Original scientific paper

Prediction of mechanical properties of composite materials based on convolutional neural network-long and short-term memory neural network

P. Huang ; Xinjiang University, Urumqi, Xinjiang, 830000, China
J. H. Dong ; Xinjiang University, Urumqi, Xinjiang, 830000, China
X. C. Han ; Xinjiang University, Urumqi, Xinjiang, 830000, China
Y. P. Qi ; Xinjiang University, Urumqi, Xinjiang, 830000, China
Y. M. Xiao ; Xinjiang University, Urumqi, Xinjiang, 830000, China
H. Y. Leng ; Xinjiang University, Urumqi, Xinjiang, 830000, China

Abstract

Convolutional neural networks (CNNs) have the advantage of processing complex images and extracting feature information from the images, while long and short term memory networks (LSTMs) are good at processing data with sequential features. In this paper, based on the deep material network, we propose to apply the CNN-LSTM neural network model to the prediction of mechanical properties of carbon fibre composites. Then the experimental results are compared with the model prediction results, and the results show that the CNN-LSTM prediction of the mechanical properties of carbon fibre composites is within 5% of the corresponding tensile mechanical experimental results, which proves the accuracy of the CNN-LSTM neural network model in the prediction of the mechanical properties of carbon fibre composites.

Keywords

artificial neural networks; deep learning; performance prediction

Hrčak ID:

315677

URI

https://hrcak.srce.hr/315677

Publication date:

1.7.2024.

Visits: 868 *