Search

Data-driven simulation models can be used for optimal control and decision making in building systems only if they are developed accurately and reliably. An unsolved issue pertains to the data collected from BEMS that generally contain errors owing to sensors, malfunctioning systems, and other unknown reasons. These anomalies generally occur in a multidimensional space and cannot be easily detected. Unfortunately, these anomalies can prevent the simulation model from being accurate, reliable, and scalable. This paper presents an automatic anomaly detection method using a support vector data description (SVDD). The data obtained from a compression chiller in a real office building were classified into raw and filtered datasets based on SVDD. Two artificial neural network (ANN) models were developed (ANNraw and ANNSVDD). Based on the comparison between the two ANNs, the ANNSVDD is found to be better than ANNraw in terms of the model reliability and reproducibility. It is also interesting that the model accuracy differences between the two ANN models are marginal. However, the accuracy of the two ANN models can be improved as long as they are tested against filtered data by SVDD. The improvements in accuracy signify the importance of the elimination of these anomalies. [ABSTRACT FROM AUTHOR]

A deep Q-network (DQN) was applied for model-free optimal control balancing between different HVAC systems. The DQN was coupled to a reference office building: an EnergyPlus simulation model provided by the U.S. Department of Energy. The building was air-conditioned with four air-handling units (AHUs), two electric chillers, a cooling tower, and two pumps. EnergyPlus simulation results for eleven days (July 1–11) and three subsequent days (July 12–14) were used to improve the DQN policy and test the optimal control. The optimization goal was to minimize the building's energy use while maintaining the indoor CO2 concentration below 1,000 ppm. It was revealed that the DQN—a reinforcement learning method—can improve its control policy based on prior actions, states, and rewards. The DQN lowered the total energy usage by 15.7% in comparison with the baseline operation while maintaining the indoor CO2 concentration below 1,000 ppm. Compared to model predictive control, the DQN does not require a simulation model, or a predetermined prediction horizon, thus delivering model-free optimal control. Furthermore, it was demonstrated that the DQN can find balanced control actions between different energy consumers in the building, such as chillers, pumps, and AHUs. [ABSTRACT FROM AUTHOR]

Window adjustment is one of the most common ways employed by occupants to control the indoor environment, and window adjustment behavior (WAB) is known to be a crucial factor for predicting building energy consumption. In recent decades, many studies have attempted to develop a reliable WAB model with the following three issues. First, most current studies have treated WAB models in an "average occupant" fashion. This average WAB model is prone to ignoring the variability of individual preferences. The second issue is the analysis of the impact of these factors, which still considers how occupants respond to environmental and non-environmental factors. Finally, regarding the prediction models, there are diverse modeling approaches such as probabilistic and machine learning approaches. Each modeling method has advantages and disadvantages in terms of adaptability, complexity, and explainability. This study focuses on explainable artificial intelligence (XAI) to presents quantitative analysis on the aforementioned three issues. For this purpose, occupant data (occupant presence and window state) and indoor/outdoor environmental data (temperature, humidity, illuminance, CO 2 concentration, PM 2.5 concentration, and time of day) were collected from 12 households for one year using multiple sensors installed in each household. Three XAI approaches, namely, logistic regression, XGBoost, and Shapley additive explanations (SHAP), were introduced to analyze the interaction between environmental and non-environmental factors and occupant behavioral patterns in terms of feature influence. The results showed that (1) occupant personal preferences on WAB significantly vary from household to household; (2) occupant personal preferences on WAB cannot be defined only with a single environmental parameter; (3) there are unmeasurable unknown factors such as psychological and social factors; and (4) similar to logistic regression models, the current complex black-box models can also be described by applying XAI techniques in terms of feature influence. [ABSTRACT FROM AUTHOR]

