Robust and Adaptive State Estimation of UAV Quadrotors with a High Gain Approach

Abstract

In this chapter, we study the problem of joint states and external aerodynamic disturbances as well as the problem of adaptive estimation of quadrotor systems. We investigate two realistic configurations of the quadrotor system: in the first configuration, the dynamics of longitudinal and angular velocities are corrupted by unknown time varying disturbances and in the second one, the quadrotor system is subject to constant unknown parameters. For the two latter cases of study, only positions and angles are available for measurements. In particular, we focus on the problem observer matching condition which is necessary to solve both robust and adaptive estimation problems. The latter restrictive assumption is not verified for the quadrotor system. To overcome this problem, we adopt the approach of generation and estimation of auxiliary outputs based on high gain approximate differentiators. For the case of presence of time varying disturbances, we present a first order sliding mode observer in cascade with a high gain observer to reconstruct both the states and the unknown disturbances of the quadrotor system. For the case of presence of constant unknown parameters, we introduce an adaptive observer in cascade with the same high gain observer used in the first configuration to reconstruct both the states and the unknown parameters. Numerical simulations are depicted to illustrate the effectiveness and the good performances of the proposed robust and adaptive estimation approaches.