Relaxed dissipativity conditions of neural networked with time-varying delay via generalized free-weighting-matrix approach
Hong‐Hai Lian, Shenping Xiao, Hong‐Bing Zeng, Xiaohu Zhang, Gang Chen · 2017
This paper focuses on the robust delay-dependent dissipativity problem of neural networks with a bounded time-varying delays. Firstly, a proper augmented Lyapunov-Krasovskii functional, which doesn't demand all the symmetric matrices to be positive definite, is constructed. In addition, by employing relaxed integral inequality and the combining generalized free-weighting-matrix (GFWM) approach to bound integral term in the derivative of the Lyapunov-Krasovskii functional, some less conservative sufficient conditions are deduced to ensure that considered neural networks are strictly (Q, S, R)-γ-dissipative. The advantages and effectiveness of presented techniques are demonstrated via two numerical examples.