多策略改进的麻雀搜索算法及其应用  

Multi-strategy improved sparrow search algorithm and its application

作　　者：周军[1] 孔菁菁 ZHOU Jun;KONG Jing-jing(College of Electrical Engineering,Northeast Electric Power University,Jilin 132012,China)

机构地区：[1]东北电力大学电气工程学院,吉林吉林132012

出　　处：《东北师大学报（自然科学版）》2024年第4期81-88,共8页Journal of Northeast Normal University(Natural Science Edition)

基　　金：吉林省科技发展计划项目(20230203033SF)。

摘　　要：针对麻雀搜索算法解决复杂问题时存在收敛过早、全局寻优能力不足等问题,提出一种改进的麻雀搜索算法(ISSA).首先,引入麻雀个体方向信息和种群方向信息对追随者位置更新,提高算法全局搜索能力;其次,动态调整警戒者数量来扩大麻雀搜索范围,采用二元锦标赛选择策略选择警戒者个体,增加麻雀种群多样性;最后,加入固定扰动项,为其提供跳出局部最优的机会.在CEC2013测试集中,将ISSA与其他5种优化算法在相同维度上进行仿真实验,比较各算法的Friedman检验和Wilcoxon秩和检验.结果表明,ISSA在收敛性能和稳定性等方面均具有显著优势,适用于解决复杂问题.Aiming at the problems of early convergence and insufficient global optimization ability of sparrow search algorithm when solving complex problems,an improved sparrow search algorithm(ISSA)was proposed.Firstly,the individual direction information and population direction information were introduced to update the follower s location and improve the global search ability of the algorithm.Secondly,the number of watchmen was dynamically adjusted to expand the search range of sparrows,and the binary tournament selection strategy was used to select individual watchmen to increase the diversity of sparrow population.Finally,a fixed perturbation term is added to provide an opportunity to escape the local optimal.In the CEC2013 test set,ISSA and other five optimization algorithms were simulated in the same dimension,and the Friedman test and Wilcoxon rank sum test of each algorithm were compared.The comparison results showed that ISSA had significant advantages in convergence performance and stability,and was suitable for solving complex problems.

关 键 词：麻雀搜索算法 元启发式算法 锦标赛选择策略 消防疏散路径 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]