Harnessing Digital Twin Technology for Adaptive Traffic Signal Control: Improving Signalized Intersection Performance and User Satisfaction

In this study, a digital twin (DT) technology-based adaptive traffic signal control (ATSC) framework is presented for improving signalized intersection performance and user satisfaction. Specifically, real-time vehicle trajectory data, future traffic demand prediction and parallel simulation strategy are considered to develop two DT-based ATSC algorithms, namely, digital twin 1 (DT1) and digital twin 2 (DT2). DT1 uses the delay experienced by each vehicle from all approaches connected to the target intersection, while DT2 uses the delay of each vehicle that occurred in all the approaches connected to the target intersection as well as immediate adjacent intersection. To demonstrate the effectiveness of these algorithms, the DT-based ATSC algorithms are evaluated with varying traffic demands at intersection, and individual user level. Evaluation results show that both DT1 and DT2 performs significantly better compared to the density-based baseline algorithm in terms of control delay reductions ranging from 1% to 52% for low-traffic demands. DT1 outperforms baseline algorithm for moderate traffic demands, achieving reduction in control delay ranging from 3% to 19%, while the performance of DT2 declines with increasing demand. For high-traffic demands, DT1 achieved control delay reduction ranging from 1% to 45% and DT2 achieved 8% to 36% compared to the baseline algorithm. Moreover, DT1 and DT2 effectively distribute the delay per vehicle among all the vehicles, which approach toward intersection, compared to the baseline ATSC algorithm. DT-based ATSC could help to improve user satisfaction by reducing prolonged delays at signalized intersections, specifically, for moderate and high-traffic demands.