Journal Home > Volume 6 , Issue 2

Most heating, ventilation, and air-conditioning (HVAC) systems operate with one or more faults that result in increased energy consumption and that could lead to system failure over time. Today, most building owners are performing reactive maintenance only and may be less concerned or less able to assess the health of the system until catastrophic failure occurs. This is mainly because the building owners do not previously have good tools to detect and diagnose these faults, determine their impact, and act on findings. Commercially available fault detection and diagnostics (FDD) tools have been developed to address this issue and have the potential to reduce equipment downtime, energy costs, maintenance costs, and improve occupant comfort and system reliability. However, many of these tools require an in-depth knowledge of system behavior and thermodynamic principles to interpret the results. In this paper, supervised and semi-supervised machine learning (ML) approaches are applied to datasets collected from an operating system in the field to develop new FDD methods and to help building owners see the value proposition of performing proactive maintenance. The study data was collected from one packaged rooftop unit (RTU) HVAC system running under normal operating conditions at an industrial facility in Connecticut. This paper compares three different approaches for fault classification for a real-time operating RTU using semi-supervised learning, achieving accuracies as high as 95.7% using few-shot learning.


menu
Abstract
Full text
Outline
About this article

Semi-Supervised Machine Learning for Fault Detection and Diagnosis of a Rooftop Unit

Show Author's information Mohammed G. Albayati1,Jalal Faraj2,Amy Thompson1Prathamesh Patil3Ravi Gorthala3Sanguthevar Rajasekaran2( )
Department of Mechanical Engineering, School of Engineering, University of Connecticut, Storrs, CT 06269, USA
Department of Computer Science and Engineering, University of Connecticut, Storrs, CT 06269, USA
Department of Mechanical and Industrial Engineering, Tagliatela College of Engineering, University of New Haven, West Haven, CT 06516, USA

Mohammed G. Albayati and Jalal Faraj contributed equally to this article.

Abstract

Most heating, ventilation, and air-conditioning (HVAC) systems operate with one or more faults that result in increased energy consumption and that could lead to system failure over time. Today, most building owners are performing reactive maintenance only and may be less concerned or less able to assess the health of the system until catastrophic failure occurs. This is mainly because the building owners do not previously have good tools to detect and diagnose these faults, determine their impact, and act on findings. Commercially available fault detection and diagnostics (FDD) tools have been developed to address this issue and have the potential to reduce equipment downtime, energy costs, maintenance costs, and improve occupant comfort and system reliability. However, many of these tools require an in-depth knowledge of system behavior and thermodynamic principles to interpret the results. In this paper, supervised and semi-supervised machine learning (ML) approaches are applied to datasets collected from an operating system in the field to develop new FDD methods and to help building owners see the value proposition of performing proactive maintenance. The study data was collected from one packaged rooftop unit (RTU) HVAC system running under normal operating conditions at an industrial facility in Connecticut. This paper compares three different approaches for fault classification for a real-time operating RTU using semi-supervised learning, achieving accuracies as high as 95.7% using few-shot learning.

Keywords: energy efficiency, fault detection and diagnostics, data-driven modeling, semi-supervised machine learning, fault classification, heating, ventilation, and air-conditioning

References(34)

