Neutral-Type CGNNs With Multiple Delays: A Frobenius Norm Approach to Stability

For neutral Cohen-Grossberg neural networks incorporating multiple time delays, we construct a tailored Lyapunov function to derive stability criteria independent of delay magnitudes. Despite the inherent challenge of avoiding vector-matrix representation due to delay multiplicity, this work introduces a pioneering sufficient condition based on parameter matrix Frobenius norms and network parameters. Owing to the method’s feasibility and simplicity, our stability finding not only decreases computational complexity but also lowers conservativeness when compared with several prior studies. The validity and superiority of the findings in this research are confirmed by two specific neural network models.