Projective lag synchronization of full quaternion-valued stochastic BAM neural networks with generalized delays
摘要
Synchronization issues are key research topics in the theory of stochastic neural networks. This paper is committed to investigating the fixed and preassigned-time projective (FITP and PATP for short) lag synchronization of stochastic quaternion-valued BAM neural networks (SQV-BAMNNs) with generalized delays. A generalized form of time delays, encompassing both discrete and proportional delays, is introduced. Subsequently, without employing the quaternion separation method, two discontinuous control strategies based on 1-norm and 2-norm are designed to achieve FITP/PATP lag synchronization of SQV-BAMNNs. Next, by constructing different Lyapunov functions and leveraging the quaternion-valued Itô’s differential formula, novel criteria for FITP/PATP lag synchronization of SQV-BAMNNs are derived. At last, the rationality of theoretical analysis is further validated by numerical simulation.