两类神经网络的指数稳定性-基础数学专业论文.docxVIP

PAGE PAGE IV ABSTRACT The dynamics of neural network have been extensively investigated; The discussion of stability is a topic at modern time. In this paper, we mainly discuss the exponential stability of the discrete bidirectional associative memory (BAM) neural network and Cohen-Grossberg neural network without delays. The discussion on the first one basted on the basic inequality then the second one basted on constructing proper Lyapunov definite function. On the discussion of the stability of the discrete-time bidirectional associative memory neural network. We first discuss the existence of equilibrium point, in this paper, we construct proper norm for the demonstration. We proof the existence by Brouwer fixed point theory. On the discussion of the stability, there are many methods and technology such as constructing proper Lyapunov definite function. Using the Matrix theory and so on. In this paper I discuss the exponential stability by basic inequalities and norm proper without complex function theory and matrix theory .In the discussion of Cohen-Grossberg neural network. We could depart it into three parts: the Cohen-Grossberg neural network without delays; the Cohen-Grossberg neural network with discrete delay; the Cohen-Grossberg neural network distributed delay. first We shall demonstrate the existence of equilibrium point of Cohen-Grossberg neural network without delays,then we give three different criteria for the stability of Cohen-Grossberg neural network without delays, We demonstration then one by one, In the discussion of Cohen-Grossberg neural network with discrete delay and distributed delay,We construct proper Laypunov definition function by the feature of neural network to demonstrate the stability . When we discuss the convergence, We also pay attention to the extent 安徽大学硕士学位论文：两类神经网络的指数稳定性 of the convergence. As we know exponential convergence is a kind of fast convergence. We mainly discuss the exponential stability of discrete bidirecti