摘要
近些年,表面贴装技术(SMT)的自动化、信息化水平有很大提升,生产出的产品性能更优异。SMT生产工艺复杂,各道工序实施阶段均可能出现质量缺陷,若在生产线上不能尽早地洞察、处理这些缺陷,则很可能降低产品质量,严重时会造成整块印制板报废,增加产品的生产费用。文章从大数据技术角度预测基于SMT下锡膏印刷过程的产品质量,预测阶段设定了时间窗口,动态更新数据,最后创建了时间序列数据包,较明显地提升了模型预测的正确率。
In recent years,the automation and information level of surface mount technology(SMT)has been greatly improved,and products with better performance can be produced.The SMT production process is complex,and quality defects may appear in the implementation stages of each process.If these defects cannot be detected and dealt with as soon as possible on the production line,it is likely to reduce product quality,and in severe cases,the entire printed circuit board will be scrapped and the cost of products will be increased.Based on the perspective of big data technology,this article predicts the product quality of the solder paste printing process based on SMT.The prediction stage sets a time window T,dynamically updates the data,and finally builds a time series data package,which significantly improves the accuracy of the model prediction rate.
作者
张雪
王晓燕
Zhang Xue;Wang Xiaoyan(Zhengzhou Electric Power College,Zhengzhou 450000)
出处
《中阿科技论坛（中英文）》
2020年第10期50-53,共4页
China-Arab States Science and Technology Forum
基金
2018年河南省高等学校重点科研项目(19B520030)
2020年郑州电力高等专科学校校级科研项目(ZEPCKY2020-27)。
关键词
表面贴装技术
大数据
产品质量
质量预测
时间窗口
预测方法
Surface mount technology
Big data
product quality
Quality prediction
Time window
Prediction method
作者简介
王晓燕(1984-),女,讲师,硕士,研究方向:计算机应用技术、网络技术。
