Cooperative multi-vehicle trajectory planning at unsignalized intersections under mixed traffic environment

Abstract

With the advancement of communication and automation technologies, urban roadways in the near future will feature the coexistence of conventional human-driven vehicles alongside connected and autonomous vehicles. To enhance safety and efficiency for vehicles navigating unsignalized intersections in mixed traffic environments, this paper proposes a novel strategy based on a cooperative vehicle-infrastructure system for multi-vehicle trajectory planning. In this strategy, the unsignalized intersection area is divided into a preparation zone, an adjustment zone, and a conflict zone, with a corresponding planning method designed for each zone. The generation of multi-vehicle trajectories in a mixed-traffic environment is formulated as an optimization problem. Subsequently, a novel collision avoidance method and an improved particle swarm optimization algorithm combined with genetic algorithm are proposed to address the problem. Furthermore, the strategy accounts for the treatment of vehicles with different priority levels. The experimental results on multi-vehicle trajectory planning at unsignalized intersections under varying operational conditions demonstrate that this method enables the safe and efficient passage of potentially conflicting vehicles through unsignalized intersections in mixed-traffic scenarios.