One Step Anti-Windup Design for a Class of Time-Delayed Cellular Neural Networks
Miao Zha, Hanlin He · 2018
An algebraic anti-windup compensation design for a class of time-delayed cellular neural networks (CNNs) is considered in this paper. Based on Lyapunov theory, Schur complement principle and some lemma, the quadratic matrix inequality (QMI) criterion keeping system stable can be get by solving an exact Lyapunov function, which controller and compensator can be determined simultaneously. The proposed method is a one-step anti-windup strategy, it can get a more optimized solution than those methods which design controller and compensator separately. By some matrix transformation, the QMIcontrol criterion can be changed into linear matrix inequality (LMI) criterion then it can be easily solved by computer. The attraction domain and its optimization are also presented. The simulation results verify the effectiveness and feasibility of the proposed method.