Command-Filter-Based Finite-Time Control for Human-in-the-Loop UAVs With Dead-Zone Inputs

Abstract

This paper studies the adaptive neural finite-time attitude control problem for six-rotor unmanned aerial vehicles (UAVs) with dead-zone inputs. Under the assumption that control inputs of leader are provided by a human operator, the command-filter-based finite-time attitude control protocol is proposed to achieve leader-follower consensus in finite time. In the control design, the command filter technique and radial basis function neural networks (RBF NNs) are adopted to solve the problems of explosion of complexity and uncertain nonlinear dynamics, respectively. In addition, dead-zone nonlinearities of control inputs are compensated by the boundedness of dead-zone slopes. Based on the presented control scheme, the finite-time stability of UAVs is obtained via the Lyapunov stability theory. Finally, simulation results validate the control property of the proposed strategy.