基于多种群生物地理学优化的混合智能算法  

Hybrid intelligent algorithm based on multi-population biogeography optimization

作　　者：张姣姣 高岳林 郭二杨 ZHANG Jiaojiao;GAO Yuelin;GUO Eryang(School of Mathematics and Information Science,North Minzu University,Yinchuan Ningxia 750021;Ningxia Key Laboratory of Intellgent Information and Big Data Processing,Yinchuan Ningxia 750021)

机构地区：[1]北方民族大学数学与信息科学学院,宁夏银川750021 [2]宁夏科学计算与智能信息处理协同创新中心,宁夏银川750021

出　　处：《宁夏师范学院学报》2024年第10期24-38,共15页Journal of Ningxia Normal University

基　　金：宁夏自然科学基金项目(2022AAC02043);宁夏高等教育一流学科建设基金项目(NXYLXK2017B09).

摘　　要：为解决生物地理学优化算法容易陷入局部最优,难以充分平衡勘探与开发,且求解速度慢、精度低等问题,提出一种基于多种群组的混合智能算法.该算法将大陆中的栖息地个体根据适应度值排序后分为3个子种群组,针对每个子种群组的性能采用不同的更新策略.在10个基准函数的不同维度上验证改进算法的有效性,并与更具有竞争力的生物地理学优化算法变体和其他先进智能算法进行比较,得出改进的算法有效地提高了高维优化问题的精度和收敛速度的结论.同时,通过工程问题应用表明,改进的算法对约束优化问题是有效的.In order to solve the problems of easily getting stuck in local optima,difficulty in fully balancing exploration and development,slow solving speed,and low accuracy in biogeographic optimization algorithms,a hybrid intelligent algorithm based on multiple groups is proposed.This algorithm sorts the habitat individuals in the mainland into three sub species groups based on their fitness values,and adopts different update strategies for the performance of each sub species group.The effectiveness of the improved algorithm is verified on different dimensions of 10 benchmark functions,and compared with more competitive biogeography optimization algorithm variants and other advanced intelligent algorithms.It is concluded that the improved algorithm effectively improves the precision and convergence speed of high-dimensional optimization problems.At the same time,the application to engineering problems shows that the improved algorithm is also effective for constrained optimization problems.

关 键 词：生物地理学优化 Levy飞行 鲸鱼优化 工程优化 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]