Anti-saturation adaptive finite-time neural network based fault-tolerant tracking control for a quadrotor UAV with external disturbances

Abstract

This brief is focused on the finite-time tracking control of the quadrotor unmanned aerial vehicle (UAV) subject to external disturbances, parametric uncertainties, actuator faults, and input saturation. According to the principle of the Euler-Lagrangian methodology, a comprehensive dynamics is first decomposed into the position and attitude subsystems to accommodate the controller design. Different from the most of previous studies with an ideal assumption that velocity information is available, a robust exact differentiator is employed to gain the precise information of unavailable velocity in finite time. Then, by utilizing the strong approximation of the radial basis function neural network (RBFNN), the NN-based fault-tolerant control scheme is proposed to compensate for parametric uncertainties, external disturbances and actuator faults. More importantly, a novel adaptive mechanism is responsible for automatically adjusting the NN's parameters, which not only can effectively avoid the selection of large adaptive gains, but can also greatly decrease the number of online-updated leaning parameters. To solve the input saturation problem, an auxiliary dynamics system is constructed. Based on the Lyapunov theoretical framework, it is proved that all the closed-loop signals are uniformly ultimately bounded and the tracking errors can converge into small neighborhoods around the origin in finite time. Finally, simulation results are verified to intuitively reveal the good tracking performance of the introduced composite controller in terms of the finite-time error convergence, strong robustness, fault tolerance, and saturation elimination.