부산대학교 도서관

This dissertation considers the stability issue for generalized neural networks (GNNs) subject to mixed delays. Firstly, an appropriate Lyapunov-Krasovskii functionals (LKF) is constructed by containing two new delay-product-type (DPT) terms, augmented terms and double integral terms. Then, to cooperate with the established LKF effectively, an improved finite-interval quadratic polynomial inequality is utilized to handle the quadratic function negative-determination. As a consequence, a new less conservative stability condition is achieved. Finally, some well-studied simulation examples are included to show the superiority of the presented conditions.