Dynamic properties of Elman and modified Elman neural network

Abstract

The Elman network is a well-known recurrent neural network. In this paper, an overview of the structure of the Elman neural network and its modified form are presented. Both have trainable feed-forward connections. Then, by comparing with the discrete PID control algorithm, their dynamic characteristics are fully discussed using the z transform function. Finally, we conclude that the modified Elman network has better dynamic properties.