流程雁阵的多目标跟踪优化方法  

An optimization method for multi-objective tracking of Process Goose Queue

作　　者：卢爱飞 黄静雯[1] 李宏光[1] 

机构地区：[1]北京化工大学信息科学与技术学院,北京100029

出　　处：《计算机与应用化学》2014年第9期1127-1132,共6页Computers and Applied Chemistry

摘　　要：流程雁阵(Process Goose Queue,PGQ)[1]为流程系统的分解协调优化提供了一个新的方法,然而,目前PGQ方法尚存在许多不足,如简单地将个体PGQ的状态跟踪处理为单目标优化问题,这与实际流程操作不符;而多级PGQ系统优化仍采用传统的数学规划方法,对模型要求苛刻且依赖于初值的选取。为此,论文提出了一个面向流程雁阵多目标跟踪的优化方法。首先对多级PGQ系统进行了结构优化,然后将NSGA-Ⅱ用于多级PGQ系统中个体PGQ的多目标优化求解,得到Pareto解集;在此基础上,将逼近理想解捧序法(TOPSIS)和扩展傅里叶幅值灵敏度分析法(EFAST)应用于个体PGQ的多目标决策,并从Pareto解集选取最优解在多级PGQ系统中逐级传递,实现流程系统的分解协调优化。仿真实例验证了方法的可行性和有效性。Process Goose Queue is a novel approach to the decomposition-coordination optimization of process systems. However, there are still several limitations in the previous PGQ method, such as considering individual PGQ＇s optimization as single-objective optimization, which is not consistent with practical process operations; and using traditional mathematical programming methods to deal with multi-level PQGs optimization problems, which demand accurate process models and depend on the selection of initial values. In response, a novel optimization method for multi-objective tracking of PGQs is proposed in this paper. Firstly, the multi-level PGQ＇s structure is facilitated. Secondly, NSGA-II methods are applied to solve the individual PGQ＇s optimization and Pareto set is obtained. On the basis, TOPSIS and EFAST methods are employed in the individual PGQ＇s multi-objective decision-making. And the optimal solution selected from the Pareto set is delivered to the next PGQ. A case study is carried out to demonstrate the feasibility and effectiveness of the proposed approaches.

关 键 词：流程雁阵 多目标 NSGA—II TOPSIS 

分 类 号：TP202[自动化与计算机技术—检测技术与自动化装置]