Data-Driven Backstepping Sliding-Mode Control of Disturbed QUAVs

Suffering from consistent external disturbances and unavailable kinetic dynamics, a data-driven backstepping sliding-mode control (DBSMC) scheme by the combination of backstepping technique and data-driven sliding-mode control is innovatively proposed for complex quadrotor unmanned aerial vehicles (QUAVs). To maintain the salient advantages of traditional PID algorithm, a data-driven PD-type sliding-mode surface is first designed for attitude subsystem, and thereby contributing to an efficient PD control law while with strong robustness and model-independent property. Then, the virtual control strategy and backstepping technique are further incorporated into the position controller design. In light of the virtual control philosophy, the proposed attitude controller and position controller are efficiently cohered in a cascade form, and the closed-loop stability pertaining to the DBSMC scheme can be ensured theoretically. Eventually, the effectiveness of the devised DBSMC scheme is evaluated via simulation studies.