| Abstract: | In this paper, we establish some new sufficient conditions for the existence of peri- odic solutions and the global exponential stability of neural networks with time-varying delays by means of the method of variation of parameters and some analysis techniques. These conditions can be used to design globally exponentially stable and periodic oscil- latory neural networks with time-varying delays. This work has important significance in both theory and applications. In addition, two examples are given to illustrate the theory. |
