云工作流系统中能耗感知的任务调度算法

发布时间：2018-07-02 23:37

  本文选题：云计算 + 工作流调度　； 参考：《模式识别与人工智能》2016年09期

【摘要】：云工作流系统研究集中在工作流任务执行的时间效率优化,然而时间最优的任务调度方案可能存在不同能耗,因此,文中求解满足时间约束时能耗最优的调度方案.首先改进任务执行能耗模型,设计适用于评价任务调度方案执行能耗的适应度计算方法.然后基于精准调整粒子速度的自适应权重,提出解决任务调度能耗优化问题的自适应粒子群算法.实验表明,文中算法收敛稳定,调度方案执行能耗较低.
[Abstract]:The research of cloud workflow system focuses on the optimization of time efficiency of workflow task execution, but the scheduling scheme of task scheduling with optimal time may have different energy consumption. Therefore, the optimal scheduling scheme of energy consumption when meeting the time constraints is solved in this paper. Firstly, the model of task execution energy consumption is improved, and a method for evaluating the performance energy consumption of task scheduling scheme is designed. Then an adaptive particle swarm optimization (APSO) algorithm is proposed to solve the problem of task scheduling energy consumption optimization based on the adaptive weight of accurately adjusting particle velocity. Experiments show that the proposed algorithm is stable in convergence and low in energy consumption.
【作者单位】： 安徽大学计算机科学与技术学院;
【基金】：国家自然科学基金项目(No.61672034) 教育部社科研究青年基金项目(No.16YJCZH048) 安徽省教育厅自然科学研究重点项目(No.KJ2016A024)资助~~
【分类号】：TP18
，

本文编号：2091370