Optimization-Based Motion Planning for Vector Field Following in Dynamic Environments

This paper proposes a method for integrating trajectory optimization with vector field-based motion planning methods. This approach aims to address motion planning challenges, particularly in scenarios like UAV navigation, where vector fields are efficient but struggle with dynamic obstacles and motion constraints. Such challenges also include scenarios where there is no defined goal configuration such as border following, loitering, and curve circulation. Several vector field methods have been proposed to solve these problems but they are prone to failure when encountering previously unmodeled or dynamic obstacles as well as no-fly zones. The method proposed in this paper uses a vector field for high-level planning. The vector field is used to create paths for the vehicle, which are optimized for smoothness, obstacle avoidance, and vector field adherence before they are followed. The result is a smooth path that is fast to plan and easy to follow for a motionconstrained vehicle. A series of simulations was used to validate this methodology, which is compared with a previous method that uses RRT*.