论文部分内容阅读
基于BP神经网络,以挤压速度、挤压道次和挤压方式为输入层参数,以抗拉强度为输出层参数,构建了BP神经网络模型用于分析ECAP强变形对Cu-3Cr合金性能影响,并进行了试验验证以及金相组织和SEM分析。结果表明,BP神经网络输出的抗拉强度预测值与试验值之间的相对误差均小于2%,平均预测误差为1.4%,模型的预测精度高、实用性强。从改善合金抗拉强度出发,Cu-3Cr合金的最佳ECAP工艺参数:挤压速度为5mm/s、挤压道次为4次、挤压方式为每次挤压后旋转180°再进入下一道次。
Based on BP neural network, BP neural network model was constructed by taking extrusion speed, extrusion pass and extrusion as input layer parameters and tensile strength as output layer parameters to analyze the effect of strong ECAP deformation on the properties of Cu-3Cr alloy Effect, and the experimental verification and microstructure and SEM analysis. The results show that the relative errors between the predicted value and the experimental value of BP neural network output are both less than 2% and the average prediction error is 1.4%. The model has high prediction accuracy and practicality. In order to improve the tensile strength of the alloy, the optimum ECAP process parameters of Cu-3Cr alloy are as follows: the extrusion speed is 5 mm / s, the extrusion pass is 4 times, the extrusion method is rotated 180 ° after each extrusion and then entered One time.