[1]
What’s on Your Roof? Rooftop Unit (RTU) Efficiency Advice and Guidance from the Advanced RTU Campaign, https://www.energy.gov/eere/buildings/articles/what-s-your-roof-rooftop-unit-rtu-efficiency-advice-and-guidance-advanced, 2021.
[2]
A. Ebrahimifakhar, A. Kabirikopaei, and D. Yuill, Data-driven fault detection and diagnosis for packaged rooftop units using statistical machine learning classification methods, Energy Build., vol. 225, p. 110318, 2020.
[3]
J. Braun and D. Yuill, FDD Evaluator 0.1, Ray W. Herrick Laboratories-Purdue University, West Lafayette, 2013.
DOI
[4]
S. Deshmukh, S. Samouhos, L. Glicksman, and L. Norford, Fault detection in commercial building VAV AHU: A case study of an academic building, Energy Build., vol. 201, pp. 163–173, 2019.
[5]
J. Granderson, R. Singla, E. Mayhorn, P. Ehrlich, D. Vrabie, and S. Frank, Characterization and Survey of Automated Fault Detection and Diagnostics Tools. Berkeley, CA, USA: Lawrence Berkeley National Laboratory, 2017.
[6]
J. Wall and Y. Guo, Evaluation of Next-Generation Automated Fault Detection & Diagnostics (FDD) Tools for Commercial Building Energy Efficiency. CSIRO Energy for Low Carbon Living CRC, 2018.
[7]
J. Granderson, G. Lin, R. Singla, E. Mayhorn, P. Ehrlich, D. Vrabie, and S. Frank, Commercial Fault Detection and Diagnostics Tools: What They Offer, How They Differ, and What’s Still Needed. Berkeley, CA, USA: Lawrence Berkeley National Laboratory, 2018.
[8]
W. Kim and S. Katipamula, A review of fault detection and diagnostics methods for building systems, Sci. Technol. Built Environ., vol. 24, no. 1, pp. 3–21, 2018.
[9]
C. Robinson, B. Dilkina, J. Hubbs, W. Zhang, S. Guhathakurta, M. A. Brown, and R. M. Pendyala, Machine learning approaches for estimating commercial building energy consumption, Appl. Energy, vol. 208, pp. 889–904, 2017.
[10]
K. Yan, J. Huang, W. Shen, and Z. Ji, Unsupervised learning for fault detection and diagnosis of air handling units, Energy Build., vol. 210, p. 109689, 2020.
[11]
K. Yan, C. W. Zhong, Z. Ji, and J. Huang, Semi-supervised learning for early detection and diagnosis of various air handling unit faults, Energy Build., vol. 181, pp. 75–83, 2018.
[12]
N. V. Chawla, K. W. Bowyer, L. O. Hall, and W. P. Kegelmeyer, SMOTE: Synthetic minority over-sampling technique, J. Artif. Intell. Res., vol. 16, no. 1, pp. 321–357, 2002.
[13]
K. Yan, A. Chong, and Y. Mo, Generative adversarial network for fault detection diagnosis of chillers, Build. Environ., vol. 172, p. 106698, 2020.
[14]
K. Yan, Chiller fault detection and diagnosis with anomaly detective generative adversarial network, Build. Environ., vol. 201, p. 107982, 2021.
[15]
A. Ebrahimifakhar, D. Yuill, and A. Kabirikopaei, Application of machine learning classification methods in fault detection and diagnosis of rooftop units, in Proc. 18t⁢h Int. Refrigeration and Air Conditioning Conf., West Lafayette, IN, USA, 2021, p. 2137.
[16]
F. Cheng, W. Cai, X. Zhang, H. Liao, C. Cui, Fault detection and diagnosis for air handling unit based on multiscale convolutional neural networks, Energy Build., vol. 236, p. 110795, 2021.
[17]
K. P. Lee, B. H. Wu, and S. L. Peng, Deep-learning-based fault detection and diagnosis of air-handling units, Build. Environ., vol. 157, pp. 24–33, 2019.
[18]
H. Wang, D. Feng, and K. Liu, Fault detection and diagnosis for multiple faults of VAV terminals using self-adaptive model and layered random forest, Build. Environ., vol. 193, p. 107667, 2021.
[19]
R. Chintala, J. Winkler, and X. Jin, Automated fault detection of residential air-conditioning systems using thermostat drive cycles, Energy Build., vol. 236, p. 110691, 2021.
[20]
I. Cohen, F. G. Cozman, N. Sebe, M. C. Cirelo, and T. S. Huang, Semisupervised learning of classifiers: Theory, algorithms, and their application to human-computer interaction, IEEE Trans. Pattern Anal. Mach. Intell., vol. 26, no. 12, pp. 1553–1566, 2004.
[21]
M. S. Mirnaghi and F. Haghighat, Fault detection and diagnosis of large-scale HVAC systems in buildings using data-driven methods: A comprehensive review, Energy Build., 229, p. 110492, 2020.
[22]
C. Fan, X. Liu, P. Xue, and J. Wang, Statistical characterization of semi-supervised neural networks for fault detection and diagnosis of air handling units, Energy Build., vol. 234, p. 110733, 2021.
[23]
B. Li, F. Cheng, X. Zhang, C. Cui, and W. Cai, A novel semi-supervised data-driven method for chiller fault diagnosis with unlabeled data, Appl. Energy., vol. 285, p. 116459, 2021.
[24]
M. Albayati, R. Gorthala, A. Thompson, P. Patil, and A. Hacker, Bringing automated fault detection and diagnostics tools for HVAC&R into the mainstream, J. Eng. Sustain. Build. Cities, vol. 1, no. 3, p. 030902, 2020.
[25]
W. Kim and J. E. Braun, Performance evaluation of a virtual refrigerant charge sensor, Int. J. Refrig., vol. 36, no. 3, pp. 1130–1141, 2013.
[26]
W. Kim and J. E. Braun, Extension of a virtual refrigerant charge sensor, Int. J. Refrig., vol. 55, pp. 224–235, 2015.
[27]
H. Li and J. E. Braun, Development, evaluation, and demonstration of a virtual refrigerant charge sensor, HVAC&R Res., vol. 15, no. 1, pp. 117–136, 2009.
[28]
M. Mehrabi and D. Yuill, Generalized effects of faults on normalized performance variables of air conditioners and heat pumps, Int. J. Refrig., vol. 85, pp. 409–430, 2018.
[29]
M. Mehrabi and D. Yuill, Generalized effects of refrigerant charge on normalized performance variables of air conditioners and heat pumps, Int. J. Refrig., vol. 76, pp. 367–384, 2017.
[30]
D. P. Yuill and J. E. Braun, Evaluating the performance of fault detection and diagnostics protocols applied to air-cooled unitary air-conditioning equipment, HVAC&R Res., vol. 19, no. 7, pp. 882–891, 2013.
[31]
A. Thompson, R. Gorthala, M. Albayati, and P. Pati, RTU-HVAC real-time operating data from unit in field, https://data.mendeley.com/datasets/9h6gpbhj5k/1, 2021.
[32]
L. M. R. Baccarini, V. V. R. E Silva, B. R. de Menezes, and W. M. Caminhas, SVM practical industrial application for mechanical faults diagnostic, Expert Syst. Appl., vol. 38, no. 6, pp. 6980–6984, 2011.
[33]
V. N. Vapnik, The Nature of Statistical Learning Theory, New York, NY, USA: Springer, 2000.
DOI
[34]
P. Arora, , and S. Varshney, Analysis of K-means and K-medoids algorithm for big data, Proced. Comput. Sci., vol. 78, pp. 507–512, 2016.
Publication history
Copyright
Acknowledgements
Rights and permissions

Publication history

Received: 16 June 2022
Accepted: 30 June 2022
Published: 26 January 2023
Issue date: June 2023

Copyright

© The author(s) 2023.

Acknowledgements

This work was supported in part by the US Department of Energy (No. DE-EE0008189) and the National Science Foundation (Nos. 1743418 and 1843025).

Rights and permissions

The articles published in this open access journal are distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/).

Return