Fuzzy-Based Synchronization Control for Coupled Neural Networks Under Cyber Attacks via Intelligent Impulsive Algorithm

This paper studies the fuzzy-based synchronization control problem of coupled neural networks under cyber attacks, where the considered attacks can block the communication links between fuzzy sub-neural networks. Firstly, an improved fuzzy network model is developed, which takes into account the inherent vulnerabilities of network. We design a fuzzy logic-based event-triggered delayed impulsive controller, where impulsive signals are generated by a dependent-Lyapunov intelligent impulsive selection algorithm. Particularly, it can mitigate attack effects, ensure the desired performance of fuzzy networks, and effectively exclude the Zeno behavior. Then, based on the proposed algorithm, some delay-dependent synchronization criteria are established for fuzzy networks on different scales delays, respectively. Finally, a practical example about resistance-capacitance circuit network is provided in different scenarios to show the validity of the theoretical results. Note to Practitioners—This paper was motivated by existing results on impulsive synchronization control of neural networks. Most existing results ignore the effects of external cyber attacks and delay during signal transmission, which is quite difficult to simulate the practical network model. This paper constructs a more general model to consider the inherent vulnerabilities of networks. Then, an intelligent impulsive selection algorithm is designed to resist the risks of attacks and obtain the expected performance. The obtained results are applied to resistance-capacitance circuit systems to verify the effectiveness, and it is expected that the proposed approach can be extended to more mechanical systems.