A Dynamic Spatiotemporal Coordination Optimization Method for Intersections Based on Variable Approach Lanes

Abstract

This study proposes a dynamic spatiotemporal coordination framework for signalized intersections with variable approach lanes (VALs), aiming to enhance road resource utilization and mitigate demand-supply imbalances. Lane-switching decisions are determined by switching intervals, movement saturation, and average delay, with transitions implemented through coordinated control. The framework adopts dual objectives of minimizing average delay and lane-changing frequency, while analyzing the effect of lane-switching intervals. Detector data on queue lengths and arrival rates are used to construct arrival-departure diagrams, from which an analytical delay formulation is derived to support a delay-based signal timing optimization model. Case studies on a Shanghai intersection using SUMO simulations show that, compared with fixed-time control, the proposed strategy reduces approach delay by 5.0-7.5%, increases throughput by 2.6-9.7%, and decreases overall intersection delay by 2.8-6.0%. These findings demonstrate that the proposed method effectively alleviates congestion and improves efficiency at intersections with unbalanced traffic demand.