Finite-time synchronization of quaternion-valued fuzzy memristive neural networks by an non-separation method via event-triggered control.

In this paper, the event-triggered control method is applied to study a class of generalized delayed quaternion-valued fuzzy Memristive Neural Networks. By constructing a novel Lyapunov function and designing an appropriate event-triggered controller, new event-triggered conditions and synchronization criteria are established. To handle the inherent complexity of quaternion operations, a nonlinear scaling technique is employed to compare distinct quaternion components, while an improved quaternion norm together with a symbolic function enables a holistic analysis of the quaternion-valued fuzzy memristive neural networks without decomposing them into real-valued subsystems. The effectiveness of the proposed theoretical results is finally verified by numerical simulations.