基于LCA与NSGA-II的混合算法解多目标优化问题  

A hybrid algorithm based on LCA and NSGA-II solving multi-objective optimization problem

作　　者：刘海燕[1] 潘欣[1] 廖安[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

出　　处：《计算机与应用化学》2014年第12期1461-1466,共6页Computers and Applied Chemistry

基　　金：国家自然科学基金项目(No.21376185);国家高技术研究发展计划项目(2011AA02A206)

摘　　要：针对非支配遗传算法(NSGA-Ⅱ)存在易过早收敛和计算时间长的问题,本文提出一种列队竞争算法(LCA)与非支配排序遗传算法的混合算法(LCA-NSGA-Ⅱ)。通过将列队竞争算法引入NSGA-Ⅱ,均衡全局搜索和局部搜索,以解决NSGA-Ⅱ早熟的问题,并利用一种快速排序方法,降低非支配排序的时间复杂度,采用动态的拥挤度计算方法,保持外部存档集的均匀性。将LCA-NSGA-Ⅱ用于经典测试函数的计算,在收敛性、分布性和运算效率方面都取得了比NSGA-Ⅱ更好的效果。最后,在ASPEN PLUS与MATLAB集成平台上,用LCA-NSGA-Ⅱ算法以烯烃回收率和能耗为目标优化MTO烯烃分离过程,为实际操作优化提供指导方向。Aimed at the drawbacks of premature convergence and low computational efficiency in genetic algorithm,a hybrid algorithm is proposed which couples line-up competition algorithm（LCA） with non-dominant sorting genetic algorithm（NSGA-Ⅱ）,namely LCA-NSGA-Ⅱ.The balance between the global and local search is better adjusted via combining LCA with NSGA-Ⅱ,thus avoiding the ＂precocious＂ phenomena.A fast non-dominant sorting strategy is employed to reduce time complexity of non-dominated sorting,and a dynamic crowding-distance is applied to maintain the diversity of external archive set.LCA-NSGA-Ⅱ performs better in convergence,distribution and operating efficiency on the test problems than NSGA-Ⅱ.Double objective of recovery of olefins and energy consumption in MTO separation process are optimized by LCA-NSGA-Ⅱ on the ASPEN PLUS and MATLAB integration platform,which providing meaningful guidance for the practical operation.

关 键 词：多目标优化 进化算法 非支配排序 烯烃分离 

分 类 号：TQ021.8[化学工程] TP18[自动化与计算机技术—控制理论与控制工程]