摘要
采用生产流程分析法对零件分组时,通常难以为新零件确定所属零件族,故提出了基于神经网络的零件族智能搜索算法。该方法将工艺与零件族的关系转化为零件编码与零件族之间的关系,基于现有零件的编码、工艺过程和零件族划分情况,用神经网络构建零件编码与零件族之间的智能非线性映射模型,从而实现新零件的零件族智能搜索。给出了算法的基本原理、神经网络设计和训练过程,并结合实例验证了算法的可行性和有效性。
It is usually difficult to search for a part family belonged to a new part when production flow analysis is used for part family classification.This paper presented a new approach for part family search based on artificial neural network.In this approach,the relationship between manufacturing process and part family was converted into the relationship between part codes and part family,an artificial neural network was employed to realize the intelligent and nonlinear mapping from part code to part family based on the existing part code,manufacturing process and parts family and then intelligent part family search can be realized.The principle of the approach,artificial neural network samples,and training process were discussed.The feasibility and effectiveness of this approach was tested with real cases.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2012年第21期2597-2600,共4页
China Mechanical Engineering
关键词
成组技术
神经网络
生产流程分析
零件编码
零件族搜索
group technology
neural network
production process analysis
part coding
part family search
