Reconfigurable Path Planning for Fixed-wing Unmanned Aircraft Using Free-Space Roadmaps

It has been shown that unmanned aircraft path planning problems can be solved efficiently with roadmap methods. But roadmap planning approaches are not flexible on changing environmental and kinematic constraints, which might be wind or the kinematic model. The proposed approach addresses the reconfigurable path planning of kinematically constrained unmanned aerial fixed-wing vehicles with roadmap based methods. In particular, reconfigurable means that environmental conditions, kinematic models, and the flight envelope can be adjusted for every planning request without the need of updating or rebuilding the roadmap. The local planner solves straight-curve-straight local planning problems and uses convex free-space volumes stored in a free-space roadmap for run-time efficient collision checking and path planning. Finally, the application of the approach is demonstrated for two vehicles.