Building simulation tools have been widely used for performance assessment. However, many studies [1] have reported that a performance gap exists between the reality and simulation output, mainly caused by unknown simulation inputs. Therefore, model calibration needs to be introduced. Calibration attempts can fail for the following reasons: coarse initial simulation model, long sampling time, uncertainty in the model, and sensor errors. The aim of this paper is to address the abovementioned issues. For this study, an existing office building was selected and two calibration approaches were presented: deterministic vs. stochastic. For stochastic calibration, a Gaussian Process Emulator (GPE) was introduced as a surrogate of the EnergyPlus model. The stochastically calibrated model performs better than the deterministically calibrated model. It is concluded in the paper that (1) the calibration quality is influenced by the degree of the details of the initial model, (2) the accumulated measured data under a sampling time of up to one day (e.g. gas energy consumption) might be unsuitable for calibration work due to the lack of ‘time-series trend’, and (3) the calibration quality is also influenced by sensor errors and further calibration needs to take these into account. [ABSTRACT FROM AUTHOR]

Although it is widely acknowledged that occupants play a critical role in building energy consumption, the characteristics of occupants are not well-represented in building simulation. Many statistic and data-mining technologies have been applied to develop a reliable occupant model. In contrast, rather than attempting to develop the occupant model, this study aims to investigate the correlation between the occupant behavior and energy consumption based on a series of experiments. First, this paper discusses the randomness of the occupants’ presence and behavior. The degree of randomness was verified using a Normalized Cumulative Periodogram (NCP) based on a random walk hypothesis. In addition, the correlation between occupant and energy consumption was investigated using the wavelet coherence. This study shows that, while the occupants’ presence has a randomness, it is not strongly correlated to the energy consumption. The occupants’ active action to control a heating/cooling system (turn on/off) is correlated to the energy consumption. In contrast to the occupant's presence, the occupants’ active action does not follow the random walk, and it has no particular frequency. This means that it is difficult to predict the control action of occupants with a specific time interval. [ABSTRACT FROM AUTHOR]

Abstract: This paper presents the follow-up in the development of occupant responsive optimal control for double-skin systems that was published previously in Park et al. [1]. The presented occupant responsive optimal control optimizes in real-time the performance of the facade in terms of energy, visual comfort (average daylight interior illuminance, uniformity, average window luminance, and outward visibility) and thermal comfort (Predicted Percentage Dissatisfied, PPD). In the aforementioned approach, the double-skin façade system was viewed as an ‘isolated’ system and hence treated as a local control problem, i.e., based purely on information about the state of the façade and its immediate environment. The paper extends the local control problem to an integrated control problem in which room environmental control and façade control are dealt with simultaneously. It was found that the local control leads to sub-optimality, albeit of moderate proportions. [Copyright &y& Elsevier]

Abstract: The CR-39 [diethylene glycol bis-(allylcarbonate)] neuron detection method was used to measure the dose of neutrons generated in X-ray (photon) therapy conducted in a linear accelerator, and to use high-energy photons as part of the clinical applications to examine the problems associated with the dose for patients caused by the generation of neutrons from high-energy photons used for cancer therapy. According to the experimental results, 0.35mSv, 0.65mSv 1.82mSv of fast neutrons on average were generated from 1Gy, 2Gy and 5Gy of photon irradiation, respectively, whereas 0.26mSv, 0.56mSv and 1.23mSv of thermal neutrons were generated. Both fast neutrons and thermal neutrons produced an increase in the dose of neutrons generated with increasing irradiation dose. With in regard to the dose generated within and around the irradiation area of the photon rays, it was confirmed that more neutrons were generated within the irradiation area. A wedge filer was used to measure the generation of neutrons. According to the measurement results, the generation of neutrons increased when a wedge filter was used. When the SRS cone that required a high dose was used, more neutrons were detected than those in the previous experiment. When fast neutrons were used, 2.85mSv neutrons on average were generated from 5Gy of photon irradiation. When thermal neutrons were used, 1.37mSv neutrons on average were generated from 5Gy of photon irradiation. Overall, approximately 1.6 times and 1.12 times more fast and thermal neutrons, respectively, were generated than in the case of a general treatment with 5Gy of photon irradiation. [Copyright &y& Elsevier]

