基于BP神经网络的SCR连铸连轧法生产Cu合金电车线坯的成分与性能关系预测  

Prediction of Relationship Between Composition and Performance of Copper Alloy Wire Blanks Made by SCR Continuous Casting and Rolling Process Based on BP Neural Network Progress on BP Neural Networks

作　　者：周朝萱[1] 

机构地区：[1]攀枝花学院,四川攀枝花617000

出　　处：《热加工工艺》2011年第19期49-51,54,共4页Hot Working Technology

摘　　要：对SCR连铸连轧铜合金电车线坯的成分和性能进行测定,以其结果作为BP神经网络模拟样本。结果表明:Cu合金线坯中Cu-Ag的电学性能优于Cu-Sn,而力学性能较差;所选用的BP神经网络模型能预测Cu合金的成分和性能的关系,抗拉强度预测误差低于10%;电阻率预测误差低于5%,达到了预期目标。The composition and performance of copper alloys made by SCR continuous casting and rolling process were measured,and the the results were considered as a sample of BP neural network.The simulation results show that the electrical propertie of Cu-Ag is better than that of Cu-Sn,but Cu-Ag has poor mechanical properties.Choosing BP neural network model can predict the relationship between the composition and performance of Cu alloys,and the prediction error of the tensile strength is less than 10%;the prediction error of the resistive is less than 5%.The model can meet the expected goals.

关 键 词：SCR连铸连轧 铜合金 组成 性能 BP神经网络 

分 类 号：TG244[金属学及工艺—铸造]