Adaptive robust backstepping attitude control for a multi-rotor unmanned aerial vehicle with time-varying output constraints

Output constraints and uncertainties are the main factors that degrade the control performance of the multi-rotor unmanned aerial vehicle (MUAV). In this paper, an adaptive neural network backstepping dynamic surface control algorithm based on asymmetric time-varying Barrier Lyapunov Function is proposed for the attitude system of a novel MUAV under asymmetric time-varying output constraints, model uncertainties and external disturbances. The asymmetric time-varying Barrier Lyapunov Function, which will grow infinite when its arguments approach some limit, is introduced to keep the output under time-varying asymmetric constraints. Considering the derivation problem of the virtual control function in backstepping, the dynamic surface control is applied to simplify the algorithm. The adaptive neural network is used to approximate the dynamic model of the attitude system, and the minimal learning parameters are employed at the same time to reduce online computation burden. In order to balance out the external disturbance and further reduce the approximate error of the adaptive neural network, a robust term is designed to compensate the above negative impacts. The proposed algorithm guarantees that all the signals of the closed-loop system bounded by Lyapunov theory. Finally, some contrast simulation experiments are given to illustrate the effectiveness and superiority of the control scheme.