Blind systems have been introduced to provide visual and thermal comfort, as well as to reduce energy use in buildings. A wide variety of such systems exist in terms of thermal and optical properties, location and physical configuration. The current problem with blinds is that their operation is not based on the dynamics of the room but on the static or manual control operated by occupants, although many studies have recognized that dynamic control can far outperform static control. One reason for the lack of dynamic control is that it is not easy to combine the room dynamics with any possible optimization algorithm. Hence, in this study, a whole building simulation tool, EnergyPlus, was integrated with the MATLAB optimization toolbox to solve for optimal control of blind systems. This paper addresses the difference between static versus optimal control of interior and exterior blind systems in office buildings. [ABSTRACT FROM AUTHOR]

A study of the surface energetics of the thinnest substratefree liquid films, fluid molecular monolayer and multilayer smectic liquid crystal films suspended in air, is reported. In films having monolayer and multilayer domains, the monolayer areas contract, contrary to predictions from the van der Waals disjoining pressure of thin uniform slabs. This discrepancy is accounted for by modeling the environmental asymmetry of the surface layers in multilayer films, leading to the possibility that preferential end-for-end polar ordering of the rod shaped molecules can reduce the surface energy of multilayers relative to that of the monolayer, which is inherently symmetric. [ABSTRACT FROM AUTHOR]

Abstract: Currently, several whole-building simulation tools (e.g., esp-r, EnergyPlus, TRNSYS, TAS, IES VE, IDA ICE, VA114, BSim, etc.) are used to assess the energy performance of double-skin façade (DSF) buildings. The aforementioned tools are well suited to assess energy performance of conventional building systems or whole buildings; however, it is questionable whether such tools can accurately describe the transient heat and mass transfer phenomena that occur in the complex three-dimensional geometry of DSFs. This paper describes an empirical validation of the EnergyPlus simulation tool for performance simulation of a DSF. A series of experiments were conducted for cavity airflow and thermal behavior of the DSF and then compared with simulation outputs. In this paper, it is shown that there are significant differences in both thermal and airflow behavior of DSFs between the measurements and simulation predictions by EnergyPlus. This study investigates three cases causing the differences and elucidates what should be considered when modeling DSFs using EnergyPlus. [Copyright &y& Elsevier]

While facility design is increasingly playing a role addressing strategic organizational objectives, issues pertaining to facility maintenance have typically been left out of the decision-making process. Reasons include the traditional disconnection between facility design and facility maintenance, mainly originating from the lack of a modality to meaningfully represent facility maintenance information during design decision making. This paper discusses two sets of facility maintenance indicators that have the potential to bridge this traditional divide. The objective of this paper is threefold: (1) to illustrate how maintenance performance indicators can support higher-level decision making; (2) to explain how the methodology could be mapped across building sectors; and (3) to show how facility maintenance could play a crucial role in informing strategic decision making. Two types of indicators are introduced based on: (1) normative models in biophysics and physiology and (2) empiricist models of environment-behavior studies. Using examples from hospital, courthouse, and office building design, the paper articulates the manner in which facility performance indicators could be developed and used to support organizational strategic decision making. The paper demonstrates that facility maintenance indicators could be developed for all types of buildings and could be meaningfully represented for consideration during strategic decision making. Moreover, using an example of a healthcare setting, the paper emphasizes how facility maintenance strategies have an impact on higher-level organizational objectives, and vice versa, thereby underscoring the importance to consider maintenance performance during strategic decision making. The paper shows how the two sets of key performance indicators—the hard and the soft—are designed to address different scales of decision making, thereby allowing facility maintenance performance to be considered at all phases of a procurement cycle. [ABSTRACT FROM AUTHOR]

