Study On Almost Periodic Function-Like Solutions For Several Neural Networks On Time Scales

In this dissertation,we will mainly research the existence and stability of almost periodic function-like solutions for several neural networks on time scales.For the given types of quaternion-valued neural networks and Clifford-valued neural networks,we consider that the self-feedback coefficients are quaternions or Clifford number.This dissertation provides effective methods for almost periodic function-like solutions of the considered neural networks.In addition,we study the basic properties of Sp-almost periodic functions and Sp-pseudo almost periodic functions on time scales.Furthermore,we base on the definition of fuzzy operation on the quaternion and proved its related properties.Without decomposing the considered neural networks into real-valued systems,base on the B anach fixed point theorem,the theory of calculus and inequality analysis techniques on time scales,we first consider quaternion-valued shunting inhibitory cellular neural networks of neutral type with time delays and quaternion-valued high-order Hopfield neural networks with time-varying delays and leakage delays on time scales,we obtain some sufficient conditions for the existence and global exponential stability of almost periodic solutions and pseudo almost periodic solutions for these two types of networks neural networks.Then,the existence and exponential stability of Sp-almost periodic solutions for a class of quaternion-valued fuzzy cellular neural networks with time-varying delays on time scales are considered,and we study the problem of quaternion-valued neural networks with time-varying delays on time scales.Next,the problem of the Sp-almost periodic synchronization of Clifford-valued cellular neural networks on time scales are studied.Subsequently,we consider a class of neutral Clifford-valued neural networks with leakage delays on time scales,and study the existence and stability of Sp-pseudo almost periodic solutions of the considered neural networks.Finally,we give numerical examples and simulations to illustrate the feasibility and effectiveness of our results.