Autonomous traffic at intersections: an optimization-based analysis of possible time, energy, and CO2 savings

In the growing field of autonomous driving, traffic-light controlled intersections as the nodes of large traffic networks are of special interest. We want to analyze how much an optimized coordination of vehicles and infrastructure can contribute to a more efficient transit through these bottlenecks. In addition, we are interested in sensitivity of the results with respect to traffic density, turning behavior, or certain regulations of traffic lights. To this end, we develop a mixed-integer linear programming (MILP) model to describe the interaction between traffic-lights and discretized traffic flow. It is based on a microscopic traffic model with centrally controlled autonomous vehicles and extended formulations for different switching regulations. We aim to determine a globally optimal traffic flow for given scenarios on a simple urban road network. This amounts to finding controls for the movement of each car as well as for each traffic light such that an objective function is optimized and collisions are avoided. The resulting models are very challenging to solve to global optimality, in particular when involving additional realistic traffic light regulations such as minimum red and green times. An evaluation of the numerical results with a traffic simulation tool indicates that the performance indicators time, energy, and emissions could be concurrently reduced by a significant amount. Potentially, the same models and algorithms might be the basis for future traffic control systems.

Citation

Published in Networks: 06 October 2021, https://doi.org/10.1002/net.22078

Article

Download

View Autonomous traffic at intersections: an optimization-based analysis of possible time, energy, and CO2 savings