SVR-Based Predictive Model for 2519 Aluminum Alloy

Abstract

For 2519 aluminum alloy, there are very complex nonlinear relations among the thermal dynamical parameters in the process of deforming. In this paper, the support vector regression (SVR) approach is proposed to establish a model for predicting flow stress of 2519 alloy base on the flows tress experimental data of 2519 aluminum alloy under two influential factors, including strain and strain rate. Research showed that the prediction precision of SVR model is high enough: the mean absolute error (MAE) is 0.181, mean absolute percentage error (MAPE) is 0.434%, root mean square error (RMSE) is 0.22, multiple correlation coefficient (R2) is 0.998. This research suggests that SVR is an effective and powerful tool for predicting the flow stress of 2519 aluminum alloy.