基于BP神经网络的覆盖件模具机加工成本估计  

Cost Prediction for the Machining of Automobile Panel Die Based on BP Neural Networks

作　　者：邢宪才[1] 刘渝[1] 王安兵 

机构地区：[1]西华大学机械工程与自动化学院,四川成都610039 [2]四川天元模具制造有限公司,四川成都610019

出　　处：《机械与电子》2008年第5期61-64,共4页Machinery & Electronics

摘　　要：在分析常用的模具成本估计方法的基础上,提出了构造模具机械加工成本神经网络计算方法;同时,引入了测差样本方法,用区间估计来为模具机加工成本报价;以铸造整体结构拉延模具为例,利用企业的历史数据组成训练样本集,训练所形成的神经网络;并对前壁板上梁铸造整体结构拉延模具的机加工成本进行了估计。计算结果显示了该方法的可行性。After analyzi cost forecasting of stamp method based on BP neural ward in this paper. At the ng the methods of the ing die, an evaluation networks was put forsame time, the sample for error measuring was introduced to make a range estimating of the machining cost of the stamping die. As an example, the historical machining cost data was used to train the neural net formed, and the trained neural net was used to predict the machining cost of the forming die of one automobile panel. The results showed the proposed method could well predict the value of machining cost.

关 键 词：BP神经网络 成本估计 区间估计 

分 类 号：TB115[理学—数学]