Skip to main content
SpringerLink
Log in

Interpretable data-driven fault diagnosis method for data centers with composite air conditioning system

  • Research Article
  • Building Systems and Components
  • Published:
Building Simulation Aims and scope Submit manuscript

Abstract

Fault detection and diagnosis are essential to the air conditioning system of the data center for elevating reliability and reducing energy consumption. This study proposed a convolutional neural network (CNN) based data-driven fault detection and diagnosis model considering temporal dependency for composite air conditioning system that is capable of cooling the high heat flux in data centers. The input of fault detection and diagnosis model was an unsteady dataset generated by the experimentally validated transient mathematical model. The dataset concerned three typical faults, including refrigerant leakage, evaporator fan breakdown, and condenser fouling. Then, the CNN model was trained to construct a map between the input and system operating conditions. Further, the performance of the CNN model was validated by comparing it with the support vector machine and the neural network. Finally, the score-weighted class mapping activation method was utilized to interpret model diagnosis mechanisms and to identify key input features in various operating modes. The results demonstrated in the pump-driven heat pipe mode, the accuracy of the CNN model was 99.14%, increasing by around 8.5% compared with the other two methods. In the vapor compression mode, the accuracy of the CNN model achieved 99.9% and declined the miss rate of refrigerant leakage by at least 61% comparatively. The score-weighted class mapping activation results indicated the ambient temperature and the actuator-related parameters, such as compressor frequency in vapor compression mode and condenser fan frequency in pump-driven heat pipe mode, were essential features in system fault detection and diagnosis.

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

A :

area (m2)

b :

bias parameter

C :

CIC score

C D :

flow coefficient

c p :

specific heat capacity (J/(kg·°C))

H :

heat transfer coefficient (W/(m2·°C))

h :

enthalpy (J/kg)

K :

valve opening

M :

convolution kernel height

m :

mass (kg)

\(\dot{m}\) :

mass flow rate (kg/s)

N :

convolution kernel width

n :

rotation speed (rpm)

P :

pressure (Pa)

S :

heat transfer diameter (m)

T :

temperature (°C)

t :

time (s)

u :

velocity (m/s)

V :

volume (m3)

v :

specific volume (m3/kg)

X :

input

x :

system features

Y :

output

z :

axial length (m)

η :

efficiency (%)

ρ :

density (kg/m3)

σ :

activation function

φ :

activation map element

ω :

weighty parameters

a:

ambient

acl:

acceleration pressure drop

c:

condensing

com:

compressor

cs:

cross sectional

dis:

discharge

e:

evaporating

f:

frictional pressure drop

in:

inlet

o:

outlet

p:

refrigerant pump

r:

refrigerant

s:

equal entropy

sh:

superheating

vol:

volume

w:

heat exchanger wall

CACS:

composite air conditioning system

CAM:

class activation mapping

CF:

condenser fouling

CIC:

channel-wise increase of confidence

CNN:

convolutional neuron network

DC:

data center

EB:

evaporator fan breakdown

EEV:

electronic expansion valve

FDD:

fault detection and diagnosis

FF:

fault free

MWM:

moving window method

NN:

neural network

PHP:

pump-driven heat pipe

SVM:

support vector machine

References

  • Asgari S, Gupta R, Puri IK, et al. (2021). A data-driven approach to simultaneous fault detection and diagnosis in data centers. Applied Soft Computing, 110: 107638.

    Article  Google Scholar 

  • Chen Y, Ebrahimifakhar A, Hu Y, et al. (2023a). Generalizability of machine learning-based fault classification for residential air-conditioners. Energy and Buildings, 295: 113263.

    Article  Google Scholar 

  • Chen K, Chen S, Zhu X, et al. (2023b). Interpretable mechanism mining enhanced deep learning for fault diagnosis of heating, ventilation and air conditioning systems. Building and Environment, 237: 110328.

    Article  Google Scholar 

  • Deng Z, Chen Y, Yang J, et al. (2022). Archetype identification and urban building energy modeling for city-scale buildings based on GIS datasets. Building Simulation, 15: 1547–1559.

    Article  Google Scholar 

  • Deng Z, Javanroodi K, Nik VM, et al. (2023). Using urban building energy modeling to quantify the energy performance of residential buildings under climate change. Building Simulation, 16: 1629–1643.

    Article  Google Scholar 

  • Ding T, Chen X, Cao H, et al. (2021). Principles of loop thermosyphon and its application in data center cooling systems: A review. Renewable and Sustainable Energy Reviews, 150: 111389.

    Article  Google Scholar 

  • Du Z, Chen K, Chen S, et al. (2023). Deep learning GAN-based data generation and fault diagnosis in the data center HVAC system. Energy and Buildings, 289: 113072.

