基于数据驱动的B公司保险杠质量控制  

Bumper Quality Control of Company B Based on Data Driving

作　　者：吕悦 岳立柱 崔亚华 LV Yue;YUE Li-zhu;CUI Ya-hua(School of Business Administration,Liaoning Technical University,Huludao 125105;School of Economics and Management,Huangshan University,Huangshan 245021,China)

机构地区：[1]辽宁工程技术大学工商管理学院,辽宁葫芦岛125105 [2]黄山学院经济管理学院,安徽黄山245021

出　　处：《物流工程与管理》2023年第2期100-105,共6页Logistics Engineering and Management

基　　金：国家社会科学基金资助项目(17BSH076);辽宁省教育厅基金资助项目(LJ2020JCL028)。

摘　　要：保险杠是汽车不可缺少的一个重要零部件,由于生产环节较多,不可避免地会出现质量问题,因此降低次品率是提升企业竞争力的一个有力手段。利用保险杠在其生产过程中积累的大量生产数据和各生产环节之间的逻辑关系,运用统计分析方法和CART决策树提前预判故障点,对故障率较高的注塑环节进行数据挖掘。结果表明,班次、注塑机台、物料、机台直接人工人数均会影响保险杠的质量,其中物料和注塑机台是影响保险杠质量最关键的两个因素。Bumper is an essential part of automobile, of which quality problems will inevitably occur due to many production links.How to reduce the defective rate is a powerful means to improve enterprises’ competitiveness.By using a large amount of production data generated during the production process of bumpers and the logical relationship between various production links, and applying statistical analysis method and CART decision tree to predict the fault point in advance, data mining is carried out for the injection molding link with high failure rate.The results show that the shift, injection molding machine, materials, number of direct workers will affect the quality of the bumper, among which materials and injection molding machine are the two key factors affecting the quality of the bumper.

关 键 词：保险杠 质量控制 数据驱动 CART算法 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]