Collision Avoidance of Fixed-Wing UAVs in Dynamic Environments Based on Spline-RRT and Velocity Obstacle

Abstract

It is crucial to online plan a smooth, continuous and collision-free path to navigate a fixed-wing unmanned aircraft through complex environments. In a static environment, given priori knowledge of the map, the sampling-based RRT (rapidly exploring random tree) method and its variants are efficient to provide global path planning to navigate the fixed-wing aircraft through static obstacles. However, in complex and dynamic situations, the fixed-wing aircraft may encounter dynamic obstacles when following the planned global path. In the presence of dynamic obstacles within the sensing range of the fixed-wing aircraft, it is challenging to on-line generate a new collision-free and smooth path. In this paper, a real-time collision-free path planning strategy is proposed for fixed-wing aircraft in dynamic environments. Specifically, the proposed collision-free path planner named as Spline-RRT-VO is presented incorporating the spline-RRT algorithm and the velocity obstacle (VO) method to avoid a high-speed dynamic obstacle. In the proposed spline-RRT-VO approach, a random tree grows in the local area, meanwhile, the VO method is used to extend tree edges and reject unavailable nodes. It improves the tree growing in a more efficient and smooth manner. Simulation results verify the effectiveness of the spline-RRT-VO method to navigate the fixed-wing UAVs through dynamic environments.