Composite observer-based robust model predictive control technique for ducted fan aerial vehicles

The ducted fan aircraft are an important type of hybrid aerial vehicle due to their annular safety mechanism called duct. Despite their numerous advantages, developing a control system for these tail-sitter configurations is a difficult task in the presence of disturbances and model inaccuracies. This paper presents a composite control strategy for these aerial robots subject to disturbances and model uncertainties that aims to construct a flight control algorithm with effective control functionality. Initially, a ducted fan aerial vehicle with adequate dynamic capabilities is utilized. A disturbance observer is employed to estimate the time-varying external disturbances and uncertainties. Moreover, two observers (extended and unscented Kalman filters) are used separately for state estimation to counter the unwanted noise. An online-based model predictive control framework is presented, where at each step, the nominal system’s state is updated by the real state. After the optimization problem is computed, the final composite control signal incorporates the optimal control action and observer estimates, thereby reducing the computational cost. Stability and feasibility are thoroughly analyzed. Comparative analysis based on simulation is carried out to reveal the efficacy of the proposed scheme.