Optimal trajectory generation of a drone for wheelchair tracking using mixed-integer programming

This study considers the optimal-trajectory generation problem in which a drone recognizes a wheelchair with an onboard camera while avoiding collision with obstacles. The optimal-trajectory generation was developed as a mixed-integer programming problem. The same method was used to recognize the wheelchair. In this regard, the optimal-trajectory generation problem was solved at each time step using model predictive control, and the first element of the optimal input was applied. The MATLAB optimization toolbox was used to solve this problem. As a result, the drone could avoid obstacles while recognizing the wheelchair that moved to the target trajectory considering the signals transmitted by the drone.