基于粗糙集与神经网络集成的制造过程质量诊断技术研究  

The Research on Quality Diagnosis Technology of Manufacture Process Based upon Rough Set and Neural Networks

作　　者：江大伟 周平[2] 李丽华[2] 

机构地区：[1]威高集团,山东威海264300 [2]济南铸造锻压机械研究所有限公司,山东济南250022

出　　处：《中国铸造装备与技术》2014年第2期63-65,共3页China Foundry Machinery & Technology

摘　　要：结合粗糙集理论和神经网络在信息处理方面的优势,建立了一个基于粗糙集理论和神经网络结合的机械制造过程质量诊断模型;并以42CrMo轴高硬度磨削工序为例分析各因素对其粗糙度的影响程度,表明可以简化网络训练样本,优化神经网络结构,提高质量诊断效率,验证了模型的可行性与有效性。Combined superiority of rough set theory with neural networks in information processing, a model of manufacturing process quality diagnosis system has been built based upon rough set theory and artificial neural networks. The influencing level of each factor on the roughness of the workpiece has been analysed upon high hardness grinding procedure of an example of a 42CrMo axis. It has demonstrated that the network training sample could be simplified to optimize the neural network structure and improve the quality of diagnostic efficiency hence the feasibility and effectiveness of the model have been test and verified.

关 键 词：质量诊断 粗糙集 可辨识矩阵 神经网络 

分 类 号：TH165.4[机械工程—机械制造及自动化]