基于佳点集和惯性权重的改进麻雀算法

Improved Sparrow Algorithm Based on Good Point Set and Inertia Weight

摘要: 针对麻雀算法(SSA)局部搜索能力差的问题，提出一种改进的麻雀算法(GSSA)。首先，采用佳点集的方法初始化麻雀个体，增强种群多样性；其次，在发现者位置更新上采用对数惯性权重来协调局部搜索和全局搜索能力，加快收敛速度；同时在跟随者位置更新方式中引入t分布策略，加强全局搜索能力；最后，对6个基准测试函数进行仿真实验表明，GSSA寻优精度与SSA算法相比可提高约51个数量级，与同类改进算法相比精度可提高2个数量级，且寻优速度加快。

Abstract: In order to solve the problem of poor local search ability of sparrow algorithm, this paper proposes an improved sparrow algorithm (GSSA). Firstly, the best point set method is used to initialize individual position of sparrow, which lays the foundation for the diversity of global search; Secondly, log inertia weight is used to coordinate the ability of local search and global search in the location update of discoverer, and the convergence speed is accelerated; Then, the t distribution strategy is introduced into the follower position update mode to enhance the global search ability; Finally, the simulation experiments on six benchmark functions show that the optimization accuracy of GSSA can be improved by about 51 orders of magnitude compared with SSA algorithm, 2 orders of magnitude compared with similar improved algorithms, and the optimization speed is accelerated.