Building Performance Data Research Articles

Rateworkspace: BIM integrated post-occupancy evaluation system for office buildings

The feedback obtained from occupants regarding their comfort needs and performance of buildings is critical for assessing occupant satisfaction, identifying the operation and maintenance (O&M) issues in time and for improving resource efficiency in buildings. Current facility management (FM) systems and occupant feedback collection practices, however, have limitations in supporting effective decision-making in FM, as they lack the necessary contextual data related to the occupant feedback (e.g., building geometry, systems, elements). Building Information Modeling (BIM)-enabled FM systems are used for combining different types of FM information with building models; however, occupant feedback is still not effectively utilized in FM since it is not integrated with BIM. In this study, a BIM integrated post-occupancy evaluation system prototype is developed for: (1) collecting occupant feedback along with the contextual information related to the feedback items in a structured way, and (2) presenting this information as integrated with BIM to the facility managers. This enables conducting spatio-temporal queries and supports effective decision-making by visualizing the collected feedback. The prototype was designed by using qualitative shadowing with FM teams to identify information needs and use case analysis to determine how contextual data integrated with BIM could be collected from office occupants who are non-technical persons with limited information on building models. This paper identifies the FM query categories that are required to process the occupant feedback and describes the RateWorkSpace prototype developed for office buildings. The deployment of the prototype in a real-world office demonstrates that the proposed system is applicable, practical, usable, and that real-time building performance data can be both collected and analysed with the developed system. This has the potential to increase the effectiveness of the FM and O&M processes, and help to create office spaces with optimized energy use and occupant comfort that also supports occupant well-being and productivity.

A novel image-based transfer learning framework for cross-domain HVAC fault diagnosis: From multi-source data integration to knowledge sharing strategies

Data-driven classification models have gained increasing popularity for fault detection and diagnosis (FDD) tasks considering their advantages in implementation flexibility and modeling accuracies. To tackle the wide existence of data shortage challenges for individual buildings, transfer learning can be adopted to enhance the applicability of data-driven approaches. At present, limited studies have been conducted to explore the potentials of transfer learning in HVAC FDD tasks, leaving the following two key questions unanswered, i.e., (1) whether the tabular data collected from different building systems can be effectively integrated and utilized as the source data for transfer learning, and (2) whether the operational patterns learnt from a specific building system can be interchangeably applied for FDD tasks of other systems. This study proposes a novel image-based transfer learning framework to tackle the multi-source data compatibility challenge in the building field, while investigating the value of transfer learning in cross-domain FDD tasks. Data experiments have been designed to quantify the value of transfer learning given different data amounts, imbalance ratios, and transfer learning strategies. The research results validate the usefulness of image-based transfer learning for HVAC FDD tasks. The insights obtained are valuable for multi-source building operational data integration and cross-domain knowledge sharing.

Machine-learning-based model predictive control with instantaneous linearization – A case study on an air-conditioning and mechanical ventilation system

Machine-learning (ML) –based building models have been gaining popularity in constructing model predictive control (MPC) for building energy management applications. However, ML-based building models are usually nonlinear so to capture the building dynamics, leading to high computation load for MPC, prohibiting its application for real-time building control. This study proposes a ML-based MPC with an instantaneous linearization (IL) scheme, which employs real-time building operation data to linearize the nonlinear ML-based building model for constructing a linear MPC at each control interval. The proposed ML-based MPC with IL system is implemented to control an air conditioning system in an office of a general hospital building located in Singapore for experimental evaluation of its control performance. The ML-based MPC with IL is compared to a ML-based MPC that directly uses a nonlinear ML-based building model and the original reactive-control-based thermostat of the office. Results show that the ML-based MPC with IL significantly reduced the computation time (by more than 70 times) as compared to the ML-based MPC while retained most of the advantages of the ML-based MPC. The ML-based MPC with IL and the ML-based MPC achieved 31.6% and 26.0% reductions, respectively, in cooling energy consumption as compared to the original thermostat. Meanwhile, both the MPC systems significantly improved indoor thermal comfort for the office as compared to the original thermostat. The study demonstrated that using IL for ML-based MPC could substantially improve computation efficiency with no obvious performance degradation in terms of thermal comfort and energy saving.

