求解全局与局部最优解的多模态多目标进化算法研究进展与挑战

Research Progress and Challenges of Multimodal Multiobjective Evolutionary Algorithms for Global and Local Optimal Solutions

摘要: 为了揭示目前求解全局与局部最优解的多模态多目标进化算法研究与发展现状,首先,介绍了具有全局和局部最优解集的多模态多目标优化问题(multimodal multiobjective optimization problem,MMOP),说明了其相关定义和特点;其次,根据现有求解该类问题的进化算法思想给出了一种分类体系,并对其中主要方法的技术特点进行了概述;然后,介绍了目前具有全局和局部最优解集的多模态多目标测试函数集,并给出了常用的评价指标;最后,通过分析领域中的挑战性问题,展望了未来多模态多目标进化算法研究的方向。

Abstract: To reveal the recent research advances of multimodal multiobjective optimization evolutionary algorithms for global and local optimal solutions, first, the multimodal multiobjective optimization problem (MMOP) with global and local optimal solution sets is introduced and its related definitions and characteristics are provided. Second, a systematic category is presented according to the existing algorithms, and the technical characteristics of the methods are analyzed. Third, the test function sets with global and local optimal solution sets are introduced, and the common evaluation indicators are discussed. Finally, by analyzing the challenges, the future research directions of this field are identified.