Backstepping control for a UAV-manipulator tuned by Cuckoo Search algorithm

Manipulators coupled with an Unmanned Aerial Vehicle (UAV) have made it possible to perform aerial handling, transport, and picking maneuvers. One of the techniques used to control these systems is a backstepping controller that has shown high performance compared to a PID in the face of uncertainties and parametric disturbances. This paper presents the study of a backstepping controller for a mobile manipulator (MM–UAV) system tuned with the Cuckoo Search algorithm (CS) for trajectory tracking. Unlike other research, this study focuses on optimization using this metaheuristic algorithm that has never been applied in an MM–UAV. The system is divided in a novel way to implement the CS, considering the dependence of each rotation axis with the correspondence translation axis. Additionally, the tuning focuses on two critical points of the dynamic response, the overshoot and settling time. The results at the simulation and experimental level show that for all cases, a settling time of fewer than 0.8 s and overshoot is minor than 2%. This allows a balanced response of the system, which directly impacts energy consumption. The results are compared with a PID controller to verify the proposed work efficiency, showing a reduction of up to 8% of overshoots without exceeding in any experiment the maximum settling time of 0.8 s imposed to the system.