Multistability of Fractional-Order Recurrent Neural Networks With Discontinuous and Nonmonotonic Activation Functions
Open Access
- 16 August 2019
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 116430-116437
Abstract
The coexistence of multiple stable equilibria in recurrent neural networks is an important dynamic characteristic for associative memory and other applications. In this paper, the existence and local Mittag-Leffler stability of multiple equilibria are investigated for a class of fractional-order recurrent neural networks with discontinuous and nonmonotonic activation functions. By using Brouwer $^{\prime }\text{s}$ fixed point theory, several conditions are established to ensure the existence of $5^{n}$ equilibria, in which all the components of $4^{n}$ equilibria are located in the continuous intervals of the activation functions. and some of the components of $5^{n}-4^{n}$ equilibria are located at some discontinuous points of the activation functions. The introduction of discontinuous activation functions makes the neural networks have more equilibria than those with continuous activation functions. Furthermore, some criteria are proposed to ensure local Mittag-Leffler stability of $3^{n}$ equilibria. The introduction of nonmonotonic activation functions makes the neural networks have more stable equilibria than those with monotonic activation functions. Two examples are given to illustrate the effectiveness of the results.Keywords
Funding Information
- Natural Science Foundation of Zhejiang Province (LQ18F030015, LY18F030023)
- National Natural Science Foundation of China (61503338, 61873238, 61603121)
This publication has 33 references indexed in Scilit:
- Superconvergence analysis for the explicit polynomial recovery methodJournal of Computational and Applied Mathematics, 2014
- Global stability of complex-valued neural networks with both leakage time delay and discrete time delay on time scalesNeurocomputing, 2013
- Dynamical stability analysis of multiple equilibrium points in time-varying delayed recurrent neural networks with discontinuous activation functionsNeurocomputing, 2012
- Mittag–Leffler stability of fractional order nonlinear dynamic systemsAutomatica, 2009
- Multistability and convergence in delayed neural networksPhysica D: Nonlinear Phenomena, 2007
- Dynamics Analysis and Analog Associative Memory of Networks With LT NeuronsIEEE Transactions on Neural Networks, 2006
- Global Exponential Stability and Global Convergence in Finite Time of Delayed Neural Networks With Infinite GainIEEE Transactions on Neural Networks, 2005
- On discrete-time cellular neural networks for associative memoriesIEEE Transactions on Circuits and Systems I: Regular Papers, 2001
- Sparsely interconnected neural networks for associative memories with applications to cellular neural networksIEEE Transactions on Circuits and Systems II: Analog and Digital Signal Processing, 1994
- Neurons with graded response have collective computational properties like those of two-state neurons.Proceedings of the National Academy of Sciences, 1984