@article{35506,
  keywords = {HVAC system, Pattern matching, Cross-level fault, Root cause fault diagnosis, Discrete Bayesian Network},
  author = {Yimin Chen and Jin Wen and Ojas Pradhan and James Lo and Teresa Wu},
  title = {Using discrete Bayesian networks for diagnosing and isolating cross-level faults in HVAC systems},
  abstract = {<p>Fault detection and diagnosis (FDD) technologies are critical to ensure satisfactory building performance, such<br />as reducing energy wastes and negative impacts on occupant comfort and productivity. Existing FDD technologies<br />mainly focus on component-level FDD solutions, which could lead to mis-diagnosis of cross-level faults in heating,<br />ventilating, and air-conditioning (HVAC) systems. Cross-level faults are those faults that occur in one component or<br />subsystem, but cause operational abnormalities in other components or subsystems, and result in a building level<br />performance degradation. How to effectively diagnose the root cause of a cross-level fault is the focus of this study.<br />This paper presents a novel discrete Bayesian Network (DisBN)-based method for diagnosing cross-level faults in<br />an HVAC system commonly used in commercial buildings. A two-level DisBN structure model is developed in this<br />study. The parameters used in the DisBN model are obtained either from expert knowledge or through machine-<br />learning strategies from normal system operation data. Meanwhile, the probability parameters are discretized to<br />incorporate the uncertainties associated with typical expert knowledge. Thus, the developed DisBN method<br />addresses the challenges many other BN based FDD methods face, i.e., the lack of fault data for BN parameter<br />training. The developed DisBN represents causal relationships between a fault and its cross-level system impacts<br />(i.e., fault symptoms or fault indicators) by considering how fault impacts propagate across different levels in a<br />HVAC system. A weather and schedule information-based Pattern Matching (WPM) method is employed to<br />automatically create WPM baseline data sets for each incoming real time snapshot data from the building systems.<br />Consequently, BN inference and real-time diagnostics are achieved by comparing incoming snapshot data and the<br />WPM baseline data set. The proposed method is evaluated using experimental fault data collected in a campus<br />building. Fault diagnosis results demonstrate that the WPM-DisBN method is effective at locating the root causes of<br />cross-level faults in an HVAC system.</p>},
  year = {2022},
  journal = {Applied Energy},
  volume = {327},
  month = {10/2022},
  doi = {https://doi.org/10.1016/j.apenergy.2022.120050},
  language = {eng},
}