基于多策略融合的改进灰狼算法.pdfVIP

2025-07兵工自动化

44(7)OrdnanceIndustryAutomation·31·

doi:10.7690/bgzdh.2025.07.007

基于多策略融合的改进灰狼算法

122

文竹，韦杏琼，刘静怡

(1.广西警察学院信息技术学院，南宁530028；2.广西民族大学人工智能学院，南宁530006)

摘要：针对当前各类路径优化算法搜索规模较小、收敛速度较慢、全局搜索与局部搜索不平衡等问题，提出一

种多策略融合的改进灰狼优化算法(multi-strategyfusionofgreywolfoptimizationalgorithm，MGWO)。通过引入精

英反向优化策略对种群进行初始化，提高初始解的质量。采用自适应权重机制，动态调整最优狼的领导能力。通过

分段搜索方法，提升平衡局部搜索与全局探索的能力。仿真实验结果表明：该算法表现出色，能快速找到最优路径，

提高算法的整体性能，具有一定借鉴作用。

关键词：改进灰狼算法；精英反向策略；自适应权重；分段策略；路径优化

中图分类号：O221;TP183文献标志码：A

ImprovedGreyWolfAlgorithmBasedonMulti-strategyFusion

122

WenZhu,WeiXingqiong,LiuJingyi

(1.SchoolofInformationTechnology,GuangxiPoliceCollege,Nanning530028,China;

2.SchoolofArtificialIntelligence,GuangxiMinzuUniversity,Nanning530006,China)

Abstract:Inordertosolvetheproblemsofsmallsearchscale,slowconvergencespeedandimbalancebetweenglobal

searchandlocalsearchincurrentpathoptimizationalgorithms,amulti-strategyfusionofgreywolfoptimizationalgorithm

(MGWO)isproposed.Thequalityoftheinitialsolutionisimprovedbyintroducingtheelitereverseoptimizationstrategy

toinitializethepopulation.Anadaptiveweightmechanismisusedtodynamicallyadjusttheleadershipoftheoptimalwolf.

Theabilityofbalancinglocalsearchandglobalexplorationisimprovedthroughthepiecewisesearchmethod.The

simulationresultsshowthatthealgorithmperformswell,canquicklyfindtheoptimalpath,andimprovetheoverall