Security event-triggered control for Markovian jump neural networks against actuator saturation and hybrid cyber attacks

This paper studies the problem of event-triggered control for the class of Markovian jump neural networks (MJNNs) under actuator saturation and hybrid cyber attacks. In order to save the limited network bandwidth, the event-triggered mechanism (ETM) is introduce to determine whether the signal of sampler is transmitted to the remote controller through the communication network. With the aid of two sets of Bernoulli distributed random variables (BDRVs), the mathematical model of randomly occurring deception attacks (RODAs) is presented. Due to the limitations of security and technology factors and the complex network environment in practice, actuator saturation and denial-of-service (DoS) attack are also considered. In summary, the MJNNs, ETM, actuator saturation and hybrid cyber attacks are incorporated into a unified construction, and a augmented system under this construction is modeled for the first time. For this system, the existence conditions of event-triggered control are derived through LyapunovKrasovskii functional (LKF). Based on this sufficient condition, the linear matrix inequality (LMI) technique is utilized to obtain the control gain of the controller and the weight matrix of the trigger. Finally, a numerical example is given to verify the effectiveness of the proposed method in this paper. (C) 2021 The Franklin Institute. Published by Elsevier Ltd. All rights reserved.