Title
Dissipativity and quasi-synchronization for neural networks with discontinuous activations and parameter mismatches.
Abstract
In this paper, global dissipativity and quasi-synchronization issues are investigated for the delayed neural networks with discontinuous activation functions. Under the framework of Filippov solutions, the existence and dissipativity of solutions can be guaranteed by the matrix measure approach and the new obtained generalized Halanay inequalities. Then, for the discontinuous master–response systems with parameter mismatches, quasi-synchronization criteria are obtained by using feedback control. Furthermore, when the proper approximate functions are selected, the complete synchronization can be discussed as a special case that two systems are identical. Numerical simulations on the chaotic systems are presented to demonstrate the effectiveness of the theoretical results.
Year
DOI
Venue
2011
10.1016/j.neunet.2011.06.005
Neural Networks
Keywords
Field
DocType
Quasi-synchronization,Discontinuous activations,Filippov solutions,Matrix measure,Generalized Halanay inequalities
Synchronization,Mathematical optimization,Matrix (mathematics),Control theory,Chaotic systems,Artificial neural network,Mathematics,Special case
Journal
Volume
Issue
ISSN
24
10
0893-6080
Citations 
PageRank 
References 
47
1.66
9
Authors
4
Name
Order
Citations
PageRank
Xiaoyang Liu134010.74
Tianping Chen23095250.77
Jinde Cao311399733.03
Wenlian Lu437025.19