混合蛙跳算法在云计算资源调度的策略改进

混合蛙跳算法已在云计算资源调度有所运用。针对青蛙种群初始化随机性大、局部搜索盲目、容易陷入局部最优的问题,提出了一种混合蛙跳算法在云计算资源调度的改进策略。该改进策略首先运用SY-MM算法和随机生成方式结合的方法对种群进行初始化,生成适应度较好且保持多样性的青蛙种群;然后对传统蛙跳算法局部搜索中步长公式进行改进,使得能够自适应的去更新步长,进而提升局部搜索能力。通过实验证明改进算法对于云计算中资源调度的时间和负载平衡方面有良好的优化性能。

The improved strategy of shuffled frog leaping algorithm in the resource scheduling of cloud computing

Shuffled Frog Leaping Algorithm (SFLA) has already been applied to the resource scheduling in cloud computing. Here, an improving strategy for the corresponding applications are put forward aimed at some existing issues, such as the high randomicity of frog population initialization, blindness of local search as well as easily sinking into local optima. To begin with, the population initialization was carried out via the combination of the SY-MM algorithm and randomly generation approach, creating multifarious frog population with favourable fitness. Then, the stepsize formula in the local search of traditional frog algorithm was also improved to update the step size by itself and enhance the local search capacity. It has been demonstrated that the modified algorithm possesses good optimal performance for both the resource scheduling time and the load balancing ability in cloud computing.