Recently, the issue of parameter identifiability has been highlighted in Bayesian inference of building energy models. Parameter identifiability addresses the correspondence between the observed data and the model parameters, and this analysis verifies whether the model parameters can be estimated using observed data. The authors propose a model selection process for Bayesian inference involving the unidentifiable parameters based on the model evidence. The model evidence is a component of the Bayes theorem, which corresponds to the observation probability (i.e., the goodness of fit) of the observed data for a given model. In this study, nested sampling was used to estimate the model evidence. A case study with the reference office building developed by the US Department of Energy shows the model selection process that uses the model evidence as the evaluation index for the unidentifiable parameter. For comparison with the existing approach of model evaluation, the authors present the results of a comparative analysis between the proposed process and that based on model prediction error (CV(RMSE)). As a result, it was observed that the higher the model evidence in the hypothesis for the unidentifiable parameters, the more the posteriors tend to be similar to the true values. In addition, no significant relationship was observed between the model prediction error and the accuracy of the posterior inference. This indicates that the model evidence can be used as an objective evaluation index for selecting the optimal hypothesis of unidentifiable parameters in the Bayesian inference of building energy models. [ABSTRACT FROM AUTHOR]

Abstract: This paper treats rational expressions of building performance in order to better support dialogues between stakeholders in the design process. These expressions are based on the notion of objectively quantifiable performance measures, which are introduced through a set of “performance indicators”. The indicators can be used to quantify expectations and fulfillments in structured dialogues between different stakeholders. Two types of indicators are introduced based on: (1) normative models in biophysics and physiology; and (2) empiricist models of Environment-Behavior studies. The treatment is positioned to support rational decision-making during different stages of building delivery and use. The focus is specifically on the fulfillment of client expectations during design evolution. [Copyright &y& Elsevier]

A nematic liquid crystal in a slot electrohydrodynamic convection geometry exhibits, with increasing drive, enhanced noisy high-dimensional flow fluctuations and weakened low-dimensional ordered flow, reaching a turbulent state with only short range velocity correlations in space and time for sufficiently high voltage. Upon additional voltage increase this disorganized turbulent state develops spatial order with extended temporal correlations, in the form of a one-dimensional periodic array of turbulent jet-like plumes along the slot. This array becomes more robust with reduced disorder and faster fluctuations as the drive is further increased. Remarkably this ordering depends sensitively on liquid crystal alignment conditions on the surface of the slot cell. [ABSTRACT FROM AUTHOR]

The paper describes the calibration of a simulation model of double-skin façade systems with controlled rotating louvers and ventilation openings. The approach is based on a parameter estimation technique and in situ monitoring of a full-scale element mounted on the south facing façade of an existing building. Contrary to similar work that attempts to derive the behavior from detailed modeling of the physical transport phenomena, the new approach is based on a postulated “minimalistic” lumped model, which is calibrated on in-situ measurements. It is found that the calibrated model is surprisingly accurate and ideally suited for use in the ensuing optimal control and performance studies. [Copyright &y& Elsevier]

Case-based reasoning (CBR) is a methodology for problem solving and decision-making in complex and changing business environments. Many CBR algorithms are derivatives of the k-nearest neighbor (k-NN) method, which has a similarity function to generate classification from stored cases. Several studies have shown that k-NN performance is highly sensitive to the definition of its similarity function. Many k-NN methods have been proposed to reduce this sensitivity by using various distance functions with feature weights.This paper proposes an analogical reasoning structure for feature weighting using a new framework called the analytic hierarchy process (AHP)-weighted k-NN algorithm. The paper also introduces AHP methodology for assigning relative importance in case indexing and retrieving. The AHP model is a methodology effective in obtaining domain knowledge from numerous experts and representing knowledge-guided indexing. The proposed AHP weighted k-NN algorithm has been shown to achieve classification accuracy higher than the pure k-NN algorithm. This approach is applied to bankruptcy prediction involves the examination of several criteria, both quantitative (financial ratios) and qualitative (non-financial variables). [Copyright &y& Elsevier]

