Collision-free trajectory planning based on Maneuver Selection-Particle Swarm Optimization
David Alejo, J. A. Cobano, Guillermo Heredia, Anı́bal Ollero · 2015
This paper presents a system for collision-free trajectory planning with multiple Unmanned Aerial Vehicles (UAVs) which automatically identifies conflicts among them. After detecting conflicts between UAVs, the system resolves them cooperatively using a collision-free trajectory planning algorithm based on a stochastic optimization technique named Particle Swarm Optimization (PSO). The new implementation of the PSO algorithm, named Maneuver Selection Particle Swarm Optimization (MS-PSO), presents improvements with respect to previous implementations. The execution time is reduced because the dimension of the problem is reduced, and different kinds of maneuvers can be selected to solve the detected conflicts: course/heading, speed or altitude changes. The MS-PSO has been validated with simulations in scenarios with multiple UAVs in a common airspace. Also, a comparison to a genetic algorithm and a PSO algorithm has been performed to highlight the advantages of the MS-PSO. The main advantage is that MS-PSO always ensures solution from the first iteration. This requirement is essential in safe cooperative missions.