摘要
针对人工蜂群算法存在的探索与开采失衡、收敛精度不足等性能短板,提出一种耦合动态拥挤选择与迁移机制的变维扰动人工蜂群算法MABC-ICBO。该算法在雇佣蜂阶段引入基于种群历史多样性的自适应拥挤度调节机制,采用分组抽样策略构建拥挤邻域父代个体集,通过动态竞争替换维持种群多样性;在观察蜂阶段融合生物地理学优化迁移算子与基于种群历史最优的双偏好开采方程,构建混合选择机制增强局部搜索能力;同时提出线性动态维度扰动策略,通过适时调整个体搜索维度规模实现探索与开采的资源优化配置。基于CEC’2017测试集的实验表明,MABC-ICBO在19/29测试函数上显著优于6种对比算法,且Friedman排名第一。在山区静态障碍物场景下的无人机航迹规划任务中,MABC-ICBO较对比算法平均降低飞行成本1.25%~7.40%,生成的航迹路径更平滑且安全性更高。
To address the performance limitations of artificial bee colony(ABC)algorithms including exploration-exploitation imbalance and insufficient convergence accuracy,this paper proposes a variable-dimension perturbation ABC algorithm(MABC-ICBO)integrating dynamic crowding selection and migration mechanisms.In the employed bee phase,an adaptive crowding regulation mechanism based on population historical diversity is introduced,which constructs a crowded neighborhood parent set through a grouped sampling strategy and maintains population diversity via dynamic competitive replacement.During the onlooker bee phase,a hybrid selection mechanism is established by combining a biogeographybased optimization migration operator with dual-preference exploitation equations derived from population historical optima,effectively enhancing local search capability.Additionally,a linear dynamic dimension perturbation strategy is developed to optimize resource allocation between exploration and exploitation by adjusting individual search dimension scales.Experimental results on the CEC'2017 benchmark demonstrate that MABC-ICBO significantly outperforms six comparative algorithms on 19/29 test functions,achieving the first rank in Friedman test.In the task of UAV route planning within mountainous environments containing static obstacles,the MABC-ICBO algorithm reduces the average flight cost by 1.25%to 7.40%compared to other algorithms,generating smoother trajectories with higher safety.
作者
于海波
雷英铎
曾建潮
YU Haibo;LEI Yingduo;ZENG Jianchao(School of Computer Science and Technology,North University of China,Taiyuan 030051,China;School of Electrical and Control Engineering,North University of China,Taiyuan 030051,China)
出处
《计算机工程与应用》
北大核心
2026年第1期192-208,共17页
Computer Engineering and Applications
基金
国家自然科学基金青年基金项目(62106237)
国家自然科学基金联合基金项目(U21A20524)。
关键词
人工蜂群算法
拥挤选择策略
迁移算子
动态维度扰动
无人机航迹规划
artificial bee colony
crowded selection strategy
migration operator
dynamic dimension perturbation
unmanned aerial vehicle(UAV)trajectory planning
