云计算环境下任务的调度是目前研究的热点,针对任务完成时间和虚拟机资源负载的均衡情况,对云任务调度遗传算法作出改进.根据云环境下虚拟机资源的性能引入虚拟机相对适应度的概念;将标准遗传算法的随机变异操作改进为有目标的变异操作,使虚拟机相对适应度大的虚拟机资源获得更大的变异可能,加快算法的收敛.仿真实验表明,该算法在降低任务完成时间的同时提高了虚拟机资源的负载均衡,是一种有效的云任务调度算法.%Task scheduling, under Cloud computing environment, is a hotspot of current research. This paper, based on task completion time and load balancing of virtual machine resources, improves the Cloud task scheduling genetic algorithm. According to performance of virtual machine resources in a cloud environment, this paper introduces the concept of virtual machine relative fitness. The targeted mutation is improved from Random mutation operation of standard genetic algorithm, which makes virtual machine resource of the big virtual machine relative fitness gain greater mutation and accelerates the convergence of algorithm. Simulation show that the proposed algorithm is a better solution,because it not only reduces the task completion time but also improves the load balancing of the virtual machine resources.
展开▼
