摘要
对于具有时滞的广义神经网络,利用Liapunov函数方法、线性矩阵不等式以及积分不等式等技巧,给出了该神经网络模型的平衡点的存在性、惟一性以及全局指数稳定的一些充分条件。通过对这些结论和已有结果的比较,说明所得到的这些条件推广了已有文献中的结果,并得到了神经网络稳定性的新的条件。这些条件对于神经网络的设计具有指导作用。
New sufficient conditions of globally exponential stability of generalized neural networks with time delays were presented by using Liapunov algorithm, linear matrix inequality and integral inequality. The comparison between the current result and the previous result indicates that the newly- got conditions generalize the previous global exponential stability results derived in the literature and establish some new stability criteria for delayed neural networks. These conditions are of some practical significance in the design of the neural networks.
出处
《武汉理工大学学报（信息与管理工程版）》
CAS
2008年第2期193-196,共4页
Journal of Wuhan University of Technology：Information & Management Engineering
基金
中国博士后基金资助项目(20060400239)
关键词
广义神经网络
时滞
全局指数稳定
generalized neural networks
delays
globally exponential stability
