Attitude control of a multicopter using L1 augmented quaternion based backstepping

This paper presents a quaternion based backstepping controller and a L₁ piecewise-constant adaptive augmentation to address the robust attitude tracking problem of a multicopter in the presence of uncertainties. The strict-feedback form of the multicopter dynamics is utilized to formulate the quaternion based backstepping controller which is designed for nominal conditions. This control law prevents unwinding by using a well-defined quaternion error. The backstepping controller is augmented using a L₁ adaptive control to increase the performance in the presence of uncertainties. In order to compare the performance of the baseline and the augmented controllers, different uncertainty cases (within position of the center of gravity and moment of inertia) were compared in simulation. A truncated L₂-norm of the tracking error is used for a quantitative comparison. Flight tests which validate the simulation results are presented.