Multi-lane multi-platoon control with collision avoidance

Cooperative control for multiple vehicles has evolved as a promising technology to improve traffic safety and efficiency. However, due to the complexity and variability in traffic, especially multiple lanes' varied conditions and functions, nearby vehicles and obstructions always present in a manner that is unpredictable. Therefore, this paper handles the issue of controlling Connected and Autonomous Vehicles (CAVs) convoy in heterogeneous multi-lane scenarios in the presence of stochastic, dynamic surrounding vehicles and unknown obstacles. A novel approach based on Laplacian feedback control is proposed to address these issues, which aims to avoid collisions, to produce steady, smooth, and uniform trajectories and to reconfigure arbitrary shapes to respond to complex traffic conditions. The presented method is a coupling in terms of distributed graph theory and artificial potential field, in which lane-keeping, collision avoidance, and following behaviors are integrated into the proposed motion planning model. In addition, a MPC-based vehicle dynamics controller is introduced to accomplish trajectory tracking robustly. Simulation experiment indicates that the proposed method enables the convoy to successfully avoid all collisions within 30s and to follow stably. Copyright (c) 2023 The Authors.