Neural Network Model Based On Multi-Temporal BAM

Authors

  • Wei Liu Department of Electronic Engineering, Tongling Polytechnic, Tongling 244061, China

Abstract

By optimizing the control design of a Bi-Directional Associative Memory (BAM) neural network with multiple delays, the stability and reliability of the BAM neural network with multiple delays is improved. A multi-delay BAM neural network control algorithm based on a variable structure proportional, integral and derivative (PID) fuzzy neural network is proposed. By ignoring the feedforward and coupling terms of the controller, and by concentrating on the fault tolerance of the inner loop control, the differential equation of the state delay of the BAM neural network control with multiple delays is obtained. The fault-tolerant control law is selected to control the BAM neural network with multiple delays in a steady state. Combined with the
Lyapunov stability principle, the total input values of hidden layer neurons in the PID fuzzy neural network are obtained. The variable structure feedforward three-layer adaptive PID neural network model is used as a learning device to realize the optimal design of the BAM neural network control algorithm with multiple delays. The simulation results show that the state response robustness and adaptability of the output of the multi-delay BAM neural network are better when. Multi-delay BAM neural network is controlled by system - delay coupling system which shows the better control quality of the multi-delay BAM neural network.

Keywords: Neurons; BAM Neural Network with Multiple Delays; Control; PID; Learning Device.

Cite As

Y. Liu, "Neural Network Model Based On Multi-Temporal BAM", Engineering Intelligent Systems,
vol. 30 no. 1, pp. 39-48, 2022.



 

Downloads

Published

2022-01-01

Issue

Vol. 30 No. 1 (2022): International Journal of Engineering Intelligent Systems

Section

General Submission

License

The submission of a paper implies that, if accepted for publication, it will not be published elsewhere in the same form, in any language, without the prior consent of the publisher. Before publication, authors are requested to assign copyright to CRL Publishing Ltd. This allows CRL to sanction photocopying, and to authorize the reprinting of issues or volumes according to demand. Authors' traditional rights will not be jeopardized by assigning Copyright in this way, as they retain the right to reuse the material following publication, and to veto third-party publication.