ABSTRACT This paper presents a gain scheduling approach for achieving the consensus tracking of multi-agent systems with actuator saturation. We first construct a series of nesting ellipsoid invariant sets associated with consensus errors. When the consensus errors stay between the two ellipsoid invariant sets, the feedback gains keep constant, but when the consensus errors enter into the smaller ellipsoid invariant set, the feedback gains abruptly become larger. By combining this gain scheduling technique and the parametric Lyapunov equations, we, respectively, design state and output feedback gain scheduling protocols. Their main advantage, in comparison with the fixed case, is that the convergence rate of consensus tracking can be enhanced by scheduling the gain parameters. Numerical simulations verify the effectiveness of theoretical analysis.

中文翻译：

---

获得受执行器饱和影响的多智能体系统的调度共识

摘要 本文提出了一种增益调度方法，用于实现对执行器饱和的多智能体系统的一致性跟踪。我们首先构建一系列与共识错误相关的嵌套椭球不变集。当共识误差停留在两个椭球不变量集之间时，反馈增益保持不变，但当共识误差进入较小的椭球不变量集时，反馈增益突然变大。通过结合这种增益调度技术和参数李雅普诺夫方程，我们分别设计了状态和输出反馈增益调度协议。与固定情况相比，它们的主要优点是可以通过调度增益参数来提高共识跟踪的收敛速度。数值模拟验证了理论分析的有效性。