基于拥挤距离排序的铝电解多目标优化  

Optimization of aluminum electrolysis production process based on crowding distance sorting

作　　者：易军[1] 黄迪[2] 李太福[1,3] 周伟[1] 姚立忠[1] 

机构地区：[1]重庆科技学院电气与信息工程学院,重庆401331 [2]重庆科技学院安全工程学院,重庆401331 [3]重庆大学自动化学院,重庆400044

出　　处：《仪器仪表学报》2015年第11期2502-2509,共8页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(51375520;51374268;51404051;61174015);重庆市自然科学基金重点项目(cstc2012jj B40006);重庆市重大应用技术开发项目(cstc2013yykf C0034);重庆市优秀人才科技训练计划(cstc2013kjrc-qnrc40008);重庆市高校创新团队项目(KJTD201324);重庆高校优秀成果转化项目(KJZH14218);重庆市教委科学技术研究项目(KJ1401301)资助

摘　　要：针对铝电解生产过程难以快速、准确地获得节能减排多目标优化Pareto前沿问题,提出一种基于拥挤距离排序的多目标细菌觅食算法。方法在保证铝电解槽平稳运行的基础上,建立电流效率最大和温室气体排放量最小的多目标优化模型;利用拥挤距离更新外部档案及对菌群步长进行自适应动态调整,以改进种群的收敛性和多样性,最后对优化模型求解。通过实验可知,改进后的算法能快速获得分布均匀的Pareto最优解,运用优化后的决策参数指导生产,能在提高电流效率的同时减少温室气体的排放量,实现铝电解生产过程节能减排的目的。Aiming at the problem that it is hard to quickly and accurately obtain the Pareto front of energy saving and emission reduction of aluminum electrolysis production process,multi-objective bacterial foraging optimization algorithm based on crowding distance sorting（ MBFO-CDS） is proposed. On the basis of alumina reduction cell running smoothly,the multi-objective model is built to find optimal solutions of the maximum current efficiency and the minimum greenhouse gases emission. In order to improve convergence and diversity of population,the crowing distance sorting is used to update the Pareto-archived and adjust the step self-adaptively and dynamically. The multi-objective modeling is resolved with the novel algorithm. The experimental result suggests that,the improved algorithm can obtain uniformly distributed Pareto-optimal solutions quickly. Using optimized decision parameters to guide production can not only improve the current efficiency but also reduce greenhouse gas emissions,so that the purpose of energy conservation and emissions reduction for aluminum electrolysis production process is realized.

关 键 词：铝电解 多目标优化 细菌觅食算法 拥挤距离 PARETO最优解 

分 类 号：TF821[冶金工程—有色金属冶金] TP273[自动化与计算机技术—检测技术与自动化装置]