【作者】 蒲兴成；

【Author】 Pu Xingcheng (Institute of Applied Mathematics Chongqing University of Posts and Telecommunications,Chongqing 400065)

【机构】 重庆邮电大学应用数学研究所；

【摘要】 研究了一类时变时滞不确定递归神经网络在平衡点的渐近稳定问题。在将相应模型转化为系统模型下,结合Lyapunov-Krasovskii稳定性定理,线性矩阵不等式(LMI)技巧和S过程等理论,得到一类时变时滞不确定递归神经网络在平衡点的全局渐近稳定的一个充分条件。该方法克服了已有方法的保守性且结果更具有一般性,因此,该方法推广了一些相关的结论。更多还原

【Abstract】 The delay-dependent stability of recurrent neural networks with uncertainties and multiple time-varying is investigated in this paper.By transforming the delayed neural model to system model,combing Lyapunov-Krasovskii stability theorem,linear matrix inequality(LMI)technique and S procedure,a new sufficient condition is derived,which is determined by the coefficient of the model for determining the globally asymptoticai stability of recurrent neural networks with time-delay and uncertainties.Compared with the reported methods,the proposed methods gave a much less conservative and more general results.Therefore,this method has improved some relating results.更多还原