基于人工神经网络的含Nb微合金钢薄板坯CSP连轧过程流变应力预报

Prediction of flow stress of Nb micro-alloyed steel sheet billets during CSP continuous rolling based on artificial neural networks

在Gleeble-3500热模拟实验机上对含Nb微合金钢进行了一系列热压缩实验,用以模拟含Nb微合金钢的薄板坯紧凑式热带生产(CSP)连轧过程。变形温度分别为950、1 000、1 050、1 100、1 150℃,变形速率分别为1、5、10s-1,应变量为70%。基于实验得到的流变应力数据,利用Matlab软件建立了描述流变应力和变形温度、变形速率和应变量之间关系的人工神经网络模型。实验结果表明:预测值和实际值的相关系数和平均相对误差分别为0.989和2.41%,证明所建立的人工神经网络模型可以准确地预测含Nb微合金钢热变形过程中的流变应力值。该研究结果对提高含Nb微合金钢的热塑性变形能力及优化CSP连轧工艺具有指导意义。

Hot compression experiments were conducted on a Gleeble-3500thermal-mechanical simulator to simulate CSP continuous rolling process of Nb micro-alloyed steel sheet billets at 950-1 150℃and under the strain rate of 1-10s-1.Based on the flow stress data obtained from the experiments,an artificial neural network-based model was developed using Matlab software to describe the relationship between flow stress and deformation parameters,i.e.temperature,strain and strain rate.Experimental results show that the correlation coefficient and averaged relative error between the prediction and measurement data are respectively 0.989 and 2.41%,which indicates that the model has very high accuracy.Thus,the results of this study are of great significance to the improvement of hot plastic deformation performance and the optimism of CSP continuous rolling process of the studied steel.