摘要
介绍了数据挖掘技术(DMOS)优化软件在N-甲酰吗啉芳烃抽提装置上的具体运用过程及效果。通过收集工艺参数的历史数据,将所有相关变量构建一个高维空间,优类样本点形成优化区,将高维数据空间投影到两维特征平面,直观地在平面图形上看到样本点的分布和优化区形状,进而研究参数与抽余油中苯质量分数之间变化规律,提出操作调整建议,实施后取得较好效果。
The application process and effect of data mining technology optimization software(DMOS) on N-formylmorpholine aromatics extraction device are introduced.Through collecting historical data of process parameters,all relevant variables are constructed into a high-dimensional space,the optimal sample points form an optimal zone,the high-dimensional data space is projected onto the two-dimensional feature plane,and the distribution of sample points and shape of the optimal zone are visually seen on the plane graph.And the variation between the parameters and the mass fraction of benzene in the raffinate oil was studied.Suggestions on adjustment of operation were proposed and good results were obtained after implementation.
作者
朱以斌
Zhu Yibin(Aromatics Division,SINOPEC Shanghai Petrochemical Co.,Ltd.,Shanghai 20054)
出处
《石油化工技术与经济》
2018年第3期10-13,共4页
Technology & Economics in Petrochemicals