Delay‐dependent stability analysis for neural networks with additive time‐varying delay components. Issue 3 (1st February 2013)

• Record Type: Journal Article
• Authors: Zhu, Xun‐Lin; Yue, Dong; Wang, Youyi
• Is Part Of: IET control theory & applications. Volume 7:Issue 3(2013)
• Page Start: 354
• Page End: 362
• Publication Date: 2013-02-01
• Subjects: computational complexity -- delays -- linear matrix inequalities -- Lyapunov methods -- neural nets -- stability -- stability criteria -- time‐varying systems; delay‐dependent stability analysis -- neural networks -- additive time‐varying delay components -- general Lyapunov functional -- conservative delay‐dependent stability criterion -- linear matrix inequalities -- computational complexity reduction -- slack variable elimination method; computational complexity -- delays -- linear matrix inequalities -- Lyapunov methods -- neural nets -- stability -- stability criteria -- time‐varying systems; delay‐dependent stability analysis -- neural networks -- additive time‐varying delay components -- general Lyapunov functional -- conservative delay‐dependent stability criterion -- linear matrix inequalities -- computational complexity reduction -- slack variable elimination method; Control theory -- Periodicals; Automatic control -- Periodicals; 629.8312
• DOI: 10.1049/iet-cta.2012.0585
• Languages: English
• ISSNs: 1751-8644

Abstract

Abstract : This article studies the problem of stability analysis for neural networks (NNs) with two additive time‐varying delay components. By taking both the independence and the variation of the two delay components into consideration, a more general Lyapunov functional is defined. By estimating the upper bound of the derivative of the Lyapunov functional more tightly, a less conservative delay‐dependent stability criterion is established in terms of linear matrix inequalities. To reduce the computational complexity, a method for eliminating slack variables is provided, and then a simplified stability criterion is obtained. Some numerical examples are given to illustrate the effectiveness of the proposed method and the significant improvement over the existing results.