UAV Formation Flight Control Method Based on Annealing Recurrent Neural Network

To address the problem of an unmanned aerial vehicle (UAV) close-formation flight, the wingman pitch angle caused by aerodynamic interference is taken as the extreme value search variable, and the annealing recurrent neural network extremum search algorithm is used to make the wingman interference pitch angle converge to its extreme value. This minimizes the amount of power required by the wingman in a close-formation UAV flight. The problem of the control variables switching back and forth, and the phenomenon of output chatterin traditional extremum search algorithm, are eliminated. The dynamic performance of the system is improved, and the stability analysis of the system is simplified. The effectiveness of the algorithm is verified by the simulation of UAV close flying formation. © 2022 CRL Publishing Ltd.