Command-filtered robust trajectory tracking control for aggressive maneuvers of quadrotor UAV with multiple unknown disturbances

Abstract

A quadrotor unmanned aerial vehicle should have the ability to achieve the desired flight manuevers regardless of the constraints posed by inevitable disturbances. Therefore, this research presents a robust control method designed for trajectory tracking of quadrotor. The method employs a reduced-order state observer that uses measurable quadrotor position and attitude to estimate only the rate of change of these variables. A modified disturbance estimation criterion, based on measurable states and their estimated rates, is introduced. Command filters and their compensations are utilized to replace numerical derivatives with estimated derivatives, improving accuracy and stability. Tracking errors, obtained from these compensated variables, are used to design state transformations. The control technique, based on sliding mode control combined with adaptive laws and disturbance observer, is formulated to attenuate various disturbances, including band-limited white noise, constant disturbances, exogenous disturbances, nonlinear disturbances with variable frequency and magnitude, Gaussian distributed random disturbances, and uniformly distributed disturbances. The integration of the reduced-order state observer with estimated differentiation minimizes computational loss and allows for quick computation. Stability analysis, conducted using Lyapunov criteria, along with simulation results on the DJI F450 quadrotor, validates the effectiveness and stability of the developed controller.

Keywords

• Quadrotor
• Disturbance observer
• Reduced-order state observer
• Tracking control
• DJI F450