BP神经网络预测脉冲电沉积Cu-SiC镀层镀速的研究  

Forecast of plating rate of pulse electrodeposited Cu SiC coatings using BP model

作　　者：李兴远[1] 夏法锋[2] 陈达[2] 

机构地区：[1]宁波大红鹰学院,浙江宁波315175 [2]东北石油大学机械科学与工程学院,黑龙江大庆163318

出　　处：《兵器材料科学与工程》2013年第3期99-101,共3页Ordnance Material Science and Engineering

摘　　要：采用脉冲电沉积在A3钢表面制备Cu-SiC镀层,采用BP网络模型对脉冲电沉积Cu-SiC镀层的镀速进行预测研究。结果表明:本BP模型为3×9×1型神经网络模型,预测值与试验值曲线吻合较好;其相对误差较小,最大误差为2.7%,其相关系数为0.999,能较好地预测脉冲电沉积Cu-SiC镀层的镀速。The Cu SiC coatings were deposited on the surface of A3 steel by pulse electrodeposition method. And a BP model was used for forecasting the plating rate of Cu SiC coating. The results indicate that the BP model is 3×9×1 neural networkmodel.The predicted values agree well with experimental values,the relative error is small,and the max error is 2.7%,the coefficient of correlation is 0.999. This BP model can be used to predict the plating rate of pulse electrodeposited Cu SiC coatings.

关 键 词：BP网络模型 脉冲电沉积 Cu-SiC镀层 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论] TG174.4[自动化与计算机技术—计算机科学与技术]