Research On Stochastic Control Of Feedback Neural Network

The research of stochastic control of feedback neural network is an interdisciplinary between theoretical disciplines such as mathematics,stochastic process and modern control,and it has broad application prospects.With the development of science and technology,the application research field of feedback neural network becomes wider and wider,and the most important indicator of a system is its stability.Therefore,the stability of feedback neural network has attracted the attention of many scholars.Aiming at the problem of stochastic control of feedback neural network,discrete stochastic feedback is used to study the classic network form of feedback neural network-H?pfield neural network.The stochastically controlled H?pfield neural network becomes almost surely exponentially stable.The specific content includes:Firstly,the basic knowledge include the type of feedback neural network,Brownian motion and It?-type stochastic differential equation.The development process of It? theory is discussed.The model of continuous H?pfield neural network is described in detail.It can be represented by an ordinary differential equation.That process lays the foundation for the theoretical proof of the following.Secondly,the basic theory of stochastic systems and the theory of stochastic stabilization of deterministic differential systems are introduced.The principle of stochastic interference for calming deterministic differential systems and other effects of stochastic interference in addition to stabilization are discussed.Thirdly,the theorems,hypotheses,lemmas and proofs of stabilizing H?pfield neural network with discrete stochastic feedback are given.This problem select the stochastic feedback control H?pfield neural network based on discrete state observations,and it introduce an auxiliary system.Then,this problem use the comparison principle method to the stochastically controlled H?pfield neural network converge exponentially,and it obtain upper limit of the duration between consecutive observation points.Finally,the numerical simulation results of the stochastic controlled H?pfield neural network system are given.The discrete stochastic feedback is simulated by Matlab.The controlled H?pfield neural network with discrete stochastic feedback is solved based on the Euler method.That results show that the controlled system converges exponentially,which further validates the conclusions of the article.Discrete stochastic feedback stabilizes the unstable H?pfield neural network,and that achieves stochastic stabilization of the feedback neural networks.Figure 9;Table 3;Reference 56...