Parameter identifiability is the concept of whether uncertain parameters can be correctly estimated from the observed data. The main cause of parameter unidentifiability in Bayesian inference is known as 'overparameterization'. In this study, the likelihood confidence interval (CI) and the likelihood confidence region (CR) were introduced for quantifying the parameter identifiability. The likelihood CI and CR can be regarded as the parameter range (one-dimensional) and parameter space (two-dimensional or higher) that can identify parameter values, respectively. For this purpose, an EnergyPlus reference office building provided by the US DOE was used in this study. Four estimation parameters in the EnergyPlus model were analyzed using the likelihood CI and CR. It was found that the closer the likelihood CI of a parameter is to the prior's parameter range, the more unidentifiable the parameter. In addition, a biplot analysis was conducted to examine a correlation between two parameters. The more correlated a parameter is with others, the more unidentifiable the parameter. It is suggested that the visual assessment of likelihood CIs and CRs can help in investigating whether Bayesian inference results can be accurately obtained. [ABSTRACT FROM AUTHOR]

The occupant behavior (OB) is known to be a significant factor that causes the performance gap. Many studies have developed OB models to address this issue, including lighting control. So far, occupant lighting on/off control (OLC) has been expressed in terms of indoor illuminance level and/or time or day, but unfortunately the predictability and periodicity of OLC have not been investigated. To address this, we collected OLC data from 29 households in a living room and main bedroom in residential buildings over a one-year period. We analyzed the predictability of OLC using the Normalized Cumulative Periodogram (NCP) and Bartlett's test for different measurement periods (7, 14, 28, and 90 days). For OLC-predictable households, we investigated the periodicity of OLC. Our findings indicate that OLC is predictable in 27 households' living rooms and 24 households' main bedrooms for 28 days of measurement period. Furthermore, the periodicity of OLC varies depending on the household and rooms. [ABSTRACT FROM AUTHOR]

Occupant behavior is regarded as one of the major factors contributing to the discrepancy between simulation prediction and real energy use. Over the past several decades, occupants have been represented as fixed profiles of occupant presence in building energy simulation tools. Recently, stochastic models have been introduced to account for dynamic occupant presence. This stochastic approach is based on the premise that occupant presence can be described by empirical and probabilistic transition rules, e.g. Markov Chain. This paper presents evidence that occupant presence in some rooms and buildings follows a “random walk” pattern. In other words, occupant presence in certain types of buildings cannot be predicted stochastically. In this study, occupants’ presence in two laboratories and three reading rooms at two universities was monitored. The hypothesis of the random walk pattern was tested using the Normalized Cumulative Periodogram (NCP) method. Based on a series of six experiments, it is shown that each occupant’s presence in the five locations follows a random walk pattern. Three different occupant models (fixed ASHRAE model, Markov Chain model, and Random Walk model) were applied in EnergyPlus simulation runs. The adjusted R 2 for three experiments between the fixed AHSRAE model and the random walk model, and between the Markov chain model and the random walk model are 0.54, 0.02, 0.01 and 0.86, 0.19, 0.41, respectively. This does not negate the need for the fixed ASHRAE model or the MC model. Rather, this signifies that, for a certain type of building, another occupant presence model should be introduced, e.g. the RW Model. [ABSTRACT FROM AUTHOR]

