KF-Based Active Modeling for a Quadrotor Slung Load System

The control of quadrotor slung load (QSL) systems is a challenging task as a result of some distinguishing features like the multiple nonlinearities and time-varying load swing. To design an effective control scheme, the top priority is to establish an excellent model. In this paper, a practical active modeling method based on the Kalman Filter (KF) for a QSL system is developed to compensate uncertainties in system dynamics. First, a simplified quadrotor model is formulated as the reference model. Second, modeling errors are introduced and the KF is adopted to actively estimate the errors involved in the reference model. Finally, some experiments are conducted to validate the remarkable performance of the proposed modeling strategy in terms of uncertainties as well as disturbances in the dynamic model.