Skip to main content
SpringerLink
Log in

Hot spot temperature prediction and operating parameter estimation of racks in data center using machine learning algorithms based on simulation data

  • Research Article
  • Advances in Modeling and Simulation Tools
  • Published:
Building Simulation Aims and scope Submit manuscript

Abstract

In this paper, models to predict hot spot temperature and to estimate cooling air’s working parameters of racks in data centers were established using machine learning algorithms based on simulation data. First, simulation models of typical racks were established in computational fluid dynamics (CFD). The model was validated with field test results and results in literature, error of which was less than 3%. Then, the CFD model was used to simulate thermal environments of a typical rack considering different factors, such as servers’ power, which is from 3.3 kW to 20.1 kW, cooling air’s inlet velocity, which is from 1.0 m/s to 3.0 m/s, and cooling air’s inlet temperature, which is from 16 °C to 26 °C The highest temperature in the rack, also called hot spot temperature, was selected for each case. Next, a prediction model of hot spot temperature was built using machine learning algorithms, with servers’ power, cooling air’s inlet velocity and cooling air’s inlet temperature as inputs, and the hot spot temperatures as outputs. Finally, based on the prediction model, an operating parameters estimation model was established to recommend cooling air’s inlet temperatures and velocities, which can not only keep the hot spot temperature at the safety value, but are also energy saving.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

Abbreviations

c p :

specific heat capacity (J/kg)

d 1 :

distance between the server and the air inlet (m or cm)

d 2 :

distance between the server and the air outlet (m or cm)

F :

body forces (N)

H :

height of rack (m or cm)

h :

height of server (m or cm)

h 1 :

height between the server and the rack bottom (m or cm)

k :

turbulent kinetic energy

k T :

heat transfer coefficient (W/(m2·°C))

L :

length of rack (m or cm)

l :

length of server (m or cm)

N :

number of servers

P :

heating power of the rack (kW)

P F :

pressure (Pa)

R 2 :

coefficient of determination

T :

air temperature (°C)

T*:

hot spot indicating temperature (°C)

t :

time (s)

V :

air velocity (m/s)

W :

width of rack (m or cm)

w :

width of server (m or cm)

Y M :

contribution of fluctuating expansion

δ :

distance between servers (m or cm)

ε :

dissipation rate

μ t :

turbulent viscosity coefficient

ρ :

air density (kg/m3)

σ k, σ ε :

turbulent Prandtl numbers

τ:

viscous stress (N)

ANN:

artificial neural network

CART:

classification and regression tree

CFD:

computational fluid dynamics

EOP:

effective operating parameters of CFD simulation

EOP*:

effective operating parameters of estimation model

GPR:

Gaussian process regression

RF:

random forest

RMSE:

root-mean-square error

SVR:

support vector regression

in:

inlet

out:

outlet

top:

top point

References

  • Ahmadi VE, Erden HS (2020). A parametric CFD study of computer room air handling bypass in air-cooled data centers. Applied Thermal Engineering, 166: 114685.

    Article  Google Scholar 

  • Asgari S, MirhoseiniNejad S, Moazamigoodarzi H, et al. (2021a). A gray-box model for real-time transient temperature predictions in data centers. Applied Thermal Engineering, 185: 116319.

    Article  Google Scholar 

  • Asgari S, Moazamigoodarzi H, Tsai PJ, et al. (2021b). Hybrid surrogate model for online temperature and pressure predictions in data centers. Future Generation Computer Systems, 114: 531–547.

    Article  Google Scholar 

  • ASHRAE (2011). ASHRAE Thermal Guidelines for Data Processing Environments—Expanded Data Center Classes and Usage Guidance. Atlanta, GA, USA: American Society of Heating, Refrigerating and Air-Conditioning Engineers.

    Google Scholar 

  • ASHRAE (2015). Thermal Guideline for Liquid Cooled Data Processing Environments. Atlanta, GA, USA: American Society of Heating, Refrigerating and Air-Conditioning Engineers.

    Google Scholar 

  • Athavale J, Yoda M, Joshi Y (2019). Comparison of data driven modeling approaches for temperature prediction in data centers. International Journal of Heat and Mass Transfer, 135: 1039–1052.

    Article  Google Scholar 

  • Ban D (2019). Research on Energy-saving Control system of intelligent street lamp based on neural network. Master Thesis, University of Electronic Science and Technology of China. (in Chinese)

  • Chang Y, Niu W, Wang W (2009). Numerical simulation and flow analysis of labyrinth path of drip irrigation emitters. Journal of Northwest A & F University (Natural Science Edition), 37(2): 203–208. (in Chinese)

    Google Scholar 

  • Chen X, Tu R, Li M, et al. (2022). Prediction models of air outlet states of desiccant wheels using multiple regression and artificial neural network methods based on criterion numbers. Applied Thermal Engineering, 204: 117940.

