The Application of an Annealing Recurrent Neural Network for Extremum Seeking Algorithm to Optimize UAV Tight Formation Flight

To solve the problem of unmanned aerial vehicle (UAV) tight formation flight, a new autopilot for UAV tight formation flight is designed by sliding mode control. If the pitch angle thetav_w of the wingman induced by aerodynamic interference is considered as a seeking object, the minimum power demand of the wingman can be gained by an annealing recurrent neural network for extremum seeking algorithm (ESA) presented in the paper. In tight formation flight, the wingman motion is affected by the aerodynamic interference of the adjacent lead UAV. The forces produced by the aerodynamic interference are complex functions of the relative positions of the UAVs. The maximum forces can be gained by minimizing the pitch angle of the wingman, at the same time, the optimal configuration of UAV tight flight formation and minimum power demand of the wingman are also attained by the algorithm proposed. The algorithm of an annealing recurrent neural network for ESA can be realized by combining the annealing recurrent neural network with ESA, which can solve the "chatter" problem of the output and the switching of the control law of the general ESA and improve the dynamic performance of the system greatly and simplify the stability analysis of ESA