Presentation 2015-03-02
Temperature prediction for energy optimization in data centers by machine learning approaches
Kazuyuki HASHIMOTO, Yuya TARUTANI, Go HASEGAWA, Kazuhiro MATSUDA, Takumi TAMURA, Yutaka NAKAMURA, Morito MATSUOKA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this report, we propose a temperature prediction method for energy optimization in data centers by machine learning approaches. In the proposed method, we acquire operating data of data center such as air conditioners' setting, server powers, and values of temperature sensors. We then construct regression equations for estimating temperature distributions at several minutes future by machine learning approaches. We evaluate the proposed method by using collected data from actual operating data center in terms of precision and accuracy of the regression equations and calculation time required for learning and estimation. Evaluation results show that we can estimate the tempera- ture distribution at 10 minutes future after changing the operational parameters with 0.095 degree C of accuracy and 0.107 degree C of precision.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Data center / Power consumption reduction / Temperature prediction / Machine learning / Air conditioner
Paper # NS2014-191
Date of Issue

Conference Information
Committee NS
Conference Date 2015/2/23(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Network Systems(NS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Temperature prediction for energy optimization in data centers by machine learning approaches
Sub Title (in English)
Keyword(1) Data center
Keyword(2) Power consumption reduction
Keyword(3) Temperature prediction
Keyword(4) Machine learning
Keyword(5) Air conditioner
1st Author's Name Kazuyuki HASHIMOTO
1st Author's Affiliation Grad. Sen. of Information Science and Technology, Osaka Univ.()
2nd Author's Name Yuya TARUTANI
2nd Author's Affiliation Cybermedia Center, Osaka Univ.
3rd Author's Name Go HASEGAWA
3rd Author's Affiliation Cybermedia Center, Osaka Univ.
4th Author's Name Kazuhiro MATSUDA
4th Author's Affiliation NTT Advanced Technology Corporation
5th Author's Name Takumi TAMURA
5th Author's Affiliation NTT Advanced Technology Corporation
6th Author's Name Yutaka NAKAMURA
6th Author's Affiliation Grad. Sch. of Engineering Science, Osaka Univ.
7th Author's Name Morito MATSUOKA
7th Author's Affiliation ATR
Date 2015-03-02
Paper # NS2014-191
Volume (vol) vol.114
Number (no) 477
Page pp.pp.-
#Pages 6
Date of Issue