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

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.