Queue Estimation and Consideration in Vehicle Trajectory Optimization at Actuated Signalized Intersections

• Published in: 2024 IEEE International Conference on Smart Mobility (SM) pp. 61 - 66
• Main Authors: Shafik, Amr K., Rakha, Hesham A.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 16-09-2024
• Subjects: Actuated traffic signals; back of the queue estimation; eco-driving; Estimation; Fuels; Heuristic algorithms; optimizing vehicle trajectories; Probes; Queueing analysis; Real-time systems; stochastic optimization; Timing; Trajectory optimization; uncertain switching times; Uncertainty; Vehicle dynamics
• DOI: 10.1109/SM63044.2024.10733497

This research develops a Green Light Optimal Speed Advisory (GLOSA) EcoDrive system operating in the proximity of actuated traffic signals considering the vehicle queuing effects. The spatiotemporal back- and front-of-queue positions are estimated in real-time using loop detector and probe vehicle data while considering the inherent uncertainty in traffic signal timings. The fuel consumption savings of the proposed system are quantified through detailed simulation experiments of a single vehicle approaching a traffic signal and signalized intersection simulations at different market penetration levels (0% to 100%) of controlled vehicles. Findings from the single vehicle perspective show a 28.7% saving in fuel consumption is achieved without considering the queueing process, while 35.7% is achieved with the consideration of the queueing process. Moreover, it is shown that the proposed queue estimation algorithm resulted in a 6.4% saving in fuel consumption compared to the case of using shockwave theory for back-of-queue estimation. Furthermore, isolated intersection simulations demonstrate an average fuel saving of 12.1% with 100% market penetration level. In this regard, although significant savings are derived from the single vehicle perspective, the isolated intersection simulations showed only a 1% improvement when considering the queuing process. This highlights the significance and limitations of considering the surrounding traffic on the performance of the proposed GLOSA algorithm when deploying these systems in real-world traffic environments.