Periodically intermittent control strategies for $$\varvec{\alpha }$$α-exponential stabilization of fractional-order complex-valued delayed neural networks
Periodically intermittent control strategies for $$\varvec{\alpha }$$α-exponential stabilization of fractional-order complex-valued delayed neural networks
复制标题
DOI：
10.1007/s11071-018-4053-0
复制
发表时间：
2018-02
影响因子：
5.6
通讯作者：
P. Wan;Jigui Jian;Jun Mei
中科院分区：
文献类型：
--
作者：
P. Wan;Jigui Jian;Jun Mei
本文研究了一类时滞分数阶神经网络在复值域的全局指数稳定性。为此，建立了几个有用的分数阶微分不等式，它们概括并改进了现有结果。然后，提出了一种合适的时滞周期性间歇控制方案，以实现所解决网络的全局指数稳定，其中反馈控制作为特例。利用这些有用的分数阶微分不等式并结合Lyapunov方法和其他不等式技术，获得了一些新颖的实值代数不等式方面的延迟无关准则，以确保所讨论的网络的全局指数稳定性，在实践中实现起来非常简单，并且避免了计算复杂的矩阵不等式。最后，通过模拟示例验证了理论标准的可用性。
This paper studies the global-exponential stabilization of a kind of fractional-order neural networks with time delay in complex-valued domain. To end this, several useful fractional-order differential inequalities are set up, which generalize and improve the existing results. Then, a suitable periodically intermittent control scheme with time delay is put forward for the global-exponential stabilization of the addressed networks, which include feedback control as a special case. Utilizing these useful fractional-order differential inequalities and combining with the Lyapunov approach and other inequality techniques, some novel delay-independent criteria in terms of real-valued algebraic inequalities are obtained to ensure global-exponential stabilization of the discussed networks, which are very simple to implement in practice and avert to calculate the complex matrix inequalities. Finally, the availability of the theoretical criteria is verified by an illustrative example with simulations.