基于BP神经网络的表面硬度预测模型  

Forecast Model of Hardness Based on the BP Neural Network

作　　者：刘宝[1] 吴钢[1] 宋光明[1] 

机构地区：[1]上海海事大学激光应用研究室,上海200135

出　　处：《热处理技术与装备》2006年第3期37-39,49,共4页Heat Treatment Technology and Equipment

摘　　要：运用MATLAB中BP神经网络强大的数据预测功能,并结合VB编写的用户操作界面,建立了基于BP神经网络的表面硬度预测模型。对该模型进行训练和仿真后,其所得到的结果与实际测量值相差较小,能够达到模型预测的要求,有较好的工程实际意义。With BP neural network＇ s powerful function in numeric forecast and with user＇s interface designed by VB, we estabilished a forecast model. After training and simulation we can see there is a little difference between the predicted values and the tested data. It means the model can be used in experimentation and can meet the goal.

关 键 词：BP神经网络 激光相变硬化 扫描参数 预测 

分 类 号：TG179[金属学及工艺—金属表面处理]