机构地区: 北京科技大学
出 处: 《钢铁》 1998年第4期27-30,共4页
摘 要: 以Gleeble_1500热模拟机得到的实验数据为基础,采用人工神经网络方法建立了50CrV4钢变形抗力与应变、应变速率和温度对应关系的预测模型,并与多元非线性回归模型比较,具有较高的精度。 On the basis of the data obtained on Gleeble1500 Thermal Simulator,the predicting models for the relation between flow stress and deformation strain,strain rate and temperature for 50CrV4 have been developed with Artificial Neural Network method.Comparison with nonlinear regression method,the neural network gives better results.