New H∞ Controller Design for Neural Networks with Interval Time-Varing Delays in State and Observation

  • Authors:
  • Le A. Tuan;Phan T. Nam;Vu N. Phat

  • Affiliations:
  • Department of Mathematics, College of Sciences, Hue University, Hue, Vietnam;Department of Mathematics, Qui Nhon University, Binh Dinh, Vietnam;Institute of Mathematics, VAST, Hanoi, Vietnam 10307

  • Venue:
  • Neural Processing Letters
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

This article addresses the H  control problem of delayed neural networks, where the state input and observation output contain interval non-differentiable time-varying delays. Based on constructing a new set of Lyapunov---Krasovskii functionals, new delay-dependent sufficient criteria for H  control are established in terms of linear matrix inequalities. The Lyapunov---Krasovskii functional is mainly based on the information of the lower and upper delay bounds, which allows us to avoid using additional free-weighting matrices and any assumption on the differentiability of the delay function. The obtained condition is less conservative because of the technique of designing state feedback controller. The H  controller to be designed must satisfy some exponential stability constraints on the closed-loop poles. A numerical example is given to illustrate the effectiveness of our results.