摘要
针对目前智能移动终端制造过程存在产品质量一致性不高、影响产品质量指标因素不明确、质量分析智能化程度不高等问题,在智能移动终端SMT产线关键工艺机理研究基础上,关联产品质量与关键工艺参数,分析产品质量影响因素。以首道印锡工序为切入点,基于改进的k-means算法对SPI检测数据记录进行聚类分析,识别样本数据异常点。针对SPI检测异常数据基于Apriori算法构建产品质量与印锡工艺参数关联模型,以最小支持度和最小置信度为判断依据,分析产生异常质量指标的重要工艺因素。在此基础上,构建质量-工艺模型反向优化刮刀速度和压力、脱模速度和距离等印锡关键工艺参数,基于粒子群算法优化求解模型,在最优质量指标输出下最佳工艺参数动态组合,结果表明,当权重系数λ=0.1时,获得最佳工艺参数组合。
In view of the current intelligent mobile terminal manufacturing process of product quality consistency is not high,the factors affecting product quality index is not clear,quality analysis intelligent degree is not high,based on the research on the key process mechanism of intelligent mobile terminal SMT production line,the factors affecting product quality are analyzed by linking product quality with key process parameters.With the first tinning process as the starting point,the SPI detection data records are clustered based on the improved k-means algorithm to identify the sample data anomalies.For SPI detection abnormal data,the correlation model of product quality and tinning process parameters is constructed based on Apriori algorithm to analyze the important process factors generating the abnormal quality index.On this basis,a quality process model is constructed to reverse optimize key tinning process parameters such as scraper speed and pressure,demolding speed and distance,based on the particle group algorithm optimization model,under the optimal quality index output optimal process parameters dynamic combination,the results show that when the weight coefficientλis 0.1,the best process parameters combination can be obtained.
作者
任炎芳
刘志培
靳晓洋
张杨志
Ren Yanfang;Liu Zhipei;Jin Xiaoyang;Zhang Yangzhi(Guangdong Hongqin Communication Technology Co.,Ltd.,Dongguan,Guangdong 523808,China;Guangdong Intelligent Robotics Institute,Dongguan,Guangdong 523808,China)
出处
《机电工程技术》
2023年第12期186-190,共5页
Mechanical & Electrical Engineering Technology
基金
广东省基础与应用基础研究基金区域联合基金项目(粤港澳研究团队项目)(2022B1515130011)。
关键词
智能移动终端
聚类分析
质量关联模型
工艺优化建模
intelligent mobile terminal
cluster analysis
quality correlation model
process optimization modeling
