一种混合灰狼结构的麻雀搜索算法的收敛性分析及应用

打开文本图片集

中图分类号：TP301.6 文献标识码：A 文章编号：1673-9868（2025）10-0221-13

Abstract：To address the issues of the sparrow search algorithm，including its propensity for local optima trapping，slow convergence rate，and reliance on the initial solution，this study introduces an enhanced sparrow search algorithm incorporating with an adaptive mixed gray wolf hierarchy. Initially，the algorithm employed the Jiadian set strategy for population initialization，enhancing diversity and boosting the algorithm's convergence speed and precision. Subsequently，a dynamic inertia weight was applied to refine the seekers’position update mechanism，achieving a beter balance between global exploration and local exploitation. Moreover，the Levy flight strategy was utilized to optimize the position update process of the followers， broadening their search scope and mitigating the risk of getting stuck in local optima. Lastly，a hierarchical approach from the gray wolf optimization algorithm was integrated into the alert sparrow's position update，the possibility of sparrow search algorithm falling into the local optimal was reduced，and analysis of the convergence of the algorithm was made. To validate the algorithm's performance，benchmark tests were conducted against five swarm intelligence optimization algorithms，along with comparative experiments on 2D static grid map path planning，further demonstrating the improved algorithm's effectiveness and practicality.

Key words： sparrow search algorithm；Levy flight；dynamic inertia weights； grey wolf hierarchy； pathplanning

随着我国工业转型升级，在大规模设备更新和技术改造等背景下，移动机器人的应用愈加广泛，而机器人路径规划则是实际应用中的重要环节。（剩余9900字）