve the model, which enables efficient traffic signal management at each planning step. The proposed approach has been tested in simulation on historical data from a major corridor in the Kirklees region, UK, to demonstrate its computational efficiency and performance; the results show that it can effectively deal with a wide range of traffic conditions.