Real vs. simulated: Questions on the capability of simulated datasets on building fault detection for energy efficiency from a data-driven perspective

被引:21
作者
Huang, Jiajing [1 ]
Wen, Jin [2 ]
Yoon, Hyunsoo [3 ]
Pradhan, Ojas [2 ]
Wu, Teresa [1 ]
O'Neill, Zheng [4 ]
Candan, Kasim Selcuk [1 ]
机构
[1] Arizona State Univ, Sch Comp & Augmented Intelligence, Tempe, AZ 85281 USA
[2] Drexel Univ, Dept Civil Architectural & Environm Engn, Philadelphia, PA 19104 USA
[3] Yonsei Univ, Dept Ind Engn, Seoul 03722, South Korea
[4] Texas A&M Univ, J Mike Walker 66 Dept Mech Engn, College Stn, TX 77843 USA
基金
美国国家科学基金会;
关键词
Building AFDD; Machine learning; Simulated; Real; Similarity; NEURAL-NETWORK; HVAC SYSTEMS; DIAGNOSIS; PROGNOSTICS; STRATEGY; SENSORS; WAVELET;
D O I
10.1016/j.enbuild.2022.111872
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
Literature on building Automatic Fault Detection and Diagnosis (AFDD) mainly focuses on simulated system data due to high expenses and difficulties of obtaining and analyzing real building data. There is a lack of validation on performances and scalabilities of data-driven AFDD approaches using simulated data and how it compares to that from real building data. In this study, we conduct two sets of experiments to seek answers to this question. We first evaluate data-driven fault detection strategies on real and simulated building data separately. We observe that the fault detection performances are not affected by fault detection strategies, sizes of training data, and the number of cross-validation folds when training and blind test data come from the same data source, namely, simulated or real building data. Next, we conduct a cross-dataset study, that is, develop the model using simulated data and tested on real building data. The results indicate the model trained on simulated data is not generalized to be applied for real building data for fault detection. Kolmogorov-Smirnov Test is conducted to confirm that there exist statistical differences between the simulated and real building data and identify a subset of features with similarities between the two datasets. Using the subset of the feature, cross-dataset experiments show fault detection improvements on most fault cases. We conclude that even if the system produces simulated data with the same fault symptoms from physical analysis perspectives, not all features from simulated datasets may not be beneficial for AFDD but only a subset of features contains valuable information from a machine learning perspective. (c) 2022 Elsevier B.V. All rights reserved.
引用
收藏
页数:10
相关论文
共 43 条
[31]  
Marino DL, 2016, IEEE IND ELEC, P7046, DOI 10.1109/IECON.2016.7793413
[32]   THE KOLMOGOROV-SMIRNOV TEST FOR GOODNESS OF FIT [J].
MASSEY, FJ .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1951, 46 (253) :68-78
[33]   Robust model-based fault diagnosis for air handling units [J].
Mulumba, Timothy ;
Afshari, Afshin ;
Yan, Re ;
Shen, Wen ;
Norford, Leslie K. .
ENERGY AND BUILDINGS, 2015, 86 :698-707
[34]   Towards an adapted PHM approach: Data quality requirements methodology for fault detection applications [J].
Omri, N. ;
Masry, Z. Al ;
Mairot, N. ;
Giampiccolo, S. ;
Zerhouni, N. .
COMPUTERS IN INDUSTRY, 2021, 127
[35]   A review on buildings energy consumption information [J].
Perez-Lombard, Luis ;
Ortiz, Jose ;
Pout, Christine .
ENERGY AND BUILDINGS, 2008, 40 (03) :394-398
[36]  
Roth K.W, 2004, INT REFRIGERATION AI, P665
[37]  
Shawe-Tylor J., 2000, An Introduction to Support Vector Machines and Other Kernel-based Learning Methods, V30, P103
[38]   Development and implementation of automated fault detection and diagnostics for building systems: A review [J].
Shi, Zixiao ;
O'Brien, William .
AUTOMATION IN CONSTRUCTION, 2019, 104 :215-229
[39]   A cautionary note on the use of the Kolmogorov-Smirnov test for normality [J].
Steinskog, Dag J. ;
Tjostheim, Dag B. ;
Kvamsto, Nils G. .
MONTHLY WEATHER REVIEW, 2007, 135 (03) :1151-1157
[40]  
Wen J., 2011, TOOLS EVALUATING FAU