摘要
针对云计算中的任务调度的存在效率低的问题,首先分析了云计算任务调度目标,其次,对蚁群算法的信息素设置质量函数、对选择概率设置经验反馈因子进行了改进,对帝国竞争算法中的殖民地计算方式、边界值的处理都采用了新的方式,最后将两种算法进行融合,得到基于蚁群算法-帝国竞争算法的云计算任务调度算法,仿真实验中,在任务执行时间,执行成本和负载率方面,本文算法都具有一定的优越性。
Aiming at the problem of low efficiency when scheduling cloud computing tasks,the target of scheduling cloud computing tasks is analyzed at first,and then setting mass function for the pheromones of ACO,advancing the set feedback factors for selective probability,and new methods are applied for the calculation of colony in ICA and the process of boundary value. Lastly,combining these two algorithms together,and the ACO-ICA-based algorithm for scheduling cloud computing tasks is made. In the stimulation experience,this algorithm have advantages in the aspects of time consumed,cost,ratios of loading and so on.
引文
[1]王兴柱.多级优化的云计算任务智能调度算法[J],控制工程,2017,24(5):1008-1012.
[2]黄婷婷,梁意文.云工作流任务调度的模拟退火遗传改进算法[J].微电子学与计算机,2016,33(1):42-46.
[3]王俊英.基于概率自适应蚁群算法的云任务调度方法[J].郑州大学学报(工学版),2017,38(4):51-56.
[4]孙凌宇.云计算环境下基于禁忌搜索的负载均衡任务调度优化算法[J].小型微型计算机系统,2015,36(9):1948-1952.