Title: A summary of the stability of several types of neural networks

Authors: Wenbo Fei; Jianhua Zhang; Yang Li

Addresses: Hebei University of Science and Technology, Shijiazhuang, Hebei, China ' School of Information and Control Engineering, Qingdao University of Technology, Qingdao, Shandong, China; Hebei Provincial Research Centre for Technologies in Process Engineering Automation, Shijiazhuang, Hebei, China ' School of Information and Control Engineering, Qingdao University of Technology, Qingdao, Shandong, China

Abstract: Owing to the influence of factors such as signal transmission delay, external interference and parameter deviation, the stability of neural networks has always been the focus of scholars, and many related literatures have been published. This article mainly summarises and analyses the stability research of several types of neural networks (Hopfield neural network, BAM neural network, cellular neural network, Cohen-Grossberg neural network). In the study of neural network stability, in addition to the common methods such as Lyapunov-Krasovskii method and LMI technology, other more advantageous solutions are also analysed. Finally, the conclusion and prospect of neural network stability analysis are given.

Keywords: neural networks; stability.

DOI: 10.1504/IJCCPS.2022.124881

International Journal of Cybernetics and Cyber-Physical Systems, 2022 Vol.1 No.2, pp.119 - 136

Received: 18 Sep 2020
Accepted: 15 Dec 2020
Published online: 13 Aug 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article