融合精英学习与多项式变异的改进麻雀算法

Improved sparrow algorithm combining elite learning and Polynomial Mutation

针对麻雀搜索算法初期搜索效率低和易陷入局部最优的问题,提出了融合精英学习与多项式变异的改进麻雀算法。算法在麻雀搜索算法的基础上,通过维度空间均匀初始化种群,优化种群分布,均匀种群密度,提高初期搜索效率;然后引入改进的精英学习策略,对发现者搜索范围进行全局性优化,提高算法的开拓能力;最后采用多项式变异对麻雀搜索算法中的最佳位置进行更新,降低算法陷入局部最优解的概率,增强算法的全局寻优能力。通过8个基准函数的仿真和Wilcoxon秩检验,验证上述方法的有效性。最终结果表明,改进麻雀算法的搜索效率和全局寻优能力都有明显的提升。

Aiming at the problem of low efficiency of initial search of the sparrow search algorithm and easy falling into local optimum,an improved sparrow search algorithm based on elite learning and polynomial mutation is proposed.Based on the sparrow search algorithm,we initialized the population uniformly through the dimensional space,optimized the population distribution,uniformed population density,and improved the initial search efficiency;Then improved elite learning strategy was introduced to optimize the search range of the discoverer globally to improve the development of the algorithm ability;Finally,polynomial mutation was used to update the best position in the sparrow search algorithm to reduce the probability of the algorithm falling into the local optimal solution and enhance the algorithm's global optimization ability.Through simulation experiments of 8 benchmark functions and the Wilcoxon rank test,the effectiveness of the above method was verified.The final results show that the search efficiency and global optimization ability of ISSA algorithm have been significantly improved.