Distributed adaptive command filtered resilient event-triggered secure consensus control for multiagent systems under double DoS attacks

The multiagent network system has been widely developed because of its advantages in distributed computing capacity and network characteristics. By considering the network characteristics of multiagent systems, this paper proposes an anti-denial of service (DoS) adaptive command filtered resilient event-triggered secure consensus control approach, where the DoS adversary attacks both the controller and the communication channels. Because each agent has its distributed computing capacity and there are several agents in the system, the proposed control strategy has the feature of low computational burden theoretically, which mainly embodies in the following aspects. First, this work proposes an adaptive neural networks (NNs) switched observer to obtain the system states when the DoS attacker is active. By converting the state equation and combining the NNs method, the developed novel adaptive NNs switched observer can avoid calculating numerous linear matrix inequalities (LMIs) in the controller design process. Second, a command filter is introduced to obtain the derivatives of the virtual control functions, which reduces the computational burden produced from computing the derivatives and solves the problems of singularity and explosion of complexity (EOC). Moreover, the error compensation signals are designed since the filtering errors become larger when the DoS attacker is active. Third, by combining the consensus tracking errors and observer errors, a novel event -triggered function is developed to diminish the computational burden generated by the uninterrupted updates of control signals. Finally, simulation results demonstrate that the proposed control strategy can achieve the secure consensus control goal with certain tracking errors.