化工学报
化工學報
화공학보
JOURNAL OF CHEMICAL INDUSY AND ENGINEERING (CHINA)
2015年
7期
2547-2554
,共8页
徐艳%张春霞%王汝贤%吕静%赵玉军%马新宾
徐豔%張春霞%王汝賢%呂靜%趙玉軍%馬新賓
서염%장춘하%왕여현%려정%조옥군%마신빈
精馏序列%代码矩阵%算法%模拟%优化%乙二醇
精餾序列%代碼矩陣%算法%模擬%優化%乙二醇
정류서렬%대마구진%산법%모의%우화%을이순
distillation sequences%code matrices%algorism%simulation%optimization%ethylene glycol
提出了一种基于代码矩阵的精馏序列合成方法及其算法实现。该方法包括混合物组群划分、减塔配置基本搜索空间的创建、代码矩阵的生成和使用分离限制条件筛选4个步骤。由于代码矩阵衍生自矩阵法的0-1矩阵,且其元素代码包含了精馏塔分离编号和产品流股的采出位置等信息,因此,该合成法不仅能提供完整的精馏配置搜索空间,而且便于根据分离限制要求进行分离序列的筛选。文中给出该方法在乙二醇粗产品体系的应用,由VB计算共获得了8个三塔分离方案。结合Aspen模拟,得到总再沸器热负荷最小的三塔分离方案。
提齣瞭一種基于代碼矩陣的精餾序列閤成方法及其算法實現。該方法包括混閤物組群劃分、減塔配置基本搜索空間的創建、代碼矩陣的生成和使用分離限製條件篩選4箇步驟。由于代碼矩陣衍生自矩陣法的0-1矩陣,且其元素代碼包含瞭精餾塔分離編號和產品流股的採齣位置等信息,因此,該閤成法不僅能提供完整的精餾配置搜索空間,而且便于根據分離限製要求進行分離序列的篩選。文中給齣該方法在乙二醇粗產品體繫的應用,由VB計算共穫得瞭8箇三塔分離方案。結閤Aspen模擬,得到總再沸器熱負荷最小的三塔分離方案。
제출료일충기우대마구진적정류서렬합성방법급기산법실현。해방법포괄혼합물조군화분、감탑배치기본수색공간적창건、대마구진적생성화사용분리한제조건사선4개보취。유우대마구진연생자구진법적0-1구진,차기원소대마포함료정류탑분리편호화산품류고적채출위치등신식,인차,해합성법불부능제공완정적정류배치수색공간,이차편우근거분리한제요구진행분리서렬적사선。문중급출해방법재을이순조산품체계적응용,유VB계산공획득료8개삼탑분리방안。결합Aspen모의,득도총재비기열부하최소적삼탑분리방안。
A code matrix-based synthesis method for multicomponent distillation sequences is described as well as its algorism. Four steps, such as grouping the components of feed, generating the basic search space of sub-column configurations, transforming 0-1 matrices into code matrices and eliminating configurations with separation constraints, are included in this method. The code matrices, whose element code includes the information about the column sequence number and the location of product withdrawing, are transformed from 0-1 matrices determined by matrix method. Thus, this method not only provides a complete set of separation configurations with sharp or non-sharp split but also can pick out all the desirable configurations with the constraints by logical programming. Additionally, the use of this method is illustrated by applying it to the ethylene glycol purification from dimethyl oxalate hydrogenation. Eight configurations are selected out and one configuration with the lowest total reboiler heat duty is identified by Aspen Plus optimization.