Lur'e systems with multilayer perceptron and recurrent neural networks: absolute stability and dissipativity

  title={Lur'e systems with multilayer perceptron and recurrent neural networks: absolute stability and dissipativity},
  author={Johan A. K. Suykens and Joos Vandewalle and Bart De Moor},
  journal={IEEE Trans. Autom. Control.},
Sufficient conditions for absolute stability and dissipativity of continuous-time recurrent neural networks with two hidden layers are presented. In the autonomous case this is related to a Lur'e system with multilayer perceptron nonlinearity. Such models are obtained after parametrizing general nonlinear models and controllers by a multilayer perceptron with one hidden layer and representing the control scheme in standard plant form. The conditions are expressed as matrix inequalities and can… 

Absolute stability and dissipativity of continuous time multilayer recurrent neural networks

  • J. SuykensJ. Vandewalle
  • Mathematics
    Proceedings of 1997 IEEE International Symposium on Circuits and Systems. Circuits and Systems in the Information Age ISCAS '97
  • 1997
In this paper we present a sufficient condition for global asymptotic stability of continuous time multilayer recurrent neural networks with two-hidden layers. The condition is based on a

Continuous time NLq theory: absolute stability criteria

  • J. SuykensJ. Vandewalle
  • Computer Science, Mathematics
    IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339)
  • 1999
We present absolute stability (global asymptotic) criteria for continuous time multilayer recurrent neural networks with two hidden layers. Such forms arise when considering recurrent neural models

Robust local stability of multilayer recurrent neural networks

A condition for robust local stability of multilayer recurrent neural networks with two hidden layers is derived from linking theory about linearization, robustness analysis of linear systems under nonlinear perturbation and matrix inequalities and dynamic backpropagation is discussed.

Stability analysis of layered digital dynamic networks using dissipativity theory

  • N. NguyenM. Hagan
  • Mathematics
    The 2011 International Joint Conference on Neural Networks
  • 2011
Using dissipativity theory, conditions for the globally asymptotic stability of equilibrium points of Layered Digital Dynamic Networks (LDDNs), a very general class of recurrent neural networks, are found.

Nonlinear system identification using discrete-time recurrent neural networks with stable learning algorithms

  • Wen Yu
  • Computer Science, Mathematics
    Inf. Sci.
  • 2004

Discrete-time nonlinear system identification using recurrent neural networks

  • Wen YuXiaoou Li
  • Mathematics, Computer Science
    42nd IEEE International Conference on Decision and Control (IEEE Cat. No.03CH37475)
  • 2003
It is concluded that for discrete-time nonlinear system identification, the gradient descent law and the backpropagation-like algorithm for the weights adjustment are stable in the sense of L/sub /spl infin// and robust to any bounded uncertainties.

An ℋ∞ approach to stability analysis of switched Hopfield neural networks with time-delay

This paper proposes a new ℋ∞ weight learning law for switched Hopfield neural networks with time-delay under parametric uncertainty. For the first time, the ℋ∞ weight learning law is presented to not

A Stability Condition for Neural Network Control of Uncertain Systems

A stability condition for neural network control systems in which the parameters of the controlled systems are uncertain is derived and can be imposed in training processes to guarantee the stability of the control systems.

A novel approach to stability analysis of multivariable Lur'e systems

  • Chun-Mo LeeJ. Juang
  • Mathematics
    IEEE International Conference Mechatronics and Automation, 2005
  • 2005
A new approach via the real structured singular value analysis is proposed to test for the absolute stability of multivariable Lur'e systems. The approach presented in this paper holds for any



Complete controllability of continuous-time recurrent neural networks

The stability of nonlinear dissipative systems

This short paper presents a technique for generating Lyapunov functions for a broad class of nonlinear systems represented by state equations. The system, for which a Lyapunov function is required,

Gradient methods for the optimization of dynamical systems containing neural networks

An extension of the backpropagation method, termed dynamic backpropagation, which can be applied in a straightforward manner for the optimization of the weights (parameters) of multilayer neural

A family of nonlinear H∞-output feedback controllers

State-space formulas are derived for a family of controllers solving the nonlinear H/sup /spl infin//-output feedback control problem by interconnecting the "central controller" with an asymptotically stable, free system having L/sup 2/-gain /spl les//spl gamma/.

Dissipative dynamical systems Part II: Linear systems with quadratic supply rates

This paper presents the theory of dissipative systems in the context of finite dimensional stationary linear systems with quadratic supply rates. A necessary and sufficient frequency domain condition

L/sub 2/-gain analysis of nonlinear systems and nonlinear state-feedback H/sub infinity / control

Previously obtained results on L2-gain analysis of smooth nonlinear systems are unified and extended using an approach based on Hamilton-Jacobi equations and inequalities, and their relation to

H∞ control for nonlinear systems with output feedback

The basic question of nonlinear H/sup infinity / control theory is to decide, for a given two-port system, when feedback that makes the full system dissipative and internally stable exists. This

Nonlinear system identification using neural state space models, applicable to robust control design

A linear fractional transformation representation is given for neural state space models, which makes it possible to use these models, obtained from input/output measurements on a plant, in a standard robust performance control scheme.