This paper presents the results of model predictive control (MPC) using multiple deep neural network (DNN) models for the cooling system of a factory building. The target building accommodates a large space (80 m × 60 m × 9.7 m) and serves cooling via 45 diffusers connected to a direct expansion air handling unit that comprises two air supply fans and four condensing units. The authors developed 10 simulation models using a DNN: One predicts the supply air temperature of the HVAC system, while the others predict the indoor temperature of nine zones. The models can sufficiently predict the thermal behavior of the HVAC system (normalized mean bias error (NMBE): −1.2 %, coefficient of variation of the root mean square error (CVRMSE): 3.8 %) and indoor environment (NMBE: 1.1 %, CVRMSE: 1.3 %). The purpose of the MPC is to minimize the energy consumption of the condensing units while maintaining the cooling set-point temperature. The MPC was applied to the target building for a sampling time of 10 min, for four weeks, from August 23 to September 17, 2021. It was found that energy consumption decreased by 35.1 % compared with the baseline period, while satisfying the cooling set-point temperature maintenance condition. Finally, the authors highlight that the DNN-based MPC is a practical approach sufficient for predicting indoor thermal behavior and has significant potential for energy saving. [ABSTRACT FROM AUTHOR]

Highlights: [•] This paper presents multi-criteria decision making under uncertainty. [•] A stochastic inference using Markov chain Monte Carlo method was applied. [•] Bayesian MCMC approach can be used for rational decision making. [ABSTRACT FROM AUTHOR]

Highlights: [•] The paper reports stochastic comparison between ISO 13790 and EnergyPlus. [•] The paper delivers the audience the need of stochastic calibration for ISO 13790. [•] The calibrated ISO model can predict similarly to dynamic simulation. [ABSTRACT FROM AUTHOR]

Abstract: Polygonum cuspidatum water extract (PCWE) was shown to be a potent inhibitor of lipopolysaccharide (LPS)-induced expression of inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2). PCWE was compared to baicalin isolated from Scutellaria baicalensis Georgi and berberine of Coptidis rhizoma and Phellodendri cortex, for their effects on LPS-induced nitric oxide (NO) production and iNOS and COX-2 gene expressions in RAW 264.7 macrophages. Both PCWE and the compounds inhibited LPS-induced NO production in a concentration-dependent manner without a cytotoxicity. The decrease in NO production was in parallel with the inhibition of LPS-induced iNOS gene expression by PCWE and the compounds. In contrast, iNOS enzyme activity was not inhibited by PCWE and two agents. In addition, only PCWE inhibited LPS-induced prostaglandin E2 (PGE2) production and COX-2 gene expression without affecting COX-2 enzyme activity, while baicalin or berberine did not. Furthermore, N-nitro-l-arginine (NLA) and N-nitro-l-arginine methyl ester (l-NAME) pretreatment enhanced LPS-induced iNOS protein expression, which was inhibited by these PCWE and two agents, although LPS-induced COX-2 protein expression was not affected by NLA and l-NAME. PCWE inhibited PGE2 production and COX-2 protein expression in NLA/LPS or l-NAME/LPS-co-treated RAW 264.7 cell, however, baicalin or berberine did not. From the results, it was concluded that co-treatment with NOS inhibitors and PCWE effectively blocks acute production of NO and inhibits expression of iNOS and COX-2 genes. [Copyright &y& Elsevier]

It is widely acknowledged that Bayesian inference is only beneficial when prior information is properly defined. However, there is no clear rule for prior selection, and it is apparently a matter of subjective selection by the domain expert(s). In other words, because the posterior inference results can vary depending on how the prior is set, a proper definition of the prior is important in terms of objectivity and accuracy for Bayesian inference of building energy models. Hence, the authors suggest a new prior selection method using Dirichlet process Gaussian mixture model (DPGMM) and the likelihood confidence region (hereafter referred to as likelihood CR). The DPGMM is a Bayesian nonparametric clustering technique that optimizes both the cluster shape and the number of clusters. Using the DPGMM, the finite probability distributions that make up the likelihood CR can be estimated, where the distribution with the highest maximum likelihood is applied as the informative prior. In this study, a reference office building of the United States Department of Energy was selected, and a surrogate model was generated using an artificial neural network. Based on a comparison between the authors' suggestion and traditional informative (and/or non-informative) priors by domain experts, the proposed method requires only minimum information about the parameters (min and max) and performs better than the traditional approach [ABSTRACT FROM AUTHOR]

