Neuro-adaptive singularity-free finite-time attitude tracking control of quadrotor UAVs

In this paper, a neuro-adaptive finite-time control law is proposed for the quadrotor unmanned aerial vehicles. Instead of employing any piecewise continuous functions, a singularity-free ter-minal sliding mode surface and an auxiliary function are presented to overcome the singularity issues resulted from the differentiation of the sliding variable and the error-related inverse matrix in the controller design, respectively. A simple neural network is employed to estimate the unknown dynamics, such that no prior knowledge on system model uncertainties is required for designing attitude controllers. With the presented control law, the finite-time convergence of the attitude and angular velocity errors can be guaranteed by rigorously theoretical analysis, and numerical simulations are performed to demonstrate the satisfactory performance of the proposed method.