Improved model-free adaptive attitude control for UAV based on IKF

To address the challenges of strong nonlinearity, model uncertainties, and noise interference in quadrotor unmanned aerial vehicle attitude control, and to overcome the slow convergence and steady-state oscillations commonly observed in traditional model-free adaptive control, this study proposes an improved control strategy. An improved Kalman filter is integrated with an improved compact-form model-free adaptive control scheme. A novel parameter update mechanism incorporating error feedback and derivative terms is introduced to accelerate convergence and suppress steady-state oscillations. Moreover, a model-free improved Kalman filter based on input/output data is developed to mitigate the reliance on accurate system models that is inherent in conventional Kalman filters. The optimal state estimates obtained from the improved Kalman filter are utilized to enhance the performance of the improved compact-form model-free adaptive control controller. The ultimate boundedness of the closed-loop system is rigorously proven based on the contraction mapping principle. Simulation results demonstrate that using the proposed approach significantly improves attitude tracking accuracy, accelerates system convergence, and enhances noise rejection under external disturbances and measurement noise. This method offers a robust, model-free solution for unmanned aerial vehicle attitude control in complex environments.