A Novel Finite-Time-Gain-Adjustment Controller Design Method for UAVs Tracking Time-Varying Targets

Abstract

As for time-varying tracking control problems, many neural-dynamics-based control methods have been proposed because of their high efficiency. The varying-parameter convergent neural dynamics design method with the characteristic of super-exponential convergence has been applied to design controllers for unmanned aerial vehicles. Although the varying-parameter convergent neural dynamics controller has a fast convergence speed, it still needs long enough time to achieve tracking theoretically. By combining the finite-time activation function with the varying-parameter convergent neural dynamics design method, a finite-time-gain-adjustment design method is proposed and proved theoretically in this paper. This controller can make state variables of the system converge to their time-varying targets in finite time. Compared with some traditional methods, contrastive experiments and application to a multi-rotor unmanned aerial vehicle system illustrate that the proposed controller has finite convergence time, better anti-noise performance, and faster convergence speed, which enable the multi-rotor unmanned aerial vehicles to track time-varying targets more quickly and accurately, so as to achieve more complex and efficient control tasks.