作　　者： ;

机构地区： 华南理工大学工商管理学院

出　　处： 《工业工程与管理》 2021年第4期28-35,共8页

摘　　要： 物联网和5G技术的发展使得生产制造环节中产生的大量数据被收集变成可能。中国智能制造的创新和发展离不开数据驱动,基于数据驱动的离散式生产制造中特定工件的质量符合率,对制造企业的生产制造过程控制极为重要,是实现智能生产和智能决策的关键环节。通过对国内外关于质量符合率的研究进行文献综述,界定离散式制造、特定工件和质量符合率的概念,基于实际生产特定工件中收集的12 934笔数据建立质量符合率随机森林模型和Bagging模型,分别对包含20个变量的数据集进行学习。根据两个模型预测精度的对比分析,表明随机森林模型对质量符合率的预测更加有效,其预测精度达99.47%。此外,基于数据样本集,对影响特定工件质量水平的变量进行重要性分析,找出关键变量,提取不同质量水平的特定工件的重要特征,为提高特定工件质量符合率提供决策依据和指导。 The development of the Internet of Things and 5G technology has made it possible to collect a large amount of data generated in the manufacturing process.The innovation and development of China′s intelligent manufacturing are inseparable from Data-driven.The quality conformance rate of typical component in discrete manufacturing based on Data-driven was extremely important for manufacturing companies to control the manufacturing process and a key link to attain intelligent production and intelligent decision-making. A literature review of domestic and foreign research on quality conformance rate was conducted to define the concepts of discrete manufacturing,typical component,and quality conformance rate. Based on the 12 934 data collected from the actual production of a typical component,a random forest model and a bagging model of quality conformance rate were established and used to learn from a data set containing 20 variables respectively. According to the comparative analysis of the prediction accuracy of the two models,it shows that the random forest model is more effective in predicting the quality conformance rate,and its prediction accuracy is 99.47%.In addition,based on the data sample set,an importance analysis of variables that affected the quality level of a typical component was conducted and the key variables were found.The important characteristics of typical component of different quality levelsare extracted,it provides decision basis and guidance for improving the quality conformance rate of typical component.

关 键 词： 离散式制造 随机森林 质量符合率

领　　域： [经济管理—企业管理] [经济管理—国民经济]