Constraint-Decoupled Trajectory Planning for Agvs Via Safe Navigation Corridors

This paper proposes a safe navigation corridor framework to improve the trajectory planning of autonomous ground vehicle in obstacle-dense environments. By constructing two types of safety corridors, i.e., rectangular frame corridors and circular expansion corridors, around discrete waypoints along an initial <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$A^{*}$</tex>-based path, the framework enables local trajectory optimization with improved safety guarantees. A core innovation of this work lies in the constraint formulation: obstacle avoidance is enforced with respect to corridor boundaries, making the number of constraints independent of the number of obstacles within the corridor. This significantly reduces computational complexity in dense environments. Comparative analysis shows that rectangular corridors are easier to compute but introduce more constraints, while circular corridors offer a more compact representation with fewer constraints at the cost of higher construction effort. The effectiveness and feasibility of the proposed method are validated through simulations in a warehouse environment using ROS.