Multi-query Path Planning for an Unmanned Fixed-Wing Aircraft

Abstract

Path planning for fixed-wing aircraft in obstacle rich environments requires the consideration of the motion constraints of the aircraft. Most existing motion planning algorithms for systems with dynamics are single query planners, which leads to a large computation overhead if multiple path planning problems are solved. This paper presents a roadmap-based multi-query path planner which generates feasible paths for a simplified dynamics model of a fixed-wing unmanned aircraft. Due to the multi-query property multiple path queries can be solved efficiently once the roadmap is constructed. Quasi-random Hammersley sets are used to place samples in a four dimensional configuration space to account for the motion constraints of the aircraft, which is modeled by the superposition of a Dubins vehicle in the horizontal plane and a double integrator with state and input constraint in the vertical plane. The optimization of roadmap parameters and post-processing of paths are discussed in detail. Simulation results show the applicability of the suggested planner.