免疫标准化正规化约束方法及其应用

Immune normalized-normal-constraint method and its application

中文关键词  免疫算法  标准化正规化约束方法  冷连轧轧制规程  多目标优化

英文关键词  immune algorithm  normalized-normal-constraint method  rolling schedules of tandem cold rolling  multiobjective optimization

中文摘要

      为了加快准化正规化约束(normalized normal constraint, 简称NNC)方法求解多目标优化问题的速度, 将免疫算法与NNC方法相结合提出了基于免疫算法的NNC方法, 简称免疫NNC(IA NNC)方法. 该方法利用免疫算法中的免疫接种技术, 从相邻的乌托邦面上的点对应的单目标优化问题的优化过程中提取疫苗, 对初始抗体群进行疫苗接种; 使用克隆选择算法求解NNC方法中的单目标优化问题, 进而使IA NNC方法能够更快的获得多目标优化问题的Pareto解集. 之后对IA NNC方法的收敛性进行了分析. 最后应用IA NNC方法对冷连轧轧制规程进行多目标优化,结果表明与基于遗传算法的NNC方法相比, IA NNC方法用较少的运行时间获得了更好的冷连轧轧制规程多目标优化问题的Pareto解集.

英文摘要

      To accelerate the solving process of the multi-objective optimization problems by using the normalizednormal-constraint (referred to as NNC) method, we propose an immune algorithm called the IA NNC method by combining the immune algorithm with the NNC method. It uses the clonal-selection algorithm to solve the single-objective optimization problem by the NNC method; and extracts vaccines from the single-objective optimization process corresponding to the nearby points on the utopia plane. These vaccines are inoculated to the initial antibody population by using the vaccineinoculation technique of the immune algorithm. By the combination of the above two methods, the IA NNC algorithm generates the Pareto solution-set more rapidly. Furthermore, the convergence of IA NNC method is analyzed. Finally, the IA NNC method is applied to optimize the multi-objective scheduling for the tandem cold rolling; it generates the Pareto solution-set for the rolling schedules with less time consumption compared with the genetic algorithm-based NNC method.