Particle Swarm Optimization for collision-free 4D trajectory planning in Unmanned Aerial Vehicles

This paper presents a new system which automatically identifies conflicts between multiple UAVs (Unmanned Aerial Vehicles) and proposes the most effective solution considering the available computation time. The system detects conflicts using an algorithm based on axis-aligned minimum bounding box and resolves them cooperatively using a collision-free trajectory planning algorithm based on a simple one-at-a-time strategy to quickly compute a feasible but non-optimal initial solution and a stochastic optimization technique named Particle Swarm Optimization (PSO) to improve the initial solution. PSO modifies the 4D trajectories of the UAVs with an overall minimum cost. Determining optimal trajectories with short time intervals during the execution of the mission is not feasible, hence an anytime approach using PSO is applied. This approach yields trajectories whose quality improves when available computation time increases. Thus, the method could be applied in realtime depending on the available computation time. The method has been validated with simulations in scenarios with multiple UAVs in a common workspace.