The purpose of the present study was to investigate the relevance of building thermal performance and characteristics to building energy consumption. This paper reports an energy analysis of 4625 office buildings in Seoul, South Korea, using data from the Korean national building energy database and architectural database. The following four research questions were investigated: (1) Do old buildings consume more energy than new ones? (2) Have strict prescriptive building energy codes contributed to the reduction in energy use intensity (EUI, kWh/m2·year) over the past several decades? (3) What are the characteristics of building energy consumption in terms of season, age, and cooling system (electric chiller vs absorption chiller)? (4) Which factors in the Korean building energy database are relevant to building energy consumption? The analyses revealed that, contrary to common assumptions, new buildings did not always consume less energy than old buildings, and it may be wrong to attribute intensification of prescriptive building energy codes directly to building energy efficiency improvements. In addition, the building characteristics (i.e., district, year built, number of floors, number of elevators, and total floor area) available in the Korean building energy database do not adequately explain building energy consumption, and the existing data collection method needs further improvement. [ABSTRACT FROM AUTHOR]

This article compares two modeling approaches for optimal operation of a turbo chiller installed in an office building: (1) a machine learning model developed with artificial neural network (ANN) and (2) a hybrid machine learning model developed with the ANN model and available physical knowledge of the chiller. Before developing the ANN model of the chiller, the authors used Gaussian mixture model in order to check the validity of measured data. Then, the hybrid model was developed by combining the ANN model and physics-based regression equations from the EnergyPlus engineering reference. It was found that both the ANN and hybrid ANN model are satisfactory to predict the chiller's power consumption: mean bias error (MBE) = −2.63%, coefficient of variation of the root mean square error (CVRMSE) = 8.05% by the ANN model; MBE = −3.99%, CVRMSE = 11.98% by the hybrid ANN model. However, the hybrid model requires fewer inputs (four inputs) than the ANN model (eight inputs). The energy savings of both models are similar coefficient of performance (COP) = 4.32 by the optimal operation of the ANN model; COP = 4.44 by the optimal operation of the hybrid ANN model. In addition, the hybrid ANN model can be applied where the ANN model is unable to provide accurate predictions. [ABSTRACT FROM AUTHOR]

Uncertainty analysis in building energy assessment has become an active research field because a number of factors influencing energy use in buildings are inherently uncertain. This paper provides a systematic review on the latest research progress of uncertainty analysis in building energy assessment from four perspectives: uncertainty data sources, forward and inverse methods, application of uncertainty analysis, and available software. First, this paper describes the data sources of uncertainty in building performance analysis to provide a firm foundation for specifying variations of uncertainty factors affecting building energy. The next two sections focus on the forward and inverse methods. Forward uncertainty analysis propagates input uncertainty through building energy models to obtain variations of energy use, whereas inverse uncertainty analysis infers unknown input factors through building energy models based on energy data and prior information. For forward analysis, three types of approaches (Monte Carlo, non-sampling, and non-probabilistic) are discussed to provide sufficient choices of uncertainty methods depending on the purpose and specific application of a building project. For inverse analysis, recent research has concentrated more on Bayesian computation because Bayesian inverse methods can make full use of prior information on unknown variables. Fourth, several applications of uncertainty analysis in building energy assessment are discussed, including building stock analysis, HVAC system sizing, variations of sensitivity indicators, and optimization under uncertainty. Moreover, the software for uncertainty analysis is described to provide flexible computational environments for implementing uncertainty methods described in this review. This paper concludes with the trends and recommendations for further research to provide more convenient and robust uncertainty analysis of building energy. Uncertainty analysis has been ready to become the mainstream approach in building energy assessment although a number of issues still need to be addressed. [ABSTRACT FROM AUTHOR]