Multi-Fidelity Black-Box Optimization for Time-Optimal Quadrotor Maneuvers

We consider the problem of generating a time-optimal quadrotor trajectory\nthat attains a set of prescribed waypoints. This problem is challenging since\nthe optimal trajectory is located on the boundary of the set of dynamically\nfeasible trajectories. This boundary is hard to model as it involves\nlimitations of the entire system, including hardware and software, in agile\nhigh-speed flight. In this work, we propose a multi-fidelity Bayesian\noptimization framework that models the feasibility constraints based on\nanalytical approximation, numerical simulation, and real-world flight\nexperiments. By combining evaluations at different fidelities, trajectory time\nis optimized while keeping the number of required costly flight experiments to\na minimum. The algorithm is thoroughly evaluated in both simulation and\nreal-world flight experiments at speeds up to 11 m/s. Resulting trajectories\nwere found to be significantly faster than those obtained through minimum-snap\ntrajectory planning.\n