Data-driven building energy modeling with feature selection and active learning for data predictive control

Three gaps impede the development of cost-effective and accurate data-driven building energy modeling/models (DBEM) for energy forecasting and predictive control strategies. Gap 1: data bias is common in building operation data, but this topic is hardly studied in DBEM; Gap 2: high data dimensionality is common in DBEM, but a systematic and scalable methodology is lacking to solve the problem; Gap 3: the interactions between data bias and high dimensionality have not been systematically studied for DBEM and predictive control in buildings.In this paper, to address the three gaps mentioned above, we develop a framework that integrates active learning and feature selection for DBEM used for whole building data predictive control (or DPC, which is a branch of model predictive control). The framework provides a systematic methodology and automatic workflow that starts with raw data from building automation systems to the establishment of data-driven energy models for DPC controllers. The developed strategies and framework are evaluated in a virtual testbed based on EnergyPlus and BCVTB. Improved performance and reduced computational complexity are observed from the DBEM built with the developed framework, as well as the DPC controller based on that DBEM, indicating the effectiveness of the developed framework.

Analytic hierarchy process-based fuzzy post mining method for operation anomaly detection of building energy systems

Association rule mining has shown outstanding capacity in extracting operation patterns from numerous building operational data. However, only a very small portion of them is valuable for energy efficiency enhancement. Advanced post mining methods are necessary for automatically deleting most of worthless association rules. Therefore, this study proposes an analytic hierarchy process-based fuzzy post mining method. Three criteria and six corresponding sub-criteria are developed to evaluate the value of each association rule in energy efficiency enhancement. They offer new levels to assess the value of association rules. The fuzzy set theory is introduced to grade each sub-criterion of an association rule. It considers the uncertainties from the imprecise judgments, which can improve the robustness of rule assessing. Analytic hierarchy process is adopted to determine the weight of each criterion/sub-criterion for getting the weighted overall score of an association rule. It provides a solution to assessing the value of association rules from multiple aspects, which can improve the performance of post mining. The proposed method is evaluated using 117,636 association rules extracted from the one-year historical operational data of an actual chiller plant. Four common indexes (support, confidence, lift and distance correlation) are selected as a traditional method for comparison with the proposed method. The two methods can both filter out approximately 96.00% of worthless association rules. As for valuable association rules, the traditional method only extracts 17.28% of them, while this proportion is 93.51% for the proposed method. It proves that the proposed method has excellent performance of valuable association rule extraction.

Experimental investigation on thermal inertia characterization of commercial buildings for demand response

Characterizing the thermal inertia of commercial buildings is of great importance in quantifying demand flexibility, and experimental tests can be used for this purpose. However, existing studies tend to be qualitative and have a limited scope. In this study, a comprehensive experimental study has been performed to assess the thermal inertia of commercial buildings. In this field test, six buildings were selected with different sizes, vintages, and types of heating, ventilation, and air conditioning (HVAC) systems. The studied HVAC systems are used for comfort heating and cooling in over 80% of the U.S. commercial building stock. We quantified thermal inertia using building operation data collected under various thermostat excitation signals. We then studied the relationship between thermal inertia and intrinsic properties, such as floor area, HVAC system, etc., as well as operation condition indicators including outdoor air temperature, zone temperature, and occupancy. The test results indicate that the median values of the normalized charge response time and the normalized discharge response time of the five buildings are 1–5 hr/°C and − 5 to − 1 hr/°C, respectively. The results also show that the thermal inertia of commercial buildings may be sensitive to the HVAC system type but not the floor area or location of zones, i.e., core vs. perimeter. Furthermore, our results suggest that the relationships between the normalized charging/discharge response time and the zone/outdoor temperature may vary among zones and be highly nonlinear.

Attention-based interpretable neural network for building cooling load prediction

Machine learning has gained increasing popularity in building energy management due to its powerful capability and flexibility in model development as well as the rich data available in modern buildings. While machine learning is becoming more powerful, the models developed, especially artificial neural networks like Recurrent Neural Networks (RNN), are becoming more complex, resulting in “darker models” with lower model interpretability. The sophisticated inference mechanism behind machine learning prevents ordinary building professionals from understanding the models, thereby lowering trust in the predictions made. To address this, attention mechanisms have been widely implemented to improve the interpretability of deep learning; these mechanisms enable a deep learning-based model to track how different inputs influence outputs at each step of inference.This paper proposes a novel neural network architecture with an attention mechanism for developing RNN-based building energy prediction, and investigates the effectiveness of this attention mechanism in improving the interpretability of RNN models developed for 24-hour ahead building cooling load prediction. To better understand, explain and evaluate these neural network-based building energy prediction models, the obtained attention vectors (or metric) are used to visualize the influence of different parts of model inputs on the prediction result. This helps the users to understand why predictions are made by the model, as well as how input sequences proportionally influence the output sequences. Further analysis of attention vectors can provide interesting temporal information for understanding building thermal dynamics, like the thermal inertia of the building. The proposed attention-based architecture can be implemented in developing optimal operation control strategies and improving demand and supply management. The model developed based on this architecture is assessed using real building operational data, and shows improved accuracy and interpretability over baseline models (without adopting attention mechanisms). The research results help to bridge the gap between building professionals and advanced machine learning techniques. The insights obtained can be used as guidance for the development, fine-tuning, explanation and debugging of data-driven building energy prediction models.

