AI Chat Paper
Note: Please note that the following content is generated by AMiner AI. SciOpen does not take any responsibility related to this content.
{{lang === 'zh_CN' ? '文章概述' : 'Summary'}}
{{lang === 'en_US' ? '中' : 'Eng'}}
Chat more with AI
Powered by AMiner. Clicking this button will take you away from SciOpen.
Home Building Simulation Article
Article Link
Cite
EndNote(RIS) BibTeX
Collect
Submit Manuscript
Show Outline
Outline
Show full outline
Hide outline
Outline
Show full outline
Hide outline
Research Article

An interpretable graph convolutional neural network based fault diagnosis method for building energy systems

Guannan Li1,2,3,4, Zhanpeng Yao1, Liang Chen1, Tao Li1,5( ), Chengliang Xu1
School of Urban Construction, Wuhan University of Science and Technology, Wuhan 430065, China
Anhui Province Key Laboratory of Intelligent Building and Building Energy-saving, Anhui Jianzhu University, Hefei 230601, China
State Key Laboratory of Green Building in Western China, Xi'an University of Architecture & Technology, Xi'an 710055, China
Key Laboratory of Low-grade Energy Utilization Technologies and Systems (Chongqing University), Ministry of Education of China, Chongqing University, Chongqing 400044, China
Hubei Provincial Engineering Research Centre of Urban Regeneration, Wuhan University of Science and Technology, Wuhan 430081, China
Show Author Information

Abstract

Due to the fast-modeling speed and high accuracy, deep learning has attracted great interest in the field of fault diagnosis in building energy systems in recent years. However, the black-box nature makes deep learning models generally difficult to interpret. In order to compensate for the poor interpretability of deep learning models, this study proposed a fault diagnosis method based on interpretable graph neural network (GNN) suitable for building energy systems. The method is developed by following three main steps: (1) selecting NC-GNN as a fault diagnosis model for building energy systems and proposing a graph generation method applicable to the model, (2) developing an interpretation method based on InputXGradient for the NC-GNN, which is capable of outputting the importance of the node features and automatically locating the fault related features, (3) visualizing the results of model interpretation and validating by matching with expert knowledge and maintenance experience. Validation was performed using the public ASHRAE RP-1043 chiller fault data. The diagnosis results show that the proposed method has a diagnosis accuracy of over 96%. The interpretation results show that the method is capable of explaining the decision-making process of the model by identifying fault-discriminative features. For almost all seven faults, their fault-discriminative features were correctly identified.

Keywords

fault diagnosisfeaturegraph neural networkbuilding energy systeminterpretationInputXGradient

References

 

Ahmad MW, Mourshed M, Mundow D, et al. (2016a). Building energy metering and environmental monitoring–A state-of-the-art review and directions for future research. Energy and Buildings, 120: 85–102.
Crossref Google Scholar
 

Ahmad MW, Mourshed M, Yuce B, et al. (2016b). Computational intelligence techniques for HVAC systems: A review. Building Simulation, 9: 359–398.
Crossref Google Scholar
 

Apicella A, Isgrò F, Pollastro A, et al. (2023). Adaptive filters in Graph Convolutional Neural Networks. Pattern Recognition, 144: 109867.
Crossref Google Scholar
 

Beghi A, Brignoli R, Cecchinato L, et al. (2016). Data-driven Fault Detection and Diagnosis for HVAC water chillers. Control Engineering Practice, 53: 79–91.
Crossref Google Scholar
 

Betkier I, Oszczypała M, Pobożniak J, et al. (2023). PocketFinderGNN: A manufacturing feature recognition software based on Graph Neural Networks (GNNs) using PyTorch Geometric and NetworkX. SoftwareX, 23: 101466.
Crossref Google Scholar
 

Chen K, Chen S, Zhu X, et al. (2023). Interpretable mechanism mining enhanced deep learning for fault diagnosis of heating, ventilation and air conditioning systems. Building and Environment, 237: 110328.
Crossref Google Scholar
 
Comstock MC J. E. Braun JE, Bernhard R (1999). Development of analysis tools for the evaluation of fault detection and diagnostics in chillers. Purdue University.
 

Costa A, Keane MM, Torrens JI, et al. (2013). Building operation and energy performance: Monitoring, analysis and optimisation toolkit. Applied Energy, 101: 310–316.
Crossref Google Scholar
 

Daigavane A, Ravindran B, Aggarwal G (2021). Understanding convolutions on graphs. Distill, https://doi.org/10.23915/distill.00032.
Crossref Google Scholar
 

Du Z, Chen S, Li P, et al. (2023). Knowledge-extracted deep learning diagnosis and its cloud-based management for multiple faults of chiller. Building and Environment, 235: 110228.
Crossref Google Scholar
 

Ebrahimifakhar A, Kabirikopaei A, Yuill D (2020). Data-driven fault detection and diagnosis for packaged rooftop units using statistical machine learning classification methods. Energy and Buildings, 225: 110318.
Crossref 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.
Crossref Google Scholar
 

Fan C, Xiao F, Song M, et al. (2019). A graph mining-based methodology for discovering and visualizing high-level knowledge for building energy management. Applied Energy, 251: 113395.
Crossref 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.
Crossref Google Scholar
 
Gilmer J, Schoenholz SS, Riley PF, et al. (2017). Neural message passing for Quantum chemistry. In: Proceedings of the 34th International Conference on Machine Learning, Sydney, Australia.
 

