An adaptive neural design for planar rigid formation of three coleaders in unknown flowfields

Abstract

This article deals with the robust planar rigid formation control problem of three second-order coleaders with unknown flowfields acting on the velocity and acceleration respectively. To yield the uniform boundedness property of the resulting system, an adaptive projection is introduced to design the novel adaptive neural control law. To avoid the derivatives of Gaussian functions of neural networks, dynamic surface is used. Simulation results are provided to illustrate the effectiveness of the proposed control law.