摘要
企业产品质量和工艺的波动数据中蕴含了生产规律性,利用数据挖掘方法总结波动数据中的统计规律,优化控制产品质量,是企业挖潜增效的有效技术方法;同时,企业产品的智能优化需要自动化建模技术.以聚全氟乙丙烯(fluorinated ethylene propylane, FEP)产品合格率(产品熔融指数分布满足用户指定范围的达标率)为生产优化目标,在收集生产装置有关工艺数据的基础上,利用最佳投影识别法自动建立了FEP产品熔融指数与其生产工艺参数间的关系,得到了FEP产品的生产工艺优化方案,并在实际生产中进行实施.结果显示,该方案显著提高了FEP产品的熔融指数合格率,以及企业的经济效益.
The process data of production can be utilized to investigate the regularity of production quality by using data mining, which is an effective method in industrial optimization. Furthermore, it is expected that the optimal model can be constructed in a fast and automatic way. In this project, the optimal projection recognition technique was proposed to construct the pattern recognition model for classifying different qualities of fluorinated ethylene propylene(FEP) concerning melt index. It was found that the qualified accuracy of production increased remarkably after the application of the optimal model available.
引文
[1] Balz G, Schiemann G. Aromatic fluorine compounds. I. a new method for their preparation[J]. Chem Ber, 1927(60):1186-1190.
[2] Midgley T, Henne A L. Organic fluorides as refrigerants[J]. Ind Eng Chem, 1930(22):542-545.
[3] Hamilton J M. The organic fluorochemicals industry[J]. Adv Fluorine Chem, 1963(3):117-180.
[4] Hiyama T. Organo fluorine compounds, chemistry and applications[M]. Berlin:Springer, 2000.
[5] Lagow R. Fluorine chemistry at the millennium:fascinated by fluorine[M]. Amsterdam:Elsevier, 2000.
[6] McKeen L W. Fluorinated coatings and finishes handbook[M]. Boston:Elsevier, 2016.
[7]倪震宇.聚全氟乙丙烯的市场及技术进展[J].有机氟工业, 2006(2):45-46.
[8]纪晓波,刘亮,赵慧,等.最佳投影识别法用于三唑类化合物的抗真菌活性的分子筛选[J].上海大学学报(自然科学版), 2004, 10(2):191-194.
[9]李远明.聚全氟乙丙烯的制备工艺[J].有机氟工业, 2005(3):9-12.
[10]糜万俊,阚大清.聚全氟乙丙烯聚合工艺研究[J].有机氟工业, 2017(3):21-22.
[11]陆文聪,李国正,刘亮,等.化学数据挖掘方法与应用[M].第一版.北京:化学工出版社业, 2011:1-15.