An Evasive Maneuvering Algorithm for UAVs in Sense-and-Avoid Situations
David Hyunchul · InTech eBooks · 2009
In this article, we presented a collision avoidance algorithm for UAVs using nonlinear model predictive control. The preview mechanism of receding horizon control is found ideal for such cases when the obstacles are moving. The proposed algorithm was also applied to obstacle avoidance problems, where onboard sensors combined with updated local map was combined with the MPC solver to compute conflict-free trajectories. Both cases are validated first in simulation and then in realistic experiments using helicopter UAVs. In each set of experiments, the proposed NMPC-based algorithms were able to run in real-time for computing conflict-free trajectories. The proposed algorithm will be further extended to vision-based sensing as well as the avoidance problems of fixed-wing UAVs.