    Article  Google Scholar 

  • Chinese Association of Refrigeration (2021). China Data Center Cooling Technology Annual Development Research Report. Beijing: China Architecture and Building Press. (in Chinese)

    Google Scholar 

  • Cho J, Woo J (2020). Development and experimental study of an independent row-based cooling system for improving thermal performance of a data center. Applied Thermal Engineering, 169: 114857.

    Article  Google Scholar 

  • Fulpagare Y, Bhargav A, Joshi Y (2019). Dynamic thermal characterization of raised floor plenum data centers: Experiments and CFD. Journal of Building Engineering, 25: 100783.

    Article  Google Scholar 

  • Gao C, Yu Z, Wu J (2013). Thermal environment analysis and air distribution optimization for a typical data room. Heating Ventilating & Air Conditioning, 43(9): 101–106. (in Chinese)

    Google Scholar 

  • Gao C, Yu Z, Wu J (2015). Investigation of airflow pattern of a typical data center by CFD simulation. Energy Procedia, 78: 2687–2693.

    Article  Google Scholar 

  • Ge X (2021). Based on random forest regression algorithm of grass root period forecast system research. Master Thesis, Nanjing University of Information Science and Technology. (in Chinese)

  • Grand View Research (2019). Virtual Data Room Market Size & Share Report (2020–2027). Available at https://www.grandviewresearch.com/industry-analysis/virtual-data-room-market.

  • Hassan NMS, Khan MMK, Rasul MG (2013). Temperature monitoring and CFD analysis of data centre. Procedia Engineering, 56: 551–559.

    Article  Google Scholar 

  • He Z, He Z, Zhang X, et al. (2016). Study of hot air recirculation and thermal management in data centers by using temperature rise distribution. Building Simulation, 9: 541–550.

    Article  Google Scholar 

  • He Y, Schnabel MA, Mei Y (2020). A novel methodology for architectural wind environment study by integrating CFD simulation, multiple parametric tools and evaluation criteria. Building Simulation, 13: 609–625.

    Article  Google Scholar 

  • Huang Z, Dong K, Sun Q, et al. (2017). Numerical simulation and comparative analysis of different airflow distributions in data centers. Procedia Engineering, 205: 2378–2385.

    Article  Google Scholar 

  • Lee YT, Wen CY, Shih YC, et al. (2022). Numerical and experimental investigations on thermal management for data center with cold aisle containment configuration. Applied Energy, 307: 118213.

    Article  Google Scholar 

  • Liu Y (2018). Application research of finite element model modification based on Gaussian Process Regression (GPR). Master Thesis, Nanchang University. (in Chinese)

  • Liu Y, Le DV, Tan R (2022). A data-assisted first-principle approach to modeling server outlet temperature in air free-cooled data centers. Future Generation Computer Systems, 129: 225–235.

    Article  Google Scholar 

  • Meng X, Zhou J, Zhang X, et al. (2020). Optimization of the thermal environment of a small-scale data center in China. Energy, 196: 117080.

    Article  Google Scholar 

  • MOHURD (2017). GB50174-2017: Data Center Design Code. Ministry of Housing and Urban-Rural Development of China. (in Chinese)

  • Park BR, Choi YJ, Choi EJ, et al. (2022). Adaptive control algorithm with a retraining technique to predict the optimal amount of chilled water in a data center cooling system. Journal of Building Engineering, 50: 104167.

    Article  Google Scholar 

  • Qin B, Yang X, Ya T, et al. (2020). Research on air distribution of data center cold channel closed room based on thermal environment evaluation index. Heating Ventilating & Air Conditioning, 50(8): 123–128. (in Chinese)

    Google Scholar 

  • Rojas R (1996). Neural Networks: A Systematic Introduction. Berlin: Springer.

    Book  MATH  Google Scholar 

  • Rumelhart D, McClelland J (1986). Parallel Distributed Processing. Cambridge, MA, USA: MIT Press.

    Book  Google Scholar 

  • Saiyad A, Fulpagare Y, Bhargav A (2022). Comparison of detached eddy simulation and standard k-s RANS model for rack-level airflow analysis inside a data center. Building Simulation, 15: 1595–1610.

