Global Asymptotic Stability for Hopfield-Type Neural Networks With Diffusion Effects
-
摘要: 对具有扩散影响的Hopfield型神经网络平衡点的存在唯一性和全局渐近稳定性进行了研究.在激活函数单调非减、可微且关联矩阵和Liapunov对角稳定矩阵有关时,利用拓扑度理论得到了系统平衡点存在的充分条件.通过构造适当的平均Liapunov函数,分析了系统平衡点的全局渐近稳定性.所得结论表明系统的平衡点(如果存在)是全局渐近稳定的而且也蕴含着系统的平衡点的唯一性.
-
关键词:
- 扩散 /
- Hopfield型神经网络 /
- 平衡点 /
- 全局渐近稳定性
Abstract: The existence,uniqueness and global asymptotic stability of the equilibrium for Hopfield-type neural networks with diffusion were discussed.The sufficient conditions of the existence and uniqueness of the equilibrium of the system were obtained by applying the topological degree theory when the activation functions are monotonous non-decreasing and differential,and the interconnected matrix is related to the Lianupov diagonal stable matrix.By constructing the average Liapunov functions,the global asymptotic stability of the equilibrium of the system was obtained.-
Key words:
- diffusion /
- neural networks /
- equilibrium /
- global asymptotic stability
-
[1] Hopfield J J.Neural networks and physical systems with emergent collective computional abilities[J].Proc Natl Acda Sci USA,1982,79(2):2554-2558. doi: 10.1073/pnas.79.8.2554 [2] Hopfield J J. Neurons with graded response have collective computional properties like those of two-state neurons[J].Proc NatlAcda Sci USA,1984,81(10):3088-3092. doi: 10.1073/pnas.81.10.3088 [3] 廖晓昕.Hopfield神经网络的稳定性[J].中国科学(A集),1993,23(10):1032-1035. [4] 梁学斌,吴立德 .Hopfield型神经网络的全局指数稳定性及应用[J].中国科学(A集),1995,25(5):523-532. [5] Forti M, Tesi A. New conditions for global stability of neural networks with application to linear and quadratic programming problems[J].IEEE Trans Circuits Systems,1995,42(7):354-366. doi: 10.1109/81.401145 [6] LIANG Xue-bin,WU Li-de.Comment on “New conditions for global stability of neural networks with application to linear and quadratic programming problems”[J].IEEE Trans Circuits Systems,1997,44(1):1099-1101. doi: 10.1109/81.641813 [7] LIANG Xue-bin,WANG Jun.Absolute exponential stability of neural networks with a general class of activation functions[J].IEEE Trans Circuits Systems,2000,47(8):1258-1263. doi: 10.1109/81.873882 [8] HE Qi-ming,KANG Li-shan.Existence and stability of global solution for generalized Hopfield neural networks[J].Neural Parallel & Scientific Computations,1994,2(2):165-176. [9] Carpenter G A. A Geometric approach to singular perturbation to never impulse equations[J].J Differential Equations,1977,23(3):355-367. [10] Evans J W. Nerve axon egus II: Stability at rest[J].Indian Univ Math J,1972,22(1):75-90. doi: 10.1512/iumj.1972.22.22009 [11] 廖晓昕,杨叔子,程时杰,等.具有反应扩散的广义神经网络的稳定性[J].中国科学(E集),2002,32(1):88-94. [12] CAO Jin-de,WANG Jun.Absolute exponential stability of recurrent neural networks with Lipschitz-continuous activation and time delays[J].Nueral Networks,2004,17(3):379-390. doi: 10.1016/j.neunet.2003.08.007 [13] 郭大钧,孙经先,刘兆理. 非线性常微分方程的泛函方法[M].济南: 山东科技出版社,1995. [14] 胡适耕. 非线性分析和方法[M]. 武汉: 华中理工大学出版社,1996. [15] 郭大钧. 非线性泛函分析[M]. 济南: 山东科技出版社,2002. [16] Fiedler M.Special Matrices and Their Applications in Numerical Mathematics[M].Dordrecht:Martinus Nijhoff,1986. [17] Horn R A,Johnson C R.Topics in Matrix Analysis[M].Cambridge:Cambridge University Press,1991.
点击查看大图
计量
- 文章访问数: 2862
- HTML全文浏览量: 135
- PDF下载量: 829
- 被引次数: 0