摘要
利用人工神经网络对铅黄铜超塑性能进行了预测研究,通过对试样在不同超塑性拉伸条件下的性能进行学习,建立了拉伸条件与性能的BP网络预测模型。结果表明,所建模型可以较好地反映超塑性拉伸条件与性能间的内在规律,预测值和试验结果吻合良好,其最大误差不超过10%,人工神经网络用于铅黄铜超塑性能的预测具有可行性和有效性。
The superplastic performance prediction of lead brass was studied based on artificial neural network. Through studying the performance of lead brass samples under superplastic tension conditions, the prediction model of BP neural network was founded. The results show that the founded model can reflect the relationship between superplastie tension conditions and performance, and predicting values agree well with tests in accordance less than 10%. It indicates that the prediction of lead brass superplasticity using artificial neural network is effective and feasible.
出处
《热加工工艺》
CSCD
北大核心
2008年第14期92-95,共4页
Hot Working Technology
关键词
铅黄铜
超塑性
人工神经网络
预测模型
lead brass
superplasticity
artificial neural network
prediction model
