改进差分进化算法在铝电解多目标优化中的应用

针对多目标优化问题,提出一种改进的差分进化算法(DE).该改进算法首先将DE与粒子群优化算法(PSO)结合,提高DE的收敛速度,然后引入多种群进化策略,有利于维持Pareto解的多样性.同时,在综合考虑机理与工艺的基础上建立铝电解多目标优化模型,并应用改进算法进行求解.仿真结果表明:在电流效率为92％时,改进算法所得的直流功耗为14.03 MW.h/t,比NSGA-Ⅱ的直流功耗降低了1.45％,比传统DE的直流功耗降低了1.75％.表明本文改进算法有效地提高了传统进化算法的性能.

An improved hybrid differential evolution algorithm used for multi-objective optimization of aluminum electrolysis

An improved hybrid differential evolution algorithm for multi-objective optimization problems was proposed.To increase the convergence speed,differential evolution and particle swarm optimization were combined,and then multi-swarm evolutionary strategy was introduced to maintain the diversity of the Pareto solutions.A multi-objective optimization model of aluminum electrolysis cell was built based on the mechanism and was resolved based on the improved algorithm.The simulation results show that the DC power consumption is 14.03 MW-h/t by using the improved algorithm,which is 1.45% lower than consumption using NSGA-II and is 1.75% lower than the consumption by using DE algorithm.The results indicate the proposed algorithm can effectively improve the performance of traditional evolution algorithms.