“…Wang et al [21] employed virtual in situ calibration based on Bayesian inference and Markov chain Monte Carlo for photovoltaic thermal heat pump systems. Li et al [22] utilized multiple linear regression to enhance in-situ sensor calibration strategies using Bayesian inference. Chen et al [23] proposed a discrete Bayesian network-based method for diagnosing cross-level faults in HVAC systems.…”
Feature data refer to direct measurements of specific features, while feature residuals represent the deviations between these measurements and their corresponding benchmark values. Both types of information offer unique insights into the system’s behavior. However, conventional diagnostic systems often struggle to effectively integrate and utilize both types of information concurrently. To address this limitation and improve diagnostic performance, a hybrid method based on the Bayesian network (BN) is proposed. This method enables the parallel fusion of feature residuals and feature data within a unified diagnostic model, and a comprehensive framework for developing this hybrid method is also given. In the hybrid BN, the symptom layer consists of residual nodes representing feature residuals and data nodes representing measured feature data. By applying the proposed method to two chillers and comparing it with state-of-the-art existing methods, we demonstrate its effectiveness and superiority. The results highlight that the proposed method not only accommodates the absence of either type of information but also leverages both of them to enhance diagnostic performance. Compared to using a single type of node, the hybrid method achieves a maximum improvement of 24.5% in diagnostic accuracy, with significant enhancements in F-measure observed for refrigerant leakage fault (34.5%) and excessive lubricant fault (32.8%), respectively.
“…Wang et al [21] employed virtual in situ calibration based on Bayesian inference and Markov chain Monte Carlo for photovoltaic thermal heat pump systems. Li et al [22] utilized multiple linear regression to enhance in-situ sensor calibration strategies using Bayesian inference. Chen et al [23] proposed a discrete Bayesian network-based method for diagnosing cross-level faults in HVAC systems.…”
Feature data refer to direct measurements of specific features, while feature residuals represent the deviations between these measurements and their corresponding benchmark values. Both types of information offer unique insights into the system’s behavior. However, conventional diagnostic systems often struggle to effectively integrate and utilize both types of information concurrently. To address this limitation and improve diagnostic performance, a hybrid method based on the Bayesian network (BN) is proposed. This method enables the parallel fusion of feature residuals and feature data within a unified diagnostic model, and a comprehensive framework for developing this hybrid method is also given. In the hybrid BN, the symptom layer consists of residual nodes representing feature residuals and data nodes representing measured feature data. By applying the proposed method to two chillers and comparing it with state-of-the-art existing methods, we demonstrate its effectiveness and superiority. The results highlight that the proposed method not only accommodates the absence of either type of information but also leverages both of them to enhance diagnostic performance. Compared to using a single type of node, the hybrid method achieves a maximum improvement of 24.5% in diagnostic accuracy, with significant enhancements in F-measure observed for refrigerant leakage fault (34.5%) and excessive lubricant fault (32.8%), respectively.
“…To address the problem of virtual sensor errors having a negative effect on the in situ calibration accuracy of the corresponding physical sensors, Koo et al [ 33 ] proposed and adopted a simultaneous in situ calibration method. Li et al [ 34 ] proposed a general-regression-improved Bayesian inference calibration method with a calibration accuracy of over 97%. This method improved the accuracy by 2.8–20% compared with the energy conservation Bayesian inference and principal component analysis methods.…”
A lack of available information on heating, ventilation, and air-conditioning (HVAC) systems can affect the performance of data-driven fault-tolerant control (FTC) models. This study proposed an in situ selective incremental calibration (ISIC) strategy. Faults were introduced into the indoor air (Ttz1) thermostat and supply air temperature (Tsa) and chilled water supply air temperature (Tchws) sensors of a central air-conditioning system. The changes in the system performance after FTC were evaluated. Then, we considered the effects of the data quality, data volume, and variable number on the FTC results. For the Ttz1 thermostat and Tsa sensor, the system energy consumption was reduced by 2.98% and 3.72% with ISIC, respectively, and the predicted percentage dissatisfaction was reduced by 0.67% and 0.63%, respectively. Better FTC results were obtained using ISIC when the Ttz1 thermostat had low noise, a 7-day data volume, or sufficient variables and when the Tsa and Tchws sensors had low noise, a 14-day data volume, or limited variables.
“…The building sector accounts for approximately 30% of total global energy consumption [1][2][3]. Approximately 60% of building energy consumption is used by heating, ventilation and air-conditioning (HVAC) systems [4,5]. Humans spend about 90% of their time working and living indoors which as a result requires higher indoor thermal comfort [6].…”
Due to energy constraints and people’s increasing requirements for indoor thermal comfort, improving energy efficiency while ensuring thermal comfort has become the focus of research in the design and operation of HVAC systems. This study took office rooms with few people occupying them in Wuhan as the research object. The EnergyPlus-Fluent co-simulation method was used to study the impact of 12 forms of air distribution on the thermal environment and air-conditioner energy consumption. The results indicate that 3 m/s supply air velocity and 45° supply air angle are more suitable for the case model in this study. The EnergyPlus-Fluent co-simulation method used in this paper provides a reference for the study of indoor environments in offices with few people occupying them.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.