基于双向奔袭策略的改进狼群算法研究  

Research of Improved Wolf Pack Algorithm Based on Attacking Strategy Towards Two Directions

作　　者：任剑翔 梁雨佳 魏佳蓓 张硕硕 刘国华 REN Jianxiang;LIANG Yujia;WEI Jiabei;ZHANG Shuoshuo;LIU Guohua(School of Mathematics,Southeast University,Nanjing 210000,China)

机构地区：[1]东南大学数学学院,江苏南京210000

出　　处：《控制工程》2024年第8期1392-1404,共13页Control Engineering of China

基　　金：国家自然科学基金面上项目(11871144);科技部重点研发项目子课题(2021YFB3901202)。

摘　　要：基于传统狼群算法(wolf pack algorithm,WPA)的基本思想提出了双向狼群算法(towards two directions wolf pack algorithm,T2WPA)。针对原算法对已知优良信息利用率的不足,T2WPA对猛狼奔袭过程进行改进:引入双向奔袭的概念,并给出新老头狼产生规则和猛狼位置更新机制。基于以上改进思想,分别按照与文献[9]和文献[10]相同的参数,选取文献中对应的基准函数来测试T2WPA的性能,最后将测试结果与文献进行对比。仿真结果表明,T2WPA具有良好的全局搜索能力和计算精度。Based on the traditional wolf pack algorithm(WPA),the wolf pack algorithm towards two directions(T2WPA)is proposed.For the deficiency that the traditional wolf pack algorithm has a low utilization rate of existing excellent information,T2WPA proposes an improvement scheme towards the fierce wolves attacking process:the concept of two-way attacking is introduced,and the old and new leader wolves’generation rule and the fierce wolves’location updating mechanism are given.Based on the improvement given above,we use the same parameters separately as in references[9]and[10],and select the corresponding benchmark functions to test the performance of T2WPA,and then compare the test results with the references.The simulation results show that the T2WPA has great global search ability and computational accuracy.

关 键 词：狼群算法 群体智能 函数优化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]