Han H, Cui X, Fan Y, et al. (2019). Least squares support vector machine (LS-SVM)-based chiller fault diagnosis using fault indicative features. Applied Thermal Engineering, 154: 540–547.
Crossref Google Scholar
 

Han H, Gu B, Hong Y, KangJ (2011a). Automated FDD of multiple-simultaneous faults (MSF) and the application to building chillers. Energy and Buildings, 43: 2524–2532.
Crossref Google Scholar
 

Han H, Gu B, Wang T, Li ZR (2011b). Important sensors for chiller fault detection and diagnosis (FDD) from the perspective of feature selection and machine learning. International Journal of Refrigeration, 34: 586–599.
Crossref Google Scholar
 

Han Y, Li Q, Wang C, et al. (2022). A novel knowledge enhanced graph neural networks for fault diagnosis with application to blast furnace process safety. Process Safety and Environmental Protection, 166: 143–157.
Crossref Google Scholar
 

Hong Y, Yoon S, Kim Y-S, et al. (2021). System-level virtual sensing method in building energy systems using autoencoder: Under the limited sensors and operational datasets. Applied Energy, 301: 117458.
Crossref Google Scholar
 

Hu Y, Chen H, Xie J, et al. (2012). Chiller sensor fault detection using a self-Adaptive Principal Component Analysis method. Energy and Buildings, 54: 252–258.
Crossref Google Scholar
 

Jia Y, Wang J, Reza Hosseini M, et al. (2023). Temporal graph attention network for building thermal load prediction. Energy and Buildings, https://doi.org/10.1016/j.enbuild.2023.113507.
Crossref Google Scholar
 

Jiang W, Luo J (2022). Graph neural network for traffic forecasting: A survey. Expert Systems with Applications, 207: 117921.
Crossref Google Scholar
 

Katipamula S, Brambley MR (2005a). Methods for fault detection, diagnostics, and prognostics for building systems—A Review, Part Ⅰ. HVAC&R Research, 11: 3–25.
Crossref Google Scholar
 

Katipamula S, Brambley MR (2005b). Methods for fault detection, diagnostics, and prognostics for building systems—A review, part Ⅱ. HVAC&R Research, 11: 169–187.
Crossref Google Scholar
 

Keramatfar A, Rafiee M, Amirkhani H (2022). Graph Neural Networks: A bibliometrics overview. Machine Learning with Applications, 10: 100401.
Crossref Google Scholar
 

Kim M, Yoon SH, Domanski PA, et al. (2008). Design of a steady-state detector for fault detection and diagnosis of a residential air conditioner. International Journal of Refrigeration, 31: 790–799.
Crossref Google Scholar
 
Kindermans PJ, Schütt K, Müller K-R, et al. (2016). Investigating the influence of noise and distractors on the interpretation of neural networks. arXiv: 1611.07270.
 
Kipf TN, Welling M (2016). Semi-supervised classification with graph convolutional networks. arXiv: 1609.02907.
 

Lee T, Yoon S, Won K (2022). Delta-T-based operational signatures for operation pattern and fault diagnosis of building energy systems. Energy and Buildings, 257: 111769.
Crossref Google Scholar
 

Li S, Wen J (2014). A model-based fault detection and diagnostic methodology based on PCA method and wavelet transform. Energy and Buildings, 68: 63–71.
Crossref Google Scholar
 

Li G, Hu Y, Chen H, et al. (2016). An improved fault detection method for incipient centrifugal chiller faults using the PCA-R-SVDD algorithm. Energy and Buildings, 116: 104–113.
Crossref Google Scholar
 

Li G, Hu Y, Chen H, et al. (2017). Data partitioning and association mining for identifying VRF energy consumption patterns under various part loads and refrigerant charge conditions. Applied Energy, 185: 846–861.
Crossref Google Scholar
 

Li Y, O’Neill Z (2018). A critical review of fault modeling of HVAC systems in buildings. Building Simulation, 11: 953–975.
Crossref Google Scholar
 

Li D, Zhou Y, Hu G, et al. (2019). Identifying unseen faults for smart buildings by incorporating expert knowledge with data. IEEE Transactions on Automation Science and Engineering, 16: 1412–1425.
Crossref Google Scholar
 

