基于人工神经网络的热轧碳钢变形抗力预报

Prediction of Flow Stress of Carbon Steel in Hot Strip Rolling by Artifcial Neural Network

摘要: 以恒应变速率凸轮压缩试验机得到的实验数据为基础,采用人工神经网络的方法建立了碳钢变形抗力与应变、应变速率及温度对应关系的预测模型,与多元非线性回归模型比较,神经网络模型具有较高的预测精度.

Abstract: On the basis of the data obtained on Cam Plastometer with constant strain rate, the predicting model for the relation between flow stress and deformation strain, strain rate and temperature of carbon steel has been developed by Artificial Neural Network method. Compared with nonlinear regression method, the neural network model gives better results.