本文已被：浏览 2321次 下载 840次
投稿时间：2018-12-17
投稿时间：2018-12-17
中文摘要: 提出了一种基于强化学习的云计算虚拟机资源调度问题的解决方案和策略。构建了虚拟机的动态负载调度模型，将虚拟机资源调度问题描述为马尔可夫决策过程。根据虚拟机系统调度模型构建状态空间和虚拟机数量增减空间，并设计了动作的奖励函数。采用Q值强化学习机制，实现了虚拟机资源调度策略。在云平台的虚拟机模型中，对按需增减虚拟机数量和虚拟机动态迁移两种场景下的学习调度策略进行了仿真，验证了该方法的有效性。
Abstract:A solution to cloud computing resource scheduling problem based on reinforcement learning is proposed.The dynamic load scheduling model of the virtual machine is constructed,and the virtual machine resource scheduling problem is described as the Markov decision process.According to the virtual machine system scheduling model,the state space and the number of virtual machines are increased or decreased,and the reward function of the action is designed.The Q-valued reinforcement learning mechanism is used to implement the virtual machine resource scheduling strategy.Finally,in the virtual machine model of the cloud platform,the performance of the learning and scheduling strategy is enhanced under the scenarios of increasing or decreasing the number of virtual machines and virtual machine dynamic migration.The effectiveness of the method is verified.
文章编号：20194018 中图分类号：TP399 文献标志码：
基金项目:
| 作者 | 单位 | |
| 李天宇 | 国网上海电力公司信息通信公司 | litianyu@sh.sgcc.com.cn |
| Author Name | Affiliation | |
| LI Tianyu | State Grid Shanghai Municipal Electric Power Company, Shanghai 200030, China | litianyu@sh.sgcc.com.cn |
引用文本：
李天宇.基于强化学习的云计算资源调度策略研究[J].上海电力大学学报,2019,35(4):399-403.
LI Tianyu.Research on Cloud Computing Resource Scheduling Strategy Based on Reinforcement Learning[J].Journal of Shanghai University of Electric Power,2019,35(4):399-403.
李天宇.基于强化学习的云计算资源调度策略研究[J].上海电力大学学报,2019,35(4):399-403.
LI Tianyu.Research on Cloud Computing Resource Scheduling Strategy Based on Reinforcement Learning[J].Journal of Shanghai University of Electric Power,2019,35(4):399-403.