基于改进模糊诊断的SMT产线质量追溯方法研究  

Quality Tracing Method of SMT Production Line Based on Improved Fuzzy Diagnosis

作　　者：常建涛[1] 郭嘉 林庭武 杨胜康 郭静[1] 张凯磊 CHANG Jiantao;GUO Jia;LIN Tingwu;YANG Shengkang;GUO Jing;ZHANG Kailei(Xidian University,Xi’an 710071,China;ZTE Corporation,Xi’an 710065,China)

机构地区：[1]西安电子科技大学,陕西西安710071 [2]中兴通讯股份有限公司,陕西西安710065

出　　处：《电子机械工程》2021年第2期29-36,共8页Electro-Mechanical Engineering

基　　金：陕西省重点研发计划项目(2020ZDLGY07-08);国家自然科学基金资助项目(51875432)。

摘　　要：锡膏印刷是表面贴装技术(Surface Mount Technology, SMT)产线质量管控的重要环节。目前该环节的各类参数对质量的影响程度不明确,导致质量追溯的准确度较低,难以精准确定SMT产品质量问题的根原因。针对此问题,提出了一种改进模糊诊断的SMT产线质量追溯方法。该方法基于改进模糊诊断构建动态锡膏印刷质量追溯模型,并利用sigmoid函数更新隶属度,实现诊断结果的动态更新。通过某通讯企业SMT产线多批次生产数据进行了应用验证,实现了锡膏印刷环节缺陷发生的定性定量根原因的准确分析。Solder paste printing is an important part in the quality control of surface mount technology(SMT)production lines.In this part,the impact of various parameters on quality is not clear at present,which results in the low accuracy of quality tracing,and it is difficult to determine the root cause of SMT product quality problems accurately.Aiming at this problem,a quality tracing method for SMT production line based on improved fuzzy diagnosis is proposed.The method constructs the dynamic quality tracing model of solder paste printing based on improved fuzzy diagnosis,and uses the sigmoid function to update the membership degree and realize the dynamic update of the diagnosis results.The application verification is carried out using the multi-batch production data of the SMT production line of a communication company,which realizes the accurate analysis of the qualitative and quantitative root causes of defects in solder paste printing process.

关 键 词：SMT 质量追溯 模糊诊断 

分 类 号：O159[理学—数学]