Robust Formation Fencing Control of Networked Agents for Multiple Noncooperative Targets

This paper proposes an adaptive neural network-based non-cooperative targets fencing control scheme for multi-agent systems subject to model uncertainties and external disturbances. The scheme enables the multi-agent system to fence multiple non-autonomous non-cooperative targets in a specified formation geometry. At first, the dynamics model of the agents and targets containing matched and mismatched uncertainties is constructed. To cope with the complex nonlinear uncertainties, the adaptive neural network is used to approximate the nonlinear terms of the system, and the composite uncertainties are dynamically estimated by the adaptive terms updated online. According to the Lyapunov stability theorem, the adaptive parameter update laws are derived, and the global stability of the system, as well as the boundedness of the tracking error, are proved. Finally, numerical simulations verify the effectiveness of the proposed scheme. The results show that the control scheme can achieve efficient fencing of non-cooperative targets.