首页|期刊导航|江西水利电力大学学报|基于混合策略的大规模多目标人工蜂群算法

基于混合策略的大规模多目标人工蜂群算法OA

Large-scale multi-objective artificial bee colony algorithm based on hybrid strategies

中文摘要英文摘要

人工蜂群算法是求解复杂优化问题的有效方法,然而现有研究多集中于单目标优化或低维决策变量的多目标优化问题.对于大规模多目标优化问题的求解,传统的多目标人工蜂群算法面临着"维数灾难"的挑战,导致搜索效率急剧下降.针对上述问题,本文提出了一种基于混合策略的大规模多目标人工蜂群算法(LMHABC).首先,引入基于随机取样与支配关系的决策变量分组策略,将高维决策变量解耦划分为收敛性变量和多样性变量.其次,建立收敛性档案和多样性档案的双档案协同机制,分别实现收敛性增强与多样性保持.最后,结合决策变量分组结果与双档案精英引导,为不同搜索阶段构建差异化的自适应搜索策略,实现全局探索与局部开采的动态平衡.为验证所提算法的有效性,将其与多个大规模多目标进化算法在 LSMOP 基准测试问题上进行比较.实验结果表明,LMHABC 算法具有较好的收敛性及分布多样性,有效地扩展了人工蜂群算法在求解复杂优化问题上的性能.

Artificial bee colony(ABC)algorithm is a class of effective methods for solving complex optimization problems.However,existing studies mostly focus on single-objective optimization and multi-objective optimization problems with low-dimensional decision variables.For solving large-scale multi-objective optimization problems,traditional multi-objective ABC algorithms are confronted with the challenge of the"curse of dimensionality",which leads to a sharp decline in search effi-ciency.To address the above issues,this paper proposes a large-scale multi-objective artificial bee colony algorithm based on hybrid strategies(named LMHABC).First,a decision variable grouping strategy based on random sampling and dominance relation is introduced to decouple and divide high-dimensional decision variables into convergence-related variables and di-versity-related variables.Then,a dual-archive collaborative mechanism consisting of a convergence archive and a diversity archive is established to achieve convergence enhancement and diversity preservation respectively.Finally,by combining the decision variable grouping results and elite guidance from the dual archives,different adaptive search strategies are construc-ted for different search stages.It aims to make a dynamic balance between global exploration and local exploitation.To verify the effectiveness of the proposed algorithm,it is compared with multiple large-scale multi-objective evolutionary algorithms on the LSMOP benchmark problems.Experimental results show that LMHABC algorithm achieves good convergence and dis-tribution diversity,effectively extending the performance of artificial bee colony algorithm in solving complex optimization problems.

王晖;廖仕涛;钱佳瑶;王文君;王芸

江西水利电力大学 信息工程学院,江西 南昌 330099江西水利电力大学 信息工程学院,江西 南昌 330099江西水利电力大学 信息工程学院,江西 南昌 330099江西水利电力大学 工商管理学院,江西 南昌 330099江西水利电力大学 信息工程学院,江西 南昌 330099

信息技术与安全科学

人工蜂群算法决策变量分析收敛性多样性双档案大规模多目标优化

artificial bee colony algorithmdecision variable analysisconvergencediversitytwo-archivelarge-scale multi-objective optimization

《江西水利电力大学学报》 2026 (4)

68-80,13

国家自然科学基金项目(62166027)

评论