论文部分内容阅读
分馏过程能耗最低化和产量最大化是相互矛盾的目标。本文提出一种人工免疫多Agent多目标优化算法(AIMAMOA),通过重新分配分馏过程的负荷,调整分馏装置的回流比,得到一组Pareto解,从而指导分馏过程的实际生产。AIMAMOA结合了人工免疫的核心思想与多Agent技术,主要操作算子有记忆存档、邻域克隆选择、邻域竞争和邻域协作。与经典的SPEA2和NSGA-II比较效果表明,AIMAMOA在解的收敛性与分布性方面有一定的优势。在分馏系统操作优化应用中,用该算法获得的负荷分配系数和回流比能均衡塔的能耗及塔釜产品的产量。
Fractional distillation process to minimize energy consumption and output is the contradictory goal. In this paper, an artificial immune multi-agent multi-objective optimization algorithm (AIMAMOA) is proposed. By reassigning the load of the fractionation process and adjusting the reflux ratio of the fractionator, a set of Pareto solutions are obtained to guide the actual production of the fractionation process. AIMAMOA combines the core idea of artificial immune and multi-agent technology. The main operation operators are memory archiving, neighborhood cloning selection, neighborhood competition and neighborhood collaboration. Compared with the classical SPEA2 and NSGA-II, the results show that AIMAMOA has some advantages in the convergence and distribution of solutions. In the application of fractional distillation system operation optimization, the load distribution coefficient and the reflux ratio obtained by the algorithm can equalize the energy consumption of the tower and the output of the tower kettle product.