Title :
New LMI-Based Stability Condition for Neutral-Type Neural Networks with Variable and Distributed Delays
Author_Institution :
Dept. of Math., Dalian Jiaotong Univ., Dalian, China
Abstract :
This paper investigates the existence, uniqueness and stability for neutral-type neural networks with variable and unbounded distributed delays. Without assuming the differentiability, boundedness and monotonicity of the activation functions, by dividing the variation interval of the time delay into three subintervals with equal length, new LMI-based delay-dependent stability criterion is obtained by involving some free-weighting matrices.
Keywords :
delays; linear matrix inequalities; neural nets; transfer functions; LMI-based stability criterion condition; activation functions; free-weighting matrices; neutral-type neural networks; time delay; unbounded distributed delays; variable delays; Artificial neural networks; Asymptotic stability; Delay; Linear matrix inequalities; Stability criteria; Symmetric matrices;
Conference_Titel :
Intelligent System Design and Engineering Application (ISDEA), 2010 International Conference on
Conference_Location :
Changsha
Print_ISBN :
978-1-4244-8333-4
DOI :
10.1109/ISDEA.2010.42
