基于粗糙集的智能可视化优化方法用于加氢裂化反应操作优化  

Operation optimization of hydrocracking reaction process based on rough set and intelligent visualization optimization method

作　　者：杨旭[1] 汪坤[2] 肖翔[2] 周靖[1] 史彬[1] 鄢烈祥[1] 

机构地区：[1]武汉理工大学化学化工与生命科学学院,湖北武汉430070 [2]中国石油化工股份有限公司武汉分公司,湖北武汉430082

出　　处：《化工进展》2016年第3期722-726,共5页Chemical Industry and Engineering Progress

基　　金：国家自然科学基金(21376185);国家高技术研究发展计划(2011AA02A206)项目

摘　　要：加氢裂化反应的历史生产数据包含了丰富的操作规律,而目前没有被充分挖掘和利用。针对该问题,本文提出了基于粗糙集的智能可视化优化方法,并应用该方法对加氢裂化反应的操作变量进行了优化。采用自适应离散化的方法对加氢裂化反应的生产操作数据进行离散化,得到离散化的数据集。针对数据集属性约简的N-P难题,提出了基于列队竞争算法的属性约简计算方法。通过对加氢裂化反应的数据集约简计算,使预设的12个操作变量约简到了8个,除去了4个冗余变量,有效降低了优化搜索空间。在此基础上,应用智能可视化优化方法将约简后的加氢裂化反应数据降维映射到平面,并生成航煤收率的等值线,据此,直观地确定出了航煤收率的优化区域和优化操作点。结果说明,选出的优化工况点与原工况航煤收率33.98%相比,能使航煤收率提高到37.58%。The historical production data in hydrocracking contains rich operational knowledge,which is not fully mined and exploited. To solve the problem,a novel method is proposed for solving the operation optimization in hydrocracking by intelligent visualization optimization method based on rough set. The continuous operation variables in the production data from practical industrial case are discretized by adaptive discretization. Because reduction of condition attributes is an N-P hard problem,thus heuristic method based on line-up competition algorithm is proposed to carry out the reduction. Results show that the proposed heuristic method can reserve the key operation variable and reduce the redundant ones efficiently,and 4 of 12 variables are reduced in hydrocracking reaction data. On the basis of the reduction,the processed data are mapped onto the plane by intelligent visualization optimization method,thus contours are created. The optimizing region and optimal operating point of the yield of aviation kerosene can be seen intuitively. Optimization result shows that the yield of aviation kerosene is promoted from previous 33.98% to 37.58% within the constraints.

关 键 词：过程系统 优化 算法 加氢裂化反应 智能可视化优化方法 

分 类 号：TQ021.8[化学工程]