A prioritized collision avoidance methodology for autonomous driving

In this paper, we develop a centralized and automated conflict resolution methodology for vehicles that guarantees collision-free motion by coordinating the vehicles' speeds on predetermined paths. A new computationally efficient formulation for collision-free motion constraints of vehicles is proposed. We develop a methodology to maximize the speed and safety distance of vehicles subjected to the collision-free motion constraints. Our method instead of constantly checking Euclidean distances among the vehicles, which is computationally expensive, first, finds the potential collision zones in the environment and then determines collision-free constraints based on a parameter called velocity rate. We also develop a new traffic flow measurement quantity, described as motion density, to study the performance of different conflict resolution scenarios. We studied how safety and speed objectives of vehicles will affect the motion density. Using simulation results the effectiveness of our approach is verified for coordinated transportation of multiple vehicles.