Weighted Pseudo-almost Periodic Solution And Weighted Pseudo-almost Automorphic Solution For Neural Networks On Time Scales

In this paper, we introduce the concept of weighted pseudo-almost periodic function on time scales and study some property of this function, obtained that the existence of weighted pseudo-almost periodic solution of dynamic equation on time scales, for applica-tion we consider the existence and global exponential stability of neural networks (include neutral type neural networks on time scales), obtain the conclusion that if neural networks on time scales satisfy some condition, assumed that external input function be almost pe-riodic function, pseudo-almost periodic function and weighted pseudo-almost periodic function respectively, then the neural networks respectively have an almost periodic solu-tion, pseudo-almost periodic solution and weighted pseudo-almost periodic solution, and this solution is global exponential stability. At last, we study the existence and global ex-ponential stability of weighted pseudo-almost automorphic solution for neural networks on time scales. Similar with the almost periodic type function, we get that if neural networks on time scales satisfy some condition, the external input function be almost automorphic function, pseudo-almost automorphic function and weighted pseudo-almost automorphic function is sufficient condition for the existence and global exponential stability of almost automorphic solution, pseudo-almost automorphic solution and weighted pseudo-almost automorphic solution of neural networks on time scales.