Model-Free Adaptive Hierarchical Resilient Cloud Control for Multiagent Group Systems Under Aperiodic DoS Attacks
针对未知数学模型且计算能力有限的大规模异构多智能体系统，提出一种仅用输入输出数据的无模型自适应分层弹性云控制策略，通过补偿因子减轻双通道非周期性拒绝服务攻击的影响，云控制器承担所有计算负担以提升实时处理能力并降低硬件成本。
In this article, the large-scale networked heterogeneous multiagent systems (LSNH-MASs) under dual-channel aperiodic Denial-of-Service (DCA-DoS) attacks with unknown mathematical models and limited computing power are studied. In order to overcome the unknown mathematical model of the system and enhance the ability of consensus control of LSNH-MASs, a model-free adaptive hierarchical resilient cloud control (MFAHRCC) strategy, only using input and output data, is proposed. This strategy can reduce the negative impact of the DCA-DoS attacks by designing a compensation factor. The stability and consensus of LSNH-MASs in insecure network environment are strictly proved. Under the constructed cloud control framework, the computation burden of all agents is borne by the cloud controller, which not only improves the ability of real-time processing data of the system but also greatly reduces the hardware cost of agents. Finally, the effectiveness of the MFAHRCC algorithm proposed in this article is verified by simulation with comparisons.