Li B, Cheng F, Zhang X, et al. (2021a). A novel semi-supervised data-driven method for chiller fault diagnosis with unlabeled data. Applied Energy, 285: 116459.
Crossref Google Scholar
 

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

Li G, Li F, Xu C, et al. (2022a). A spatial-temporal layer-wise relevance propagation method for improving interpretability and prediction accuracy of LSTM building energy prediction. Energy and Buildings, 271: 112317.
Crossref Google Scholar
 

Li T, Zhou Y, Zhao Y, et al. (2022b). A hierarchical object oriented Bayesian network-based fault diagnosis method for building energy systems. Applied Energy, 306: 118088.
Crossref Google Scholar
 

Li T, Zhou Z, Li S, et al. (2022c). The emerging graph neural networks for intelligent fault diagnostics and prognostics: A guideline and a benchmark study. Mechanical Systems and Signal Processing, 168: 108653.
Crossref Google Scholar
 

Li G, Chen L, Fan C, et al. (2023a). Interpretation and explanation of convolutional neural network-based fault diagnosis model at the feature-level for building energy systems. Energy and Buildings, 295: 113326.
Crossref 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.
Crossref Google Scholar
 

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

Li G, Chen L, Fan C, et al. (2024). Improved convolutional neural network chiller early fault diagnosis by gradient-based feature-level model interpretation and feature learning. Applied Thermal Engineering, 236: 121549.
Crossref Google Scholar
 

Liang X, Zhu X, Chen S, et al. (2023). Physics-constrained cooperative learning-based reference models for smart management of chillers considering extrapolation scenarios. Applied Energy, 349: 121642.
Crossref Google Scholar
 

Liu J, Li G, Chen H, et al. (2017). A robust online refrigerant charge fault diagnosis strategy for VRF systems based on virtual sensor technique and PCA-EWMA method. Applied Thermal Engineering, 119: 233–243.
Crossref Google Scholar
 

Liu J, Shi D, Li G, et al. (2020). Data-driven and association rule mining-based fault diagnosis and action mechanism analysis for building chillers. Energy and Buildings, 216: 109957.
Crossref Google Scholar
 

Long S, Marjanovic O, Parisio A (2019). Generalised control-oriented modelling framework for multi-energy systems. Applied Energy, 235: 320–331.
Crossref Google Scholar
 

Lu J, Zhang C, Li J, et al. (2022). Graph convolutional networks-based method for estimating design loads of complex buildings in the preliminary design stage. Applied Energy, 322: 119478.
Crossref Google Scholar
 

Mirnaghi MS, Haghighat F (2020). Fault detection and diagnosis of large-scale HVAC systems in buildings using data-driven methods: A comprehensive review. Energy and Buildings, 229: 110492.
Crossref Google Scholar
 

Reddy TA (2007). Application of a generic evaluation methodology to assess four different chiller FDD methods (RP-1275). HVAC&R Research, 13: 711–729.
Crossref Google Scholar
 

Singh V, Mathur J, Bhatia A (2022). A comprehensive review: Fault detection, diagnostics, prognostics, and fault modeling in HVAC systems. International Journal of Refrigeration, 144: 283–295.
Crossref Google Scholar
 

Sun K, Hong T, Kim J, et al. (2022). Application and evaluation of a pattern-based building energy model calibration method using public building datasets. Building Simulation, 15: 1385–1400.
Crossref Google Scholar
 

Tang R, Fan C, Zeng F, et al. (2022). Data-driven model predictive control for power demand management and fast demand response of commercial buildings using support vector regression. Building Simulation, 15: 317–331.
Crossref Google Scholar
 

Tian Y, Wang J, Qi Z, et al. (2023). Calibration method for sensor drifting bias in data center cooling system using Bayesian Inference coupling with Autoencoder. Journal of Building Engineering, 67: 105961.
Crossref Google Scholar
 

Toub M, Reddy CR, Robinett RD Ⅲ, et al. (2021). Integration and optimal control of MicroCSP with building HVAC systems: Review and future directions. Energies, 14: 730.
Crossref Google Scholar
 

Tran DAT, Chen Y, Ao HL, et al. (2016a). An enhanced chiller FDD strategy based on the combination of the LSSVR-DE model and EWMA control charts. International Journal of Refrigeration, 72: 81–96.
Crossref Google Scholar
 

Tran DAT, Chen Y, et al. (2016b). Comparative investigations on reference models for fault detection and diagnosis in centrifugal chiller systems. Energy and Buildings, 133: 246–256.
Crossref Google Scholar
 

Vishwanathan SVN, Schraudolph SS, Kondor R, et al. (2010). Graph kernels. Journal of Machine Learning Research, 11: 1201–1242.
Google Scholar
 

