求解聚类问题的异构蝙蝠算法

摘要

为解决蝙蝠算法在较高精度要求下收敛速度慢且易于陷入局部最优等缺陷问题,在蝙蝠算法框架基础上,利用具有良好随机性的Lévy飞行来增强算法的全局搜索能力,结合单纯形法提高算法在局部开采时的性能,提出一种异构的蝙蝠算法.对聚类问题进行解的映射表示并且进行仿真实验.实验数据表明,该算法求解精度高、收敛速度快,具备有效性和可行性,为解决聚类问题提供了可参考的求解算法.%To solve the weakness of bat algorithm of slow convergence in demand of higher precision and easiness of trapping in local optimum, based on the framework of the bat algorithm, Lévy flights with good randomness was used to enhance the global search ability of the algorithm, and the simplex method was combined with bat algorithm to improve the performance during the local mining period.By mapping the solution of clustering problem in bat algorithm, results of the simulation indicate the proposed isomeric bat algorithm has outstanding efficiency and optimization performance.It also shows that the algorithm has high accuracy and convergence speed, and it can provide a reference for solving the clustering problem.