时滞神经网络对于新的二次优化的应用  

Delayed Neural Network for Quadratic Optimization Applications

作　　者：张静文[1] 刘德友[1] 吴文娟[1] 

机构地区：[1]燕山大学理学院

出　　处：《郑州大学学报（理学版）》2012年第1期6-10,共5页Journal of Zhengzhou University:Natural Science Edition

基　　金：国家自然科学基金资助项目;编号70671088;71071133

摘　　要：研究了一类新的二次优化最优解稳定性问题,并在M为广义正定矩阵时给出了判定平衡点全局指数稳定的充分条件,通过时滞神经网络解线性投影方程的模型,借助Lyapunov函数给出系统全局稳定性的充分条件,最后用仿真实例说明研究结果的有效性.A kind of new quadratic optimization problem of the stability of optimal solutions was studied.Sufficient conditions were given to determine the equilibrium point for globally exponentially stable,which was considered as the generalized positive definite matrix M.Through the delayed neural network model for solving linear projection equations,new sufficient conditions for global stability of the system were given with the Lyapunov function.The final simulation showed the validity of research results.

关 键 词：时滞神经网络 投影方程 广义正定矩阵 全局指数稳定 最优解 

分 类 号：O175[理学—数学]