噪声环境中时滞双向联想记忆神经网络指数稳定

任何系统实际上都是在噪声环境中进行工作的.对处在噪声强度已知的噪声环境下双向联想记忆(BAM)神经网络,其平衡点具有指数渐近稳定性是网络进行异联想记忆的基础.构造一个适当的Lyapunov函数,应用It^o公式、M矩阵等工具讨论了在噪声环境下具有时滞的BAM神经网络概率1指数渐近稳定,得到了指数稳定的代数判据和两个推论,此判据只需验证仅由网络参数构成的矩阵是M矩阵即可,给网络设计带来方便.本文所得结果包括相关文献中确定性结果作为特例.

Exponential stability of time-delay bi-direction associated memory neural networks in noisy environment

In reality,any system works in noisy environment.For bi-direction associated memory(BAM) neural networks in noisy environment,the disturbance intensity is estimated.It is the chief problem that the equilibrium of BAM neural networks should be exponentially stable.By constructing an appropriate Lyapunov function and by using It formula and M-matrix as analytic tools,the problem of exponential stability in probability one about noisy time-delay BAM neural networks is discussed,and some algebraic criteria are obtained.By those criteria,it is only necessary to verify the matrix to be M-matrix of the system's parameters,resulting in convenience in system design.The conclusions include those obtained in relevant literature as special cases.