Linear matrix inequality-based criteria for exponential robust stability of Cohen–Grossberg-type bidirectional associative memory neural networks with delays

Abstract

The issue of global exponential robust stability is discussed for Cohen–Grossberg-type (CG-type) bidirectional associative memory (BAM) neural networks with delays. The activation functions that are adopted contain sigmoid functions and Lipschitz functions. For CG-type BAM neural networks with parameter uncertainties, which are assumed to be time invariant and bounded, by employing a Lyapunov–Krasovskii functional and linear matrix inequality (LMI) approach, the conditions ensuring global exponential robust stability are derived, which are expressed in terms of LMIs, and can be checked easily using the MATLAB LMI toolbox. In addition, when parameter uncertainties vanish, global exponential stability as a byproduct of global exponential robust stability, can also be guaranteed. Finally, two examples are provided to illustrate the effectiveness of the obtained results.