    Article  Google Scholar 

  • Su Z (2021). Design optimization and analysis of data rack inlet air temperature for an IDC center. Building Energy & Environment, 40(4): 55–59. (in Chinese)

    Google Scholar 

  • Sun D (2004). Research on classification and regression method of support vector machine. PhD Thesis, Central South University. (in Chinese)

  • Tian Y (2005). Support Vector Regression and its application. PhD Thesis, China Agricultural University. (in Chinese)

  • Tian H (2012). Research on cooling technology of high yield heat density data room. PhD Thesis, Tsinghua University. (in Chinese)

  • Tian J (2018). State estimation of urban water supply network based on regression tree. Master Thesis, North China University of Water Resources and Electric Power. (in Chinese)

  • Tian H, Liang H, Li Z (2019). A new mathematical model for multi-scale thermal management of data centers using entransy theory. Building Simulation, 12: 323–336.

    Article  Google Scholar 

  • Wang N (2016). Numerical simulation and energy-saving optimization of thermal environment in data center. Master Thesis, Wuhan Institute of Technology. (in Chinese)

  • Wang R, Kuru E, Yan Y, et al. (2022). Sensitivity analysis of factors controlling the cement hot spot temperature and the corresponding well depth using a combined CFD simulation and machine learning approach. Journal of Petroleum Science and Engineering, 208: 109617.

    Article  Google Scholar 

  • Wang S, Tu R, Chen X, et al. (2023). Thermal performance analyses and optimization of data center centralized-cooling system. Applied Thermal Engineering, 222: 119817.

    Article  Google Scholar 

  • Watson B, Venkiteswaran VK (2017). Universal cooling of data centres: A CFD analysis. Energy Procedia, 142: 2711–2720.

    Article  Google Scholar 

  • Xu L, Zhao T (2013). Comparative study on turbulence model based on hydraulic performance of emitter. Gansu Water Resources and Hydropower Technology, 49(05): 18–20. (in Chinese)

    Google Scholar 

  • Yang S (2020). Gaussian process regression method for functional data research and application. Master Thesis, Dalian University of Technology. (in Chinese)

  • Yuan S, Lu S (2015). Simulation research on airflow in data center under various supply air temperature. Low Temperature and Superconductivity, 43(10): 55–61. (in Chinese)

    Google Scholar 

  • Yuan X, Wang Y, Liu J, et al. (2018). Experimental and numerical study of airflow distribution optimisation in high-density data centre with flexible baffles. Building and Environment, 140: 128–139.

    Article  Google Scholar 

  • Yuan X, Zhou X, Liu J, et al. (2019). Experimental and numerical investigation of an airflow management system in data center with lower-side terminal baffles for servers. Building and Environment, 155: 308–319.

    Article  Google Scholar 

  • Yuan X, Xu X, Liu J, et al. (2020). Improvement in airflow and temperature distribution with an in-rack UFAD system at a high-density data center. Building and Environment, 168: 106495.

    Article  Google Scholar 

  • Zapater M, Risco-Martín JL, Arroba P, et al. (2016). Runtime data center temperature prediction using Grammatical Evolution techniques. Applied Soft Computing, 49: 94–107.

    Article  Google Scholar 

  • Zhang L, Zhang Q, Xu F, et al. (2002). Application and research of statistical analysis model based on classification regression tree (CART) method. Journal of Zhejiang University of Technology, 30(4): 315–318. (in Chinese)

    Google Scholar 

  • Zhu Y (2018). Numerical simulation and feasibility study of combined air supply in data center computer room. Master Thesis, Nanjing Normal University. (in Chinese)

Download references

Acknowledgements

The authors appreciate support of the project from China Electronics Engineering Design Institute CO., LTD. (No. SDIC2021-08), and from the Beijing Natural Science Foundation (No. 4212040).

Author information

Authors and Affiliations

  1. School of Civil and Resources Engineering, University of Science and Technology Beijing, Beijing, 100083, China

    Xianzhong Chen, Rang Tu & Ming Li

  2. School of Automation and Electrical Engineering, University of Science and Technology Beijing, Beijing, 100083, China

    Xu Yang

  3. China Electronics Engineering Design Institute CO., LTD., Beijing, 100142, China

    Kun Jia

Authors
  1. Xianzhong Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Rang Tu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Ming Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Xu Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Kun Jia
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Rang Tu.

Ethics declarations

The authors have no competing interests to declare that are relevant to the content of this article.

Electronic Supplementary Material

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chen, X., Tu, R., Li, M. et al. Hot spot temperature prediction and operating parameter estimation of racks in data center using machine learning algorithms based on simulation data. Build. Simul. 16, 2159–2176 (2023). https://doi.org/10.1007/s12273-023-1022-4

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s12273-023-1022-4

Keywords

Search

Navigation