    Article  Google Scholar 

  • Eom YH, Yoo JW, Hong SB, et al. (2019). Refrigerant charge fault detection method of air source heat pump system using convolutional neural network for energy saving. Energy, 187: 115877.

    Article  Google Scholar 

  • Fan C, Lin Y, Piscitelli MS, et al. (2023). Leveraging graph convolutional networks for semi-supervised fault diagnosis of HVAC systems in data-scarce contexts. Building Simulation, 16: 1499–1517.

    Article  Google Scholar 

  • Guo Y, Tan Z, Chen H, et al. (2018). Deep learning-based fault diagnosis of variable refrigerant flow air-conditioning system for building energy saving. Applied Energy, 225: 732–745.

    Article  Google Scholar 

  • Han Z, Wei H, Sun X, et al. (2020). Study on influence of operating parameters of data center air conditioning system based on the concept of on-demand cooling. Renewable Energy, 160: 99–111.

    Article  Google Scholar 

  • Han Z, Xue D, Wei H, et al. (2021). Study on operation strategy of evaporative cooling composite air conditioning system in data center. Renewable Energy, 177: 1147–1160.

    Article  Google Scholar 

  • Kanbur BB, Wu C, Fan S, et al. (2021). System-level experimental investigations of the direct immersion cooling data center units with thermodynamic and thermoeconomic assessments. Energy, 217: 119373.

    Article  Google Scholar 

  • Kim D, Kang S, Yoo J, et al. (2021). A study on the quantitative single and dual fault diagnosis of residential split type air conditioners in static operation using support vector machine method. International Journal of Refrigeration, 131: 206–217.

    Article  Google Scholar 

  • Li Z, Shi S, Chen H, et al. (2020). Machine learning based diagnosis strategy for refrigerant charge amount malfunction of variable refrigerant flow system. International Journal of Refrigeration, 110: 95–105.

    Article  Google Scholar 

  • Li G, Yao Q, Fan C, et al. (2021). An explainable one-dimensional convolutional neural networks based fault diagnosis method for building heating, ventilation and air conditioning systems. Building and Environment, 203: 108057.

    Article  Google Scholar 

  • Li X, Zhang C, Sun X, et al. (2022). Experimental study on reliable operation strategy for multi-split backplane cooling system in data centers. Applied Thermal Engineering, 211: 118494.

    Article  Google Scholar 

  • Li G, Wang L, Shen L, et al. (2023a). Interpretation of convolutional neural network-based building HVAC fault diagnosis model using improved layer-wise relevance propagation. Energy and Buildings, 286: 112949.

    Article  Google Scholar 

  • Li G, Chen L, Liu J, et al. (2023b). Comparative study on deep transfer learning strategies for cross-system and cross-operation-condition building energy systems fault diagnosis. Energy, 263: 125943.

    Article  Google Scholar 

  • Li G, Wu Y, Yan C, et al. (2024). An improved transfer learning strategy for short-term cross-building energy prediction using data incremental. Building Simulation, 17: 165–183.

    Article  Google Scholar 

  • Magoulès F, Zhao H, Elizondo D (2013). Development of an RDP neural network for building energy consumption fault detection and diagnosis. Energy and Buildings, 62: 133–138.

    Article  Google Scholar 

  • Nguyen HP, Liu J, Zio E (2020). A long-term prediction approach based on long short-term memory neural networks with automatic parameter optimization by Tree-structured Parzen Estimator and applied to time-series data of NPP steam generators. Applied Soft Computing, 89: 106116.

    Article  Google Scholar 

  • Rogers AP, Guo F, Rasmussen BP (2019). A review of fault detection and diagnosis methods for residential air conditioning systems. Building and Environment, 161: 106236.

    Article  Google Scholar 

  • Song M, Chen K (2023). Numerical study on the optimal power distribution of server racks in a data center. Building Simulation, 16: 983–995.

    Article  Google Scholar 

  • Sun X, Zhang C, Han Z, et al. (2023). Experimental study on a novel pump-driven heat pipe/vapor compression system for rack-level cooling of data centers. Energy, 274: 127335.

    Article  Google Scholar 

  • Taheri S, Ahmadi A, Mohammadi-Ivatloo B, et al. (2021). Fault detection diagnostic for HVAC systems via deep learning algorithms. Energy and Buildings, 250: 111275.

    Article  Google Scholar 

  • Wang H, Wang Z, Du M, et al. (2020). Score-CAM: Score-weighted visual explanations for convolutional neural networks. In: Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Seattle, WA, USA.

