Robust trajectory tracking of an uncertain quadrotor with almost global finite-time convergence

Abstract

This paper addresses the three-dimensional trajectory tracking control problem of quadrotor unmanned aerial vehicles (UAVs) in the presence of modeling uncertainties and external disturbances. We propose two control strategies that steer the quadrotor onto its desired path within a finite time and maintain it thereafter for all future times. The proposed strategies use a cascaded control architecture, where the inner loop consists of the attitude dynamics and the outer loop consists of the translational dynamics of the quadrotor. The proposed strategies guarantee the global finite-time convergence of the tracking errors. Consequently, the quadrotor tracks its desired trajectory globally regardless of its initial engagement geometry with respect to the path. The first strategy draws motivation from the terminal sliding mode control. Meanwhile, in the second strategy, we modify the reaching law using an exponential function, which dynamically adapts to the variation of system states and guarantees a shorter reaching time with improved tracking performance. We demonstrate the robust performance of the proposed strategy in the presence of system lag and bounded external disturbances, which may emulate the effect of wind waves and other environmental imperfections neglected during the controller design.