Fuzzy predictive control for UAV attitude under disturbances and input constraints

A disturbance observer-based adaptive fuzzy model predictive control method is proposed to solve the attitude control difficulty of quadrotor UAVs in complex environments. A nonlinear dynamic model of the quadrotor UAV is established， including external disturbances. In order to effectively handle system nonlinearities and constraints， the T-S fuzzy modeling approach is employed to transform the nonlinear model into a set of local linear models which are then embedded into a model predictive control framework for rolling optimization solution. Meanwhile， an adaptive prediction horizon adjustment strategy is designed to dynamically optimize controller parameters， which is based on the system's dynamic characteristics to optimize controller parameters in real time and effectively balances dynamic response speed and robustness. Furthermore， by establishing a Lyapunov function， stability analysis of the state equations is conducted， rigorously proving the asymptotic stability of the system under this adaptive strategy. The simulation results demonstrate that outstanding stability control capability and strong robustness performance is presented by applying proposed method in complex uncertain environments.