.png){kind=icon}
返回
Multiobjective Demand Response for Internet Data Centers
.png){kind=icon}
OA
AI
DOI：10.1109/TETCI.2021.3055232.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Optimizing the workload distribution and dictating the number of active servers at Internet data centers (IDCs) over a dynamic pricing scheme in power markets can be a challenging task. This paper develops an efficient multiobjective approach to demand response at IDCs. Two objectives are considered: maximize the data center profits under some constraints and optimize quality of service (QoS) measured by a delay probability. A multiobjective optimization problem is formulated and Pareto optimality is adopted. A multiobjective solution method is proposed, which consists of an algorithm generating feasible points, a mechanism that preserves the feasibility, and a multiobjective evolutionary algorithm (MOEA) for producing Pareto optimal workload and energy scheduling. Scalability and feasibility of the proposed approach are analyzed. Simulation results using real-world data reveal that the proposed demand response approach can systematically achieve an excellent balance between the profits for IDCs and QoS for clients. In contrast with comparable MOEAs, the proposed approach finds feasible points in the decision variable space during the evolution process with a 100% success rate.
Keyword:
Quality of service
Servers
Portals
Data centers
Power demand
Optimization
Load management
Demand response
evolutionary multiobjective optimization (EMO)
Internet data centers (IDCs)
minimum Manhattan distance (MMD) approach
pareto optimality
quality of service (QoS)
multiobjective evolutionary algorithms (MOEAs)
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
I
IF:
6.5
论文数:
1.4K
被引数:
4.5K
机构
引用论文
.png){kind=icon}
分享收藏
分享收藏
查看更多内容