Distributed Algorithm for Optimal Vehicle Coordination at Traffic Intersections

Automated vehicle coordination can be used to control vehicles across traffic intersections safely and efficiently. This paper proposes a novel parallelizable algorithm, which solves the coordination problem at traffic intersections under a given precedence order by using a tailored variant of the augmented Lagrangian based alternating direction inexact Newton method (ALADIN). Here, each vehicle solves its own optimal control problem and exchanges information about arrivial and departure times at the intersection with its neighbors such that collisions are avoided. We illustrate the performance of ALADIN by analyzing two scenarios, one during rush hour and one at low-traffic conditions.