混合递进多目标进化算法及其在flow shop排序中的应用

A Hybrid Escalating Multi-objective Evolutionary Algorithm with Its Application to Flow Shop Problems

提出一种混合递进多目标进化算法(HEMEA):通过在进化搜索过程中引入递进模式的精英保留、群体重构以及可变邻域非劣解局部搜索策略,增强了算法的求解效率.将算法应用于一系列标准双目标flow shop算例及一个典型三目标flow shop问题,研究结果验证了算法的有效性.

A hybrid escalating multi-objective evolutionary algorithm（HEMEA）, which has a new evolution structure compared with the existing ones, was proposed in this paper. The new algorithm enhanced the efficiency of optimization by using an innovative escalating evolutionary scheme with an elitism selection and variable Pareto local search strategy. A series of bi-objective flow shop optimization problems from OR-Library and one typical tri-objective flow shop optimization problem which was first studied in Bagchi＇s work, were re-optimized by NSGA-II, MOGLS, ENGA and our HEMEA respectively. The comparison of the optimization results have shown the outstanding performance of HEMEA with respect to the others＇, which were well-known for their good performance in multiobjective evolutionary computation, thus, the effectiveness and efficiency of HEMEA was demonstrated.