Abstract
This paper discusses the adaptive event-triggered synchronization problem of a class of neural networks (NNs) with time-varying delay and actuator saturation. First, in view of the limited communication channel capacity of the network system and unnecessary data transmission in the NCSs, an adaptive event-triggered scheme (AETS) is introduced to reduce the network load and improve network utilization. Second, under the AETS, the synchronization error model of the delayed master-slave synchronization system is constructed with actuator saturation. Third, based on Lyapunov–Krasovskii functional (LKF), a new sufficient criterion to guarantee the asymptotic stability of the synchronization error system is derived. Moreover, by solving the stability criterion expressed in the form of a set of linear matrix inequalities (LMIs), some necessary parameters of the system are obtained. At last, two examples are expressed to demonstrate the feasibility of this method.
Highlights
neural networks (NNs) represent a computing type similar to that of the brain [1]
Because of network congestion and limited signal transmission speed between neurons, the time delay is common in NNs. erefore, the delay is introduced into NNs, and as a kind of complex nonlinear system, delayed NNs (DNNs) can exhibit complicated dynamic behaviors effectively and even chaotic phenomena, so it has received considerable attention
We will give the asymptotic stability criterion of the synchronization error system (17) using Lyapunov–Krasovskii functional (LKF) under the adaptive event-triggered scheme (AETS) in eorem 1. en considering the nonlinear terms in eorem 1, we will give its explicit expression in eorem 2
Summary
NNs represent a computing type similar to that of the brain [1]. they have been widely applied in many fields, such as secure communications, combinatorial optimization, pattern recognition, associative memories, and complex system control [2,3,4,5,6,7,8]. Complexity and problem of L2-gain analysis of delayed chaotic NNs with actuator saturation, and by solving LMIs, the intermittent linear state feedback controller was obtained. In [17], considering the actuator saturation, sampled-data synchronization of chaotic NNs was discussed; the condition of keeping local synchronization of master-slave NNs was derived; and the relevant sampling data controllers can be obtained by employing LMI. In [20], on the premise of fully considering the influence of asynchronous preconditions caused by event-triggered sampling, the event-triggered master-slave synchronization problem of delayed T-S fuzzy neural networked systems was discussed. E synchronization error model is built on the framework of NCSs, where both actuator saturation and network-induced delay are considered. It can be divided into the following points. Notations: Rn means n-dimensional Euclidean space, Rm×n means m × n real matrices, I denotes the identity matrix, block diagonal matrix is represented by diag{. . .}, and ∗ is used to denote symmetry terms in a matrix
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call