Sampled-data Synchronization of Quaternion-valued Neural Networks by Non-decomposition Method

This paper studies sampled-data control of quaternion-valued neural networks(QVNNs) with variable sampling, based on the quaternion matrices nondecomposition method. Firstly, the synchronization error model of quaternion-valued neural networks is established. Then, by adopting a refined input delay approach, and structuring a discontinuous Lyapunov function, the quaternion-valued LMI synchronization criteria for QVNNs are obtained. Furthermore, the proposed quaternion-valued LMIs are transformed into complexvalued LMIs equivalently. In addition, the event-trigger synchronization criteria are easily obtained by introducing the event-trigger scheme. Finally, the obtained results are demonstrated by a numerical simulation.