Neurodynamic Formation Maneuvering Control With Modified Prescribed Performances for Networked Uncertain Quadrotors

This article investigates a neurodynamic formation maneuvering control problem with guaranteed performances for networked uncertain quadrotors. A modified prescribed performance control (MPPC) with asymmetric behavior boundaries is proposed to enable a better regulation ability for formation synchronization error without any overshoots, and the requirement on obtaining the precise value of initial error can be removed compared with the existing PPC. To deal with uncertainties appearing in individual quadrotors, a state estimator-based minimal learning parameter (SE-MLP) observer capable of reducing online computational burden without offline training is developed, such that a chattering-free transient learning process against disturbances can be ensured. Additionally, by introducing a high-order sliding mode (HOSM) differentiator to replace the linear filter in recursive back-stepping design, fast and accurate time-derivatives of virtual control signals can be generated to further improve the control performance. By fusion of MPPC, SE-MLP, HOSM as well as Lyapunov stability analysis, an enhanced neural guaranteed performance formation maneuvering control scheme for networked uncertain quadrotors is obtained such that position consensus without any overshoots can be achieved. Simulation results are offered to verify the effectiveness of the proposed approach.