注射模精密成型过程一致性在线监控方法研究  

Research on on-line monitoring method for precision forming process of injection mould

作　　者：李华伟 李阳[1] 郭飞 张云[1] 周华民[1] 李德群[1] LI Hua-wei;LI Yang;GUO Fei;ZHANG Yun;ZHOU Hua-min;LI De-qun(State Key Laboratory of Material Processing and Die&Mould Technology,Huazhong University of Science and Technology,Wuhan,Hubei 430074,China)

机构地区：[1]华中科技大学材料成形与模具技术国家重点实验室,湖北武汉430074

出　　处：《模具工业》2020年第5期8-13,19,共7页Die & Mould Industry

基　　金：国家重点研发计划课题项目(2019YFB1704903)。

摘　　要：精密制品的注射成型易受生产环境波动的影响,实际生产过程中,难以及时监控异常的发生,导致废品率高,原材料浪费大,通过对注射过程的曲线反映制品质量的机理进行研究,提出了基于统计量的特征提取模型,并通过特征距离衡量批次差异性,建立与实际制品质量的关联。该研究克服了现有监控方法实时性差、模型迁移受限等问题,并通过试验验证特征距离与质量检测标准之间的相关性高于现有已知的经验特征,且对不同类型的波动具有较高的灵敏度,满足实际生产需求。In the actual production process,it is difficult to monitor the occurrence of abnormalities in time,resulting in a high rate of rejects and a huge waste of raw materials.Through the further research on the mechanism of the product quality reflected by the process curve,it proposed a feature extraction model based on statistics,and measures the batch difference by feature distance to establish the correlation with the actual product quality.This study overcame the problems of poor real-time performance and limited model migration of existing methods,and experimentally verified that the correlation between feature distance and quality inspection standards was higher than the existing known empirical features,and has different effects on different types of external fluctuations.Higher sensitivity to meet actual production needs.

关 键 词：精密注射 参数波动 故障诊断 在线监控 型腔压力 

分 类 号：TG76[金属学及工艺—刀具与模具] TP391.77[自动化与计算机技术—计算机应用技术]