Scheduling-driven Motion Coordination of Autonomous Vehicles at a Multi-Lane Traffic Intersection

Abstract

This paper addresses the motion coordination problem of autonomous vehicles that approach an intersection of a traffic network. The proposed approach assumes that there is a bi-directional communication link between incoming vehicles and the Intersection Controller (IC). Once the autonomous vehicles reach the communication range, the IC schedules their arrival time based on a first-come-first-serve rule and the availability of the intersection region. The appointed arrival time minimizes the delay time and ensures safety within the vehicle's motion constraints. Based on the appointed arrival time, the IC regulates the linear velocity of the vehicles in the intersection control region. The constrained optimization problem is solved by employing the metaheuristic method of Particle Swarm Optimization (PSO), enhanced with an adaptive penalty function. Simulation results demonstrate the efficacy of the proposed approach by comparison with a traditional signalized intersection.