Data mining cubes for buildings, a generic framework for multidimensional analytics of building performance data

Over the last decade, collecting massive volumes of data has been made all the more accessible, pushing the building sector to embrace data mining as a powerful tool for harvesting the potential of big data analytics. However repetitive challenges still persist emerging from the need for a common analytical frame, effective application- and insight-driven targeted data selection, as well as benchmarked-supported claims. This study addresses these concerns by putting forward a generic stepwise multidimensional data mining framework tailored to building data, leveraging the dimensional-structures of data cubes. Using the open Building Data Genome Project 2 set, composed of 3053 energy meters from 1636 buildings, we provide an online, open access, implementation illustration of our method applied to automated pattern identification. We define a 3-dimensional building cube echoing typical analytical frames of interest, namely, bottom-up, top-down and temporal drill-in approaches. Our results highlight the importance of application and insight driven mining for effective dimensional-frame targeting. Impactful visualizations were developed allowing practical human inspection, paving the path towards more interpretable analytics.

Active learning strategy for high fidelity short-term data-driven building energy forecasting

The quality of a data-driven model is heavily dependent on the quality of data. Data from building operation often have data bias problems, which means that the data sample is collected in a way that some members of the intended data population are less likely to be included than others. Data-driven energy forecasting models built on such data hence are biased and could lead to large forecasting errors. Active learning— an effective method to defying data bias—is rarely studied or applied in the area of data-driven building energy forecasting modeling. This paper attempts to fill this gap and explores the application of active learning in data-driven building energy forecasting.The developed strategy in this paper efficiently generate informative training data within a time budget and uses block design to passively consider weather disturbances. The developed active learning strategy is applied and evaluated in both virtual and real-building testbeds against traditional data-driven methods. Via these virtual and real-building evaluation cases, we have demonstrated that the data bias problem typically exists in building operation data is resolved by applying the developed active learning strategy. Building energy forecasting models trained from data generated from the active learning strategy have shown improved performances in both model accuracy and model extendibility perspectives. The effectiveness of the block design module is also validated to effectively consider the impact of weather conditions on active learning design.

A Review on Data Preprocessing Techniques Toward Efficient and Reliable Knowledge Discovery From Building Operational Data

The rapid development in data science and the increasing availability of building operational data have provided great opportunities for developing data-driven solutions for intelligent building energy management. Data preprocessing serves as the foundation for valid data analyses. It is an indispensable step in building operational data analysis considering the intrinsic complexity of building operations and deficiencies in data quality. Data preprocessing refers to a set of techniques for enhancing the quality of the raw data, such as outlier removal and missing value imputation. This article serves as a comprehensive review of data preprocessing techniques for analysing massive building operational data. A wide variety of data preprocessing techniques are summarised in terms of their applications in missing value imputation, outlier detection, data reduction, data scaling, data transformation, and data partitioning. In addition, three state-of-the-art data science techniques are proposed to tackle practical data challenges in the building field, i.e., data augmentation, transfer learning, and semi-supervised learning. In-depth discussions have been presented to describe the pros and cons of existing preprocessing methods, possible directions for future research and potential applications in smart building energy management. The research outcomes are helpful for the development of data-driven research in the building field.

Detecting deficiencies using building performance data in healthcare facilities: Improving operational efficiency with Continuous Commissioning ®

The Continuous Commissioning® version of the Existing Building Commissioning process was implemented in three large hospitals and optimized the HVAC operations to improve their energy use while maintaining comfort without any capital system investment. In the hospitals studied, measures accounting for 64% of the total annual commissioning cost savings of $592,816 were identified by analyzing trend data. Trend analysis of AHU ON/OFF status, economizer, outside air damper position, supply air temperature, and supply air static pressure corresponded to 51%, 8%, 5%, 31%, and 4% of the total projected savings, respectively. The use of trend analysis identified measures in the three case-study hospitals that are projected to reduce the EUI of the three hospitals by 16%, 21% and 19% respectively, even though the average consumption of the three hospitals was below the 5th percentile of hospital consumption. The availability and use of trend data in these facilities saved time in identifying these operational improvements and likely identified some improvements that would not have been identified using manual diagnostic techniques. Continuous engagement of this low cost and resource-intensive analysis tool can significantly reduce the operational costs for facilities.

