中文摘要：

由于云计算环境下并行数据具有不同的特征,采用不同的表达式描述客户对并行计算任务的满意度时,无法使客户所需并行计算任务和客户满意度完全符合调度条件。传统的调度方法,主要以客户满意度作为并行计算调度的前提,没有充分考虑并行计算任务和客户满意度之间的关系对任务调度准确性的影响,导致调度效率差的问题。提出一种基于优化遗传算法的云计算环境下的并行计算调度方法。建立并行计算调度任务模型,将模糊逻辑和虚拟机利用率的方差作为控制变量,利用遗传算法进行并行计算调度策略的寻优,在交叉操作中引入了参数可调的概率函数,在变异操作中引入了自适应变异的概率函数,从而提高了并行计算中的调度效率。仿真结果表明,改进算法能够提高调度效率,效果令人满意。

英文摘要：

Because the parallel computing data has different characteristics under cloud computing environment,when different expressions are used to describe the satisfaction of customer for the parallel computing task,it cannot make the parallel computing tasks required by customer and customer satisfaction to fully comply with the scheduling condition.In traditional scheduling methods,customer satisfaction is mainly used as the premise of the parallel computing scheduling,without fully considering the influence of the relationship between parallel computing task and customer satisfaction on the accuracy of the task scheduling,causing the problem of poor scheduling efficiency.In this paper,a scheduling method of parallel computing under the cloud computing environment based on optimized genetic algorithm is proposed.The scheduling task model of parallel computing is established,and the fuzzy logic and the variance of utilization ratio of virtual machine are as control variable.The optimization of parallel computing scheduling strategy by using the genetic algorithm is made,then,the parameter adjustable probability function is introduced in the crossover operation,and the adaptive mutation probability function is introduced in the mutation operation,so as to improve the scheduling efficiency of the parallel computing.The simulation results show that the improved algorithm can increase the efficiency of scheduling,and the effect is satisfactory.