时滞细胞神经网络的代数抗饱和补偿设计

摘要

This paper proposes an algebraic anti-saturation compensation design for time-delay chaotic cellular neural networks and presents an algebraic condition of control for time-delay chaotic cellular neural networks without saturable input according to M-matrix theory,Dini derivative and system pa-rameters.When the input is saturable,QMI control criterion can be obtained,from the inequality scaling principle,zoned discussion and maximum synthesis and Lyapunov theory and thus changed in-to LMI criterion.And meanwhile the attraction domain and its optimization are offered.The simula-tion results verify the effectiveness and feasibility of the proposed method.%针对时滞细胞神经网络，提出了一种基于代数判据的抗饱和补偿设计。运用 M 矩阵理论、迪尼导数、系统参数设计了无饱和输入限制系统镇定的代数判据。在此基础上，利用不等式放缩原理、分区讨论与大中取大综合方法和 Lyapunov 理论，得出了二次矩阵不等式形式的抗饱和控制设计方案，进而将之转化为线性矩阵不等式。同时，给出了吸引域及其最大化方法。仿真验证了该设计方案的可行性和有效性。