网络攻击下一类非线性切换多智能体系统的安全自适应控制

针对一类控制方向未知的非线性切换多智能体系统, 本文研究了在不确定网络攻击下的安全控制问题. 网 络攻击破坏传感器真实数据, 导致系统真实信息无法获取且不能直接用于控制设计. 为此, 通过受攻击状态构建一 组新颖辅助变量来消除网络攻击造成的影响. 此外, 所研究的系统包含更一般的不确定性, 即未知控制方向, 未知常 值参数以及不确定攻击. 这些不确定性在设计过程中相互耦合. 利用Nussbaum型函数并设计自适应律补偿耦合后 的不确定性, 极大降低了系统复杂性. 通过系统性地迭代构造出共同Lyapunov函数, 提出了一套安全自适应控制方 法, 保证了受攻击系统在任意切换下达到渐近输出一致性. 最后, 数值仿真验证了该方法的有效性.

Secure adaptive consensus control against cyber attacks for nonlinear switched multi-agent systems

In this paper, the secure adaptive consensus control problem is considered for a class of uncertain nonlinear multi-agent systems under unknown control direction against uncertain cyber attacks. Cyber attacks destroy the real data of sensors, so that the real information of the system cannot be obtained and cannot be directly used for control design. In this paper, a new set of auxiliary variables is constructed to eliminate the impact of cyber attacks. Furthermore, the considered systems include more general uncertainties, namely, the unknown control direction, the uncertain constant parameters and the uncertain cyber attacks. These uncertainties are coupled together during the control design process. To tackle the coupled uncertainties, Nussbaum-type functions are used and an adaptive law is designed, which immensely reduces the system complexity. By constructing a common Lyapunov function for all subsystems, a secure protocol is developed to guarantee that all the outputs of the multi-agent systems achieve the asymptotical consensus against cyber attacks under arbitrary switching. Finally, an example is presented to verify the effectiveness of the proposed method.