  • Wu R, Ren Y, Tan M, et al. (2024). Fault diagnosis of HVAC system with imbalanced data using multi-scale convolution composite neural network. Building Simulation, 17: 371–386.

    Article  Google Scholar 

  • Yang Y, Xu L, Han H, et al. (2023). Soft measurement and prediction of refrigerant leakage based on SVR-LSTM. International Journal of Refrigeration, 152: 303–314.

    Article  Google Scholar 

  • Zhang H, Shao S, Xu H, et al. (2014). Free cooling of data centers: A review. Renewable and Sustainable Energy Reviews, 35: 171–182.

    Article  Google Scholar 

  • Zhang P, Li X, Shi W, et al. (2018). Experimentally comparative study on two-phase natural and pump-driven loop used in HVAC systems. Applied Thermal Engineering, 142: 321–333.

    Article  Google Scholar 

  • Zhang Y, Zhao Y, Dai S, et al. (2022). Cooling technologies for data centres and telecommunication base stations—A comprehensive review. Journal of Cleaner Production, 334: 130280.

    Article  Google Scholar 

  • Zhang Y, Li M, Dong J, et al. (2023a). Study on the impacts of refrigerant leakage on the performance and reliability of datacenter composite air conditioning system. Energy, 284: 129336.

    Article  Google Scholar 

  • Zhang Y, Dong J, Zhang C, et al. (2023b). Dynamic model based study on mode switching characteristics and reliability analysis of data center composite air conditioning system. Applied Thermal Engineering, 227: 120442.

    Article  Google Scholar 

  • Zhao Y, Wen J, Xiao F, et al. (2017). Diagnostic Bayesian networks for diagnosing air handling units faults-part I: Faults in dampers, fans, filters and sensors. Applied Thermal Engineering, 111: 1272–1286.

    Article  Google Scholar 

  • Zhou Z, Li G, Wang J, et al. (2020). A comparison study of basic data-driven fault diagnosis methods for variable refrigerant flow system. Energy and Buildings, 224: 110232.

    Article  Google Scholar 

  • Zhou Z, Li G, Chen H, et al. (2021). Fault diagnosis method for building VRF system based on convolutional neural network: Considering system defrosting process and sensor fault coupling. Building and Environment, 195: 107775.

    Article  Google Scholar 

  • Zhu X, Du Z, Jin X, et al. (2019). Fault diagnosis based operation risk evaluation for air conditioning systems in data centers. Building and Environment, 163: 106319.

    Article  Google Scholar 

  • Zou S, Zhang Q, Yu Y, et al. (2019). Field study on the self-adaptive capacity of multi-split heat pipe system (MSHPS) under nonuniform conditions in data center. Applied Thermal Engineering, 160: 113999.

    Article  Google Scholar 

Download references

Acknowledgements

The authors gratefully acknowledge the support from the National Natural Science Foundation of China (Grant number 52176180). The authors gratefully acknowledge the support from “the open competition mechanism to select the best candidates” key technology project of Liaoning (Grant 2022JH1/10800008).

Author information

Authors and Affiliations

  1. SEP Key Laboratory of Eco-Industry, School of Metallurgy, Northeastern University, Shenyang, 110819, China

    Yiqi Zhang, Fumin Tao, Baoqi Qiu, Xiuming Li & Zongwei Han

  2. Energy Saving and Low-carbon Technology of Process Industry Engineering Research Center of Liaoning Province, Northeastern University, Shenyang, 110819, China

    Yiqi Zhang, Fumin Tao, Baoqi Qiu, Xiuming Li & Zongwei Han

  3. College of Civil Engineering, Hunan University, Changsha, 410082, China

    Yixing Chen

Authors
  1. Yiqi Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Fumin Tao
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Baoqi Qiu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Xiuming Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Yixing Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Zongwei Han
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Zongwei Han: conceptualization, methodology, investigation, validation, writing—review & editing, supervision. Yiqi Zhang: software, writing—original draft, formal analysis, visualization. Fumin Tao: formal analysis, visualization. Baoqi Qiu: formal analysis, visualization. Xiuming Li: writing—review & editing, investigation. Yixing Chen: writing—review & editing, investigation.

Corresponding author

Correspondence to Zongwei Han.

Ethics declarations

Declaration of competing interest

The authors have no competing interests to declare that are relevant to the content of this article. Yixing Chen is a Subject Editor of Building Simulation.

Ethical approval

This study does not contain any studies with human or animal subjects performed by any of the authors.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhang, Y., Tao, F., Qiu, B. et al. Interpretable data-driven fault diagnosis method for data centers with composite air conditioning system. Build. Simul. (2024). https://doi.org/10.1007/s12273-024-1124-7

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1007/s12273-024-1124-7

Keywords

Search

Navigation