Salmanpour Yasaman, Arefi Mohammad Mehdi, Khayatian Alireza, Yin Shen
IEEE Trans Neural Netw Learn Syst. 2024 Oct;35(10):14534-14543. doi: 10.1109/TNNLS.2023.3279890. Epub 2024 Oct 7.
In this article, an adaptive neural containment control for a class of nonlinear multiagent systems considering actuator faults is introduced. By using the general approximation property of neural networks, a neuro-adaptive observer is designed to estimate unmeasured states. In addition, in order to reduce the computational burden, a novel event-triggered control law is designed. Furthermore, the finite-time performance function is presented to improve the transient and steady-state performance of the synchronization error. Utilizing the Lyapunov stability theory, it will be shown that the closed-loop system is cooperatively semiglobally uniformly ultimately bounded (CSGUUB), and the followers' outputs reach the convex hull constructed by the leaders. Moreover, it is shown that the containment errors are limited to the prescribed level in a finite time. Eventually, a simulation example is presented to corroborate the capability of the proposed scheme.
本文介绍了一类考虑执行器故障的非线性多智能体系统的自适应神经包容控制。利用神经网络的通用逼近特性，设计了一种神经自适应观测器来估计不可测状态。此外，为了减轻计算负担，设计了一种新颖的事件触发控制律。进一步提出了有限时间性能函数，以改善同步误差的暂态和稳态性能。利用李雅普诺夫稳定性理论，将证明闭环系统是协同半全局一致最终有界的（CSGUUB），并且跟随者的输出达到由领导者构建的凸包。此外，还表明包容误差在有限时间内被限制在规定水平。最后，给出了一个仿真例子来证实所提方案的能力。
