智能可视化优化技术用于常压塔多目标操作优化  

Multi-objective optimization for atmospheric tower operation by using intelligent visual optimization

作　　者：肖翔[1] 熊佐松 刘海燕[2] 廖安[2] 史彬[2] 鄢烈祥[2] 

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

出　　处：《炼油技术与工程》2015年第4期8-10,共3页Petroleum Refinery Engineering

摘　　要：应用Aspen Plus流程模拟软件,对某炼油厂的原油常减压蒸馏装置中常压塔进行准确模拟。在此基础上,选取常底蒸汽量、侧线采出量、中段回流量等8个操作变量,在其值域范围内,通过计算机模拟实验,获得了满足常压塔约束条件的样本数据。以产值最大化和能耗最小化为优化目标,应用智能可视化优化技术将这些样本数据降维映射到二维平面,生成产值与能耗的等值线。根据数据点和等值线的降维映射图,确定出能耗小、产值高的优化区域和优化操作点。对比原工况点,优化操作点的常压塔产值提高了1.42%、能耗降低了4.7%,智能可视化优化技术为常压塔的多目标操作优化提供了一种有效的手段。The atmospheric and vacuum distillation unit in a refinery was simulated by Aspen Plus soft- ware, and the simulation results agreed well with the working conditions. The eight operating parameters, such as the amount of steam at the bottom of atmospheric tower, the side cutting capacity, the pump-around reflux and the atmospheric furnace temperature, etc, were selected as operating variables. The computing experi- ments were used to obtain the atmospheric tower operating parameters and its sample data. By using the intelli- gent visual optimization, these sample data are mapped onto a two dimensional plane, and the contours of out- put and energy consumption are generated. The optimized point and region which have lower cost and higher output can be determined intuitively based on these contours. As compared with the original operating point, the optimized point can increase the 1.42% output and reduce 4.7% energy requirement of the atmospheric tower. The study shows that the intelligent visual optimization technology provides an effective means to solve the multi-objective optimization issue of atmospheric tower.

关 键 词：智能可视化优化技术 常压塔 操作优化 

分 类 号：TE968[石油与天然气工程—石油机械设备]