文献类型：期刊文献

中文题名：基于模式识别的加氢裂化装置可视化模型

英文题名：Visualization Model of Hydrocracking Unit Based on Pattern Recognition

作者：王振宇[1];叶国华[2];吴云潇[1];朱俊涛[1];杜晓航[1];曾广威[1];李智[1]

机构：[1]华东理工大学能源化工过程智能制造教育部重点实验室,上海200237;[2]中国石化海南炼油化工有限公司,海南洋浦578101

年份：2024

卷号：31

期号：7

起止页码：1335

中文期刊名：控制工程

外文期刊名：Control Engineering of China

收录：CSTPCD;;北大核心:【北大核心2023】；CSCD:【CSCD_E2023_2024】；

基金：政府间国际科技创新合作项目(2021YFE0112800);国际(地区)合作与交流项目(61720106008);国家自然科学基金面上项目(62073142,62173144)。

语种：中文

中文关键词：加氢裂化;特征提取;路径规划;过程优化;可视化

外文关键词：Hydrocracking;feature extraction;path planning;process optimization;visualization

摘要：为了实时监控加氢裂化过程装置的运行状态,实现生产过程的优化,提出了一种基于模式识别的加氢裂化装置可视化模型。首先,使用聚类方法对加氢裂化装置的历史数据进行生产过程工况划分,并针对不同工况,基于主成分分析算法分别构建可视化模型,获得评估生产过程的有效工具;然后,根据2个重要指标划分模型区域,便于对新样本进行指标评估;最后,使用路径规划算法寻找到达最优生产模式的迁移路径。仿真结果表明,所提模型能有效根据工况与性能指标划分区域,并针对低效的生产状态得到生产过程变量的调节方法,保持生产过程产品的高收率。
In order to monitor the operating state of hydrocracking unit in real time and optimize the production process,a visualization model of hydrocracking unit based on pattern recognition is proposed.Firstly,a cluster method is used to divide the historical data of hydrocracking unit into different working conditions of the production process,and visualization models are constructed based on principal component analysis algorithm to obtain effective tools for evaluating the production process.Then,according to two important indicators,the model region is divided to facilitate the evaluation of new samples.Finally,a path planning algorithm is used to find the optimal migration path to the optimal production mode.The simulation results show that the proposed model can effectively divide the regions according to the working conditions and performance indicators,and obtain the adjustment method of the production process variables according to the inefficient production state,so as to maintain the high product yield of the production process.