A study on semi-supervised learning in enhancing performance of AHU unseen fault detection with limited labeled data

The fault detection and diagnosis (FDD) of air handling units (AHUs) serves as a major task in building operation management and energy savings. Data-driven classification methods have gained increasing popularities considering their flexibilities and effectiveness in practice. One essential challenge in developing accurate and reliable FDD classification models is the lack of sufficient labeled data. In practice, it can be highly time-consuming, labor-intensive and sometimes even infeasible to collected sufficient labeled data for all possible faulty operations. As a result, the fault detection models developed by limited and partially labeled data may not perform well in detecting any unknown or unseen faults in AHU operations. This study investigates the value of semi-supervised learning in detecting unseen faults during AHU operations. The main idea is to adopt a self-training strategy to gradually enhance the model capability by leveraging large amounts of unlabeled data. Data experiments have been designed to evaluate the unseen fault detection performance, the impacts of key semi-supervised learning parameters and the difficulties in detecting typical AHU faults. The insights obtained are valuable for the integration of data sciences with massive building operational data for smart building management.

Does historical data still count? Exploring the applicability of smart building applications in the post-pandemic period

The emergence of COVID-19 pandemic is causing tremendous impact on our daily lives, including the way people interact with buildings. Leveraging the advances in machine learning and other supporting digital technologies, recent attempts have been sought to establish exciting smart building applications that facilitates better facility management and higher energy efficiency. However, relying on the historical data collected prior to the pandemic, the resulting smart building applications are not necessarily effective under the current ever-changing situation due to the drifts of data distribution. This paper investigates the bidirectional interaction between human and buildings that leads to dramatic change of building performance data distributions post-pandemic, and evaluates the applicability of typical facility management and energy management applications against these changes. According to the evaluation, this paper recommends three mitigation measures to rescue the applications and embedded machine learning algorithms from the data inconsistency issue in the post-pandemic era. Among these measures, incorporating occupancy and behavioural parameters as independent variables in machine learning algorithms is highlighted. Taking a Bayesian perspective, the value of data is exploited, historical or recent, pre- and post-pandemic, under a people-focused view.

An anomaly detection and dynamic energy performance evaluation method for HVAC systems based on data mining

With the wide application of building automation systems (BASs), a large amount of building operation data are usually available, which provide a good basis for the optimal operation of a building’s heating, ventilation and air conditioning (HVAC) systems. In this study, a data mining (DM)-based method is proposed for the anomaly detection and dynamic energy performance evaluation of an HVAC system. In this method, first a DM technology is used to detect the abnormal operation data from historical operation data and identify the possible reasons for abnormalities. Then, the identified abnormal energy consumption data caused by faults are corrected. On this basis, a multilevel dynamic energy performance benchmark and a set of energy performance evaluation rules for the HVAC system are established. Finally, the real-time operation performance of an HVAC system is evaluated, and the causes of abnormal energy consumption are identified at multiple levels. The effectiveness of the proposed method is verified in a case study of a commercial building with a complex cooling system.

Using pattern recognition to characterise heating behaviour in residential buildings

ABSTRACT The understanding of energy-related occupant behaviour and its better reproduction in building energy analysis has recently become a primary field of interest. The combination of computing ease and the availability of big streams of high resolution building performance data enhance the ability to study this behaviour. In this context, data mining methods are increasingly employed. The aim of this study is to propose a data mining methodology for the characterization of heating behaviour in residential buildings. The methodology consists of two multivariate statistical analysis methods, namely Principal Component Analysis (PCA) followed by cluster analysis. The methods were applied on monitored gas consumption data of five dwellings in Italy. Findings support literature indicating that people heat their homes in different ways. It was found that households do not always follow a different heating schedule on weekends, have very different temperature preferences and operate the heating system at different hours during the day. In fact, some households may change heating practices over the heating season. The highlight of the proposed methodology is the insightful and simple way that PCA can extract succinct information about the heating behaviour of the user.

From Electricity and Water Consumption Data to Information on Office Occupancy: A Supervised and Unsupervised Data