| 智能工厂中的雾计算资源调度 |
| |
| 作者姓名： | 戴志明 周明拓 杨旸 李剑 刘军 |
| |
| 作者单位： | 1. 中国科学院上海微系统与信息技术研究所, 上海 200050;2. 中国科学院大学, 北京 100049;3. 上海雾计算实验室, 上海 201210;4. 上海科技大学, 上海 201210;5. 思科(中国)有限公司上海分公司, 上海 201103 |
| |
| 基金项目： | 上海市科学技术委员会项目（18511106500）资助 |
| |
| 摘 要： | 随着新一代信息技术的发展，许多传统工厂开始向智能工厂转型。如何对智能工厂中海量数据进行处理，从而提高工厂的生产效率仍然是一个严峻的问题。基于智能工厂的特性提出适用于智能工厂的雾计算框架，使用Kubernetes对容器化的智能工厂应用进行自动化部署。并且提出基于遗传算法改进的区间划分遗传调度算法（interval division genetic scheduling arithmetic，IDGSA）对智能工厂中的容器应用进行调度分配。仿真实验表明，与Kubernetes缺省的调度算法相比，IDGSA算法可使数据处理时间减少50%，雾计算资源使用率提高达60%；与传统的遗传算法相比，在迭代次数更少的情况下，可使数据处理时间减少7%，雾计算资源的使用率提高9%。
|
| 关 键 词： | 智能工厂 雾计算 容器 Kubernetes 资源调度 |
| 收稿时间： | 2019-12-13 |
| 修稿时间： | 2020-03-19 |
Fog computing resource scheduling in intelligent factories |
| |
| Authors: | DAI Zhiming ZHOU Mingtuo YANG Yang LI Jian LIU Jun |
| |
| Institution: | 1. Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Science, Shanghai 200050, China;2. University of Chinese Academy of Sciences, Beijing 100049, China;3. Shanghai Institute of Fog Computing Technology, Shanghai 201210, China;4. ShanghaiTech University, Shanghai 201210, China;5. Cisco(China) Co., Ltd. Shanghai Branch, Shanghai 201103, China |
| |
| Abstract: | With the development of next-generation information technology, many traditional factories have begun to transform into smart factories. How to deal with the massive data in smart factories and improve the production efficiency of the factory is still a serious problem. In this paper, based on the characteristics of smart factories, a fog computing framework for smart factories is proposed, and Kubernetes is used to automate the deployment of containerized smart factory applications.And an improved genetic algorithm based interval division genetic algorithm IDGSA (interval division genetic scheduling arithmetic) is proposed to dispatch and allocate container applications in smart factories. Simulation experiments show that compared with Kubernetes' default scheduling algorithm, IDGSA algorithm can reduce data processing time by 50% and increase the utilization rate of fog computing resources by 60%. Compared with traditional genetic algorithms, it has fewer iterations. In this case, the processing time of the data can be reduced by 7%, and the utilization rate of the fog computing resource can be increased by 9%. |
| |
| Keywords: | smart factory fog computing container Kubernetes resource allocation |
|
| 点击此处可从《》浏览原始摘要信息 |
| 点击此处可从《》下载免费的PDF全文 |
|
