Cooperative Trajectory Planning for Multiple Vehicles: State of the Art and Future Perspectives

Abstract

This paper surveys cooperative trajectory planning for multiple autonomous vehicles operating in a shared workspace, where “vehicles” broadly include ground vehicles, aerial drones, surface vessels, and underwater vehicles. It first defines the nominal problem as a centralized optimal control formulation that enforces kinematic feasibility, obstacle avoidance, and inter-vehicle collision avoidance while optimizing a cost function. The paper then organizes solution architectures into centralized and distributed paradigms. Centralized methods are reviewed through two branches: simultaneous strategies that solve a single coupled program for all trajectories, and joint strategies that introduce structured staging, conflict-driven refinement, or precedence fixation before continuous trajectory optimization. Distributed methods are subsequently summarized by how local subproblems are constructed and coordinated through limited information exchange. Three focused topics follow: common collisionavoidance modeling patterns across architectures, statistics of experimental setups reported in the surveyed studies, and representative application domains. Finally, the paper presents future perspectives and closes with concluding remarks.