A novel summation inequality in the stability analysis of neural
networks
Xin-Ge Liu Mei-Lan Tang
∗
School of Mathematics and Statistics, Central South University
Changsha, Hunan 410083, China
Abstract
In this paper, based on energy function and mathematical analysis techniques, a new
summation inequality is obtained which extend the Jensen inequality.
Keywords: Jensen inequality; energy function; Wirtinger inequality; summation
1 Introduction
Lyapunov-type inequalities[1-3] and the Jensen inequality have become two powerful math-
ematical tools for stability analysis of time-delay systems. The conservatism of the Jensen
inequality has been analyzed in [4] using the Gruss inequality. An alternative inequality
reducing the gap of the Jensen inequality has been proposed in [5] based on the Wirtinger in-
equality. The prop osed inequality in [5] has been successfully applied to the stability analysis
of various time-delay systems [6-7]. Recently, a novel integral inequality has been developed in
[8] which encompasses the Jensen inequality. However, these inequalities are involved integral.
In order to investigate the stability of discrete delay systems, a new summation inequality is
given in this paper.
∗
Corresp onding Author. E-mail address: csutmlang@163.com
1