Modelling of elastic modulus degradation in sheet metal forming using back propagation neural network

Abstract

The aim of this study is to develop an elastic modulus predictive model during unloading of plastically prestrained SPCC sheet steel. The model was developed using the back propagation neural networks (BPNN) based on the experimental tension unloading data. The method involves selecting the architecture, network parameters, training algorithm, and model validation. A comparison is carried out of the performance of BPNN and nonlinear regression methods. Results show the BPNN method can more accurately predict the elastic modulus at the respective prestrain levels.