权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

CNS: Core: Small: Energy and Load Management in Data Centers: Online Optimization and Learning

CNS：核心：小型：数据中心的能源和负载管理：在线优化和学习

基本信息

批准号：
1908298
负责人：
Mohammadhassan Hajiesmaili
金额：
$ 49.91万
依托单位：
University of Massachusetts Amherst
依托单位国家：
美国
项目类别：
Standard Grant
财政年份：
2019
资助国家：
美国
起止时间：
2019-10-01 至 2022-09-30
项目状态：
已结题

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1908298&HistoricalAwards=false
关键词：
CNS Core Small Energy Load

项目摘要

Data centers, the backbone of the internet, are using increasingly massive amounts of power, consuming by some estimates up to 3% of the world's power supply and doubling their appetite every four years. This increasing appetite for power has significant ramifications for both costs -- electricity counts among the largest expenses in operating a data center -- and unwelcome contributions to climate change. Major data centers are working to meet this challenge by creating their own renewable energy sources (such as wind and solar farms) and energy storage systems in addition to the electric grid. But the fluctuation and lack of reliability built into the supply of energy from renewable sources causes significant challenges for data centers attempting to efficiently procure energy from these sources in combination with the grid. This project aims to use proven methods of algorithms and machine learning to optimize the energy procurement and load management of data centers under these challenges. In doing so, it will enable data centers to minimize their energy costs, maximize robustness against uncertainty, and improve their energy footprints. As a result, the successful implementation of this project will have a significant impact on lowering the overall cost of internet services. More broadly, by modeling how to more efficiently use renewable sources and energy storage systems, this research has significance beyond data centers. This project will facilitate the efficient and reliable incorporation of renewables into the operations of other large energy customers, playing a key role in moving toward a greener and more sustainable electric grid. The project will broaden participation by involving undergraduates in the research and will include outreach activities on introducing data centers to local K-12 students.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

数据中心是互联网的支柱，正在使用越来越大量的电力，据估计消耗了全球3%的电力供应，并且每四年就会翻一番。对电力的需求不断增加，这对成本（电力是运营数据中心的最大支出之一）和对气候变化的不受欢迎的贡献都有重大影响。主要的数据中心正在努力通过创建自己的可再生能源（如风能和太阳能发电场）和储能系统来应对这一挑战。但是，可再生能源能源供应的波动和可靠性的缺乏，给试图结合电网从这些来源有效获取能源的数据中心带来了重大挑战。该项目旨在使用经过验证的算法和机器学习方法来优化数据中心在这些挑战下的能源采购和负载管理。通过这样做，它将使数据中心能够最大限度地降低能源成本，最大限度地提高对不确定性的鲁棒性，并改善其能源足迹。因此，该项目的成功实施将对降低互联网服务的总体成本产生重大影响。更广泛地说，通过对如何更有效地使用可再生能源和储能系统进行建模，这项研究的意义超越了数据中心。该项目将促进可再生能源高效、可靠地融入其他大型能源客户的运营，在迈向更绿色、更可持续的电网方面发挥关键作用。该项目将通过让本科生参与研究来扩大参与，并将包括向当地K-12学生介绍数据中心的外联活动。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

期刊论文数量（21）

专著数量（0）

科研奖励数量（0）

会议论文数量（0）

专利数量（0）

Distributed Bandits with Heterogeneous Agents

DOI：
10.1109/infocom48880.2022.9796901
发表时间：
2022-01
期刊：
IEEE INFOCOM 2022 - IEEE Conference on Computer Communications
影响因子：
0
作者：
Lin Yang;Y. Chen;M. Hajiesmaili;John C.S. Lui;D. Towsley
通讯作者：
Lin Yang;Y. Chen;M. Hajiesmaili;John C.S. Lui;D. Towsley

Online Peak-Aware Energy Scheduling with Untrusted Advice

DOI：
10.1145/3447555.3464860
发表时间：
2021-06
期刊：
Proceedings of the Twelfth ACM International Conference on Future Energy Systems
影响因子：
0
作者：
Russell Lee;Jessica Maghakian;M. Hajiesmaili;Jian Li;R. Sitaraman;Zhenhua Liu
通讯作者：
Russell Lee;Jessica Maghakian;M. Hajiesmaili;Jian Li;R. Sitaraman;Zhenhua Liu

Hierarchical Learning Algorithms for Multi-scale Expert Problems

多尺度专家问题的分层学习算法