MADER: Trajectory Planner in Multiagent and Dynamic Environments

This article presents MADER, a 3-D decentralized and asynchronous trajectory planner for UAVs that generates collision-free trajectories in environments with static obstacles, dynamic other planning agents. Real-time collision avoidance obstacles or agents is done by performing outer polyhedral representations of every interval the then including plane separates each pair polyhedra as decision variable optimization problem. MADER uses our recently developed MINVO basis to obtain volumes 2.36 254.9 times, respectively, smaller than Bernstein B-Spline bases used extensively literature. Our algorithm guarantees safety respect their committed constraints executing check-recheck scheme. Finally, extensive simulations challenging cluttered show up 33.9% reduction flight time, 88.8% number stops compared bases, shorter distances centralized approaches, total times on average synchronous approaches.