In this paper, the author analyzes further problems of global stability for a class of cellular neural networks with delays by means of the Lyapunov functional method, inequalities $a^2{+b}^2>~2ab$ and $a^3{+b}^3{+c}^3>~3abc\hspace{0.5em}(a,b,c>~0)$ analysis technique, some stability criteria are obtained under more general conditions. These criteria can be applied to design globally stable networks and thus have important significance in both theory and applications.

• Received 12 October 1998

DOI:https://doi.org/10.1103/PhysRevE.59.5940