Wang P, Yoon S, Wang J, et al. (2019). Automated reviving calibration strategy for virtual in situ sensor calibration in building energy systems: Sensitivity coefficient optimization. Energy and Buildings, 198: 291–304.
Crossref Google Scholar
 

Wen H, Guo W, Li X (2023). A novel deep clustering network using multi-representation autoencoder and adversarial learning for large cross-domain fault diagnosis of rolling bearings. Expert Systems with Applications, 225: 120066.
Crossref Google Scholar
 

Wu Z, Pan S, Chen F, et al. (2021). A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems, 32: 4–24.
Crossref Google Scholar
 

Xu X, Xiao F, Wang S (2008). Enhanced chiller sensor fault detection, diagnosis and estimation using wavelet analysis and principal component analysis methods. Applied Thermal Engineering, 28: 226–237.
Crossref Google Scholar
 

Yan R, Ma Z, Zhao Y, et al. (2016). A decision tree based data-driven diagnostic strategy for air handling units. Energy and Buildings, 133: 37–45.
Crossref Google Scholar
 

Yan K, Huang J, Shen W, et al. (2020). Unsupervised learning for fault detection and diagnosis of air handling units. Energy and Buildings, 210: 109689.
Crossref Google Scholar
 

Yang Z, Liu Z, Zhou J, et al. (2023). A graph neural network (GNN) method for assigning gas calorific values to natural gas pipeline networks. Energy, 278: 127875.
Crossref Google Scholar
 
Ying R, Bourgeois D, You J, et al. (2019). GNNExplainer: Generating explanations for graph neural networks. arXiv: 1903.03894.
 
Yuan H, Yu H, Gui S, et al. (2022). Explainability in graph neural networks: A taxonomic survey. arXiv: 2012.15445v3.
Crossref
 

Zhang R, Hong T (2017). Modeling of HVAC operational faults in building performance simulation. Applied Energy, 202: 178–188.
Crossref Google Scholar
 

Zhang D, Stewart E, Entezami M, et al. (2020). Intelligent acoustic-based fault diagnosis of roller bearings using a deep graph convolutional network. Measurement, 156: 107585.
Crossref Google Scholar
 

Zhang Y, Yu J (2022). Pruning graph convolutional network-based feature learning for fault diagnosis of industrial processes. Journal of Process Control, 113: 101–113.
Crossref Google Scholar
 

Zhang C, Tian X, Zhao Y, et al. (2022a). Causal discovery-based external attention in neural networks for accurate and reliable fault detection and diagnosis of building energy systems. Building and Environment, 222: 109357.
Crossref Google Scholar
 

Zhang H, Li C, Wei Q, et al. (2022b). Fault detection and diagnosis of the air handling unit via combining the feature sparse representation based dynamic SFA and the LSTM network. Energy and Buildings, 269: 112241.
Crossref Google Scholar
 

Zhang J, Xiao F, Li A, et al. (2023). Graph neural network-based spatio-temporal indoor environment prediction and optimal control for central air-conditioning systems. Building and Environment, 242: 110600.
Crossref Google Scholar
 

Zhao Y, Wang S, Xiao F, et al. (2013). A simplified physical model-based fault detection and diagnosis strategy and its customized tool for centrifugal chillers. HVAC&R Research, 19: 283–294.
Crossref Google Scholar
 

Zhao Y, Li T, Fan C, et al. (2019). A proactive fault detection and diagnosis method for variable-air-volume terminals in building air conditioning systems. Energy and Buildings, 183: 527–537.
Crossref Google Scholar
 

Zhou Q, Wang S, Xiao F (2009). A novel strategy for the fault detection and diagnosis of centrifugal chiller systems. HVAC&R Research, 15: 57–75.
Crossref Google Scholar
 

Zhou J, Cui G, Hu S, et al. (2020). Graph neural networks: a review of methods and applications. AI Open, 1: 57–81.
Crossref Google Scholar
 

Zhou Z, Chen H, Li G, et al. (2021). Data-driven fault diagnosis for residential variable refrigerant flow system on imbalanced data environments. International Journal of Refrigeration, 125: 34–43.
Crossref Google Scholar
Building Simulation
Volume 17 Issue 7,
July 2024
Pages 1113-1136
DOI: 10.1007/s12273-024-1125-6
Cite this article:
Li G, Yao Z, Chen L, et al. An interpretable graph convolutional neural network based fault diagnosis method for building energy systems. Building Simulation, 2024, 17(7): 1113-1136. https://doi.org/10.1007/s12273-024-1125-6

99

Views

1

Crossref

1

Web of Science

1

Scopus

0

CSCD

Altmetrics
Received: 27 November 2023
Revised: 01 March 2024
Accepted: 18 March 2024
Published: 20 June 2024
© Tsinghua University Press 2024
Return