摘要
基于人工神经网络理论 ,建立了各种硬度值之间相互转换的完整统一的神经网络模型 .该模型可实现各种硬度之间的线性与非线性映射 ,并对硬度表中没有的数据给出准确的预测 ,克服了查表法的不足 ,较线性回归具有更高的精度 .网络模型可嵌入机械 CAD系统中 ,方便地实现各种硬度之间的转换 .
An ANN( artificial neural network) model fortransformation of all kinds of hardnesswas estab- lished based on ANN theory. Itcan realize the linearand nonlinearmapping between hardness values,and predict precisely the hardness value that is not present in the table.So it overcomes the shortcomings of hardness transformation table. Compared with the linear regression method,this model can not only deal with the nonlinear mapping between hardness values,butalso reach much higher accuracy.In addition,it can be embedded into a mechanical CAD system to realize hardness transformation conveniently.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2000年第10期1376-1377,1381,共3页
Journal of Shanghai Jiaotong University
关键词
硬度
人工神经网络
转换模型
CAD
hardness
artificial neural network( ANN)
transformation
