Distributed Formation Control for Car-Like Robots Navigating in Complex Environments

Formation control of car-like robots has garnered significant attention due to its broad applications in logistics, transportation, and resource exploration. However, previous research has mostly considered environments with a limited number of obstacles, thereby restricting the practical deployment of robot formations in complex real-world scenarios. To address this limitation, we propose an integrated spatio-temporal trajectory optimization method for the navigation of car-like robot formations in complex and densely obstructed environments, effectively balancing formation maintenance and obstacle avoidance. Precisely, we introduce a graph-based formation error function that quantitatively captures the deviation between the current and desired formations, enabling robots to dynamically adjust their positions for optimal formation maintenance performance. By integrating the formation error function, obstacle avoidance penalty, and other factors such as control effort and dynamical feasibility, into unconstrained trajectory optimization, our method concurrently resolves conflicting objectives. Experimental results demonstrate the effectiveness of our approach and its superiority over advanced baselines.