【摘要】 针对传统飞蛾捕焰(MFO)算法求解复杂函数时后期收敛速度慢与求解精度较低等问题,提出了一种基于快速收敛的飞蛾捕焰(RMFO)算法。采用最大最小距离积的方法来初始化飞蛾群,能够提高算法全局收敛速度并且优化解的质量,同时构造出MFO算法的适应度函数作为寻优函数。将RMFO算法和有K均值聚类算法(KMC)进行交叉迭代,构建基于RMFO优化的KMC算法,求解聚类中心时能够改善聚类性能,可以解决现有KMC算法选取初始聚类中心不确定陷入结果局部最优的问题。实验结果表明,通过用UCI国际通用测试数据库的Iris、Wine和Glass 3种数据集,对RMFO算法和优化KMC算法进行性能测试,提出的RMFO算法更加精准,收敛速度快,不易陷入局部最优解,同时,优化KMC算法的聚类性能更好。更多还原

【Abstract】 In view of the problems of slow convergence speed and low accuracy of the solution of the traditional Moth-flame Optimization(MFO) algorithm in the later stage of solving complex functions,a Moth-flame Capture Optimization(RMFO) algorithm based on rapid convergence is proposed.Firstly,the method of maximum and minimum distance product is used to initialize moth swarm,which can improve the global convergence speed of the algorithm and can optimize the quality of the solution.At the same time,the adaptability function of MFO algorithm is constructed as the optimizing function.Then the cross interation is carried out with RMFO algorithm and KMC algorithm to construct.The optimal KMC algorithm based on RMFO.The solution of the clustering centers can improve the clustering performance and can solve the problem of immersion in locally optimal results KMC algorithm in selecting the uncertainly initial clustering center.The performance of RMFO algorithm and KMC based on RMFO algorithm are tested by IRIS,Wine and Glass data sets of UCI International General Test Database.The results show that the results of RMFO algorithm are more accurate,the convergence speed is faster,and it is not easy to fall into the local optimal solution.And the clustering performance of the optimalized KMC based on RMFO algorithm is better.更多还原