Improved snow geese algorithm for engineering applications and clustering optimization

作     者：Bian, Haihong Li, Can Liu, Yuhan Tong, Yuxuan Bing, Shengwei Chen, Jincheng Ren, Quance Zhang, Zhiyuan 

作者机构：Nanjing Inst Technol Coll Elect Engn Chunhua St Nanjing 211167 Jiangsu Peoples R China Nanjing Inst Technol Jiangsu Prov Key Construct Lab Act Distribut Netwo Chunhua St Nanjing 211167 Jiangsu Peoples R China State Grid Zhejiang Elect Power Co Ltd Cixi Power Supply Co Gutang St Cixi 315300 Zhejiang Peoples R China 

出 版 物：《SCIENTIFIC REPORTS》 (Sci. Rep.)

年 卷 期：2025年第15卷第1期

页      面：1-95页

核心收录：

基　　金：Jiangsu Provincial Key Research and Development Program [BE2020688] Jiangsu Provincial Key Research and Development Program 

主　　题：Snow geese algorithm Meta-heuristic algorithm Engineering and clustering optimization Lead goose rotation mechanism Honk-guiding mechanism Outlier boundary 

摘      要：The Snow Goose Algorithm (SGA) is a new meta-heuristic algorithm proposed in 2024, which has been proved to have good optimization effect, but there are still problems that are easy to fall into local optimal and premature convergence. In order to further improve the optimization performance of the algorithm, this paper proposes an improved Snow Goose algorithm (ISGA) based on three strategies according to the real migration habits of snow geese: (1) Lead goose rotation mechanism. (2) Honk-guiding mechanism. (3) Outlier boundary strategy. Through the above strategies, the exploration and development ability of the original algorithm is comprehensively enhanced, and the convergence accuracy and convergence speed are improved. In this paper, two standard test sets of IEEE CEC2022 and IEEE CEC2017 are used to verify the excellent performance of the improved algorithm. The practical application ability of ISGA is tested through 8 engineering problems, and ISGA is employed to enhance the effect of the clustering algorithm. The results show that compared with the comparison algorithm, the proposed ISGA has a faster iteration speed and can find better solutions, which shows its great potential in solving practical optimization problems.