参数化复杂度
人工神经网络
平衡点
同胚(图论)
独特性
指数稳定性
数学
理论(学习稳定性)
应用数学
类型(生物学)
计算机科学
控制理论(社会学)
拓扑(电路)
非线性系统
算法
数学分析
微分方程
人工智能
离散数学
机器学习
物理
组合数学
生物
控制(管理)
量子力学
生态学
作者
Xian Zhang,Yantao Wang,Xin Wang
标识
DOI:10.1016/j.neucom.2021.08.068
摘要
For a class of neutral-type Cohen–Grossberg neural networks with multiple discrete and neutral delays, the existence and uniqueness of equilibrium point are derived by the homeomorphism mapping theory between topology spaces. By proposing a direct parameterized approach that is based on a parameterized estimation of solutions, it is the first time to investigate a sufficient condition guaranteeing that the unique equilibrium point is globally exponentially stable. The stability condition contains only some very simple inequalities, which is easily solved by applying the toolbox YALMIP of MATLAB.Three numerical examples in literature are employed to demonstrate the effectiveness of the obtained criterion over the previously achieved ones. In addition, the proposed approach is different from the popular linear matrix inequality approach, since no Lyapunov–Krasovskii functional is required. Therefore, the obtained criterion can be evaluated as an important contribution to the stability issue of the considered neural networks. It is potential that the proposed approach is applied to stability issues of various types of neural networks.
科研通智能强力驱动
Strongly Powered by AbleSci AI