This paper addresses the global asymptotic stabilization of delayed fractional complex-valued neural networks(FCVNNs) subject to bounded parameter uncertainty. The problem is proposed for two reasons: 1) The availablemethods for uncertain dynamical systems may be too conservative; 2) The existing algebraic conditions willlead to huge computational burden for large-scale FCVNNs. To surmount these difficulties, the delayed FCVNNswith interval parameters are transformed into a tractable form at first. Then, a simple and practical controller–linearstate feedback controller is designed to achieve the global asymptotic stabilization. By constructing different Lyapunovfunctions and utilizing the fractional-order comparison principle and interval matrix method, two sufficientglobal asymptotic stabilization criteria expressed in LMI forms, are established. The obtained results in this paperimprove and extend some previous published results on FCVNNs. Finally, two numerical examples are provided toillustrate the correctness of the theoretical results.
This paper addresses the global asymptotic stabilization of delayed fractional complex-valued neural networks(FCVNNs) subject to bounded parameter uncertainty. The problem is proposed for two reasons: 1) The availablemethods for uncertain dynamical systems may be too conservative; 2) The existing algebraic conditions willlead to huge computational burden for large-scale FCVNNs. To surmount these difficulties, the delayed FCVNNswith interval parameters are transformed into a tractable form at first. Then, a simple and practical controller–linearstate feedback controller is designed to achieve the global asymptotic stabilization. By constructing different Lyapunovfunctions and utilizing the fractional-order comparison principle and interval matrix method, two sufficientglobal asymptotic stabilization criteria expressed in LMI forms, are established. The obtained results in this paperimprove and extend some previous published results on FCVNNs. Finally, two numerical examples are provided toillustrate the correctness of the theoretical results.