Feedback linearization-based model reference adaptive control for multirotor UAVs under uncertainties from fuel and payload

This paper proposes a feedback linearization-based model reference adaptive control (FL-MRAC) scheme to handle a multirotor unmanned aerial vehicle (UAV) equipped with a gasoline-battery-type hybrid system under uncertainties. The gasoline-battery-type hybrid propulsion systems are increasingly used to increase operational time. This multirotor system suffers from uncertainties that may originate from offcentered payloads, or fuel consumption during flight. These uncertainties can change model parameters such as mass, moment of inertia, and center of gravity. Therefore, a new control strategy is applied using feedback linearization as a baseline design augmented with a model reference adaptive control strategy on a multirotor UAV. To formulate the structure of uncertainties, the mathematical model of the multirotor UAV with fuel and payload is constructed considering the effect of the mass change rate and the off-diagonal inertia tensor. FL-MRAC improves the performance of attitude and altitude tracking using the closed-loop reference model under model parameter uncertainties and time-varying nature. To guarantee stability, the uniform ultimate boundedness of the closed-loop system is proved considering the time-varying parameters. Numerical simulations are performed using the root-mean-square error and the fast Fourier transform. In Case 1, FL-MRAC with a closed reference model exhibits a more accurate tracking performance and transient response than FL-MRAC with an open reference model under the uncertainties of time-varying parameter uncertainties. In Case 2, the proposed method of tracking performance demonstrates an improved attitude and altitude tracking performance compared to sliding mode control and nonlinear disturbance observer under time-varying parameter uncertainties and external disturbances due to fuel sloshing effect.