Model Predictive Contouring Control for Time-Optimal Quadrotor Flight

In this article, we tackle the problem of flying time-optimal trajectories through multiple waypoints with quadrotors. State-of-the-art solutions split into a planning task—where global trajectory is generated—and control accurately tracked. However, at current state, generating that considers full quadrotor model requires solving difficult time allocation via optimization, which computationally demanding (in order minutes or even hours). This detrimental for replanning in presence disturbances. We overcome issue by and concurrently Model Predictive Contouring Control (MPCC). Our MPCC optimally selects future states platform runtime, while maximizing progress along reference path minimizing distance to it. show that, when tracking simplified trajectories, proposed results approaches true one, can be generated real time. validate our approach world, where method outperforms both state art world-class human pilot terms lap achieving speeds up 60 km/h.