Generalized Attitude Model Identification of a Quadrotor

Abstract

Abstract This paper studies the generalized attitude model (GAM) identification of a controlled quadrotor by using flight experimental data. The GAM describes the dynamics of the quadrotor and its driving motors. Its mechanism analysis will be first carried out for guiding the parametrization of its estimation model. Before the prediction error method (PEM) is applied on the GAM identification, the involved closed-loop identifiability problem is addressed. Finally, the accuracy of the estimated GAM is validated by comparisons of Bode diagrams and controller performances achieved by a redesigned robust controller.