Dissipative Synchronization Control for Two-Time-Scale Markov Jump Neural Networks Subject to Redundant Channels: A Hidden-Markov-Model-Based Method

This work studies the synchronization issue for two-time-scale Markov jump neural networks subject to redundant channels. In such systems, mode information may not be directly available (e.g., packet loss), and traditional synchronous control methods cannot meet this challenge. The hidden Markov model can deal with the situation that the systems state cannot be accessed directly, and estimate the current state of the system through the “observation” mode, so as to improve the controller design and advance the stability and robustness of the systems. Therefore, the controller is designed based on a hidden Markov model for the above scenarios. Meanwhile, the redundant channels are built to reduce the influence of packet loss. Moreover, the two-time-scale phenomenon of the plant is considered by using the singular perturbation parameter. Then, the Lyapunov function construction is associated with the singular perturbation parameter and some sufficient conditions to guarantee the stability of the plant are obtained. Finally, the designed control law is available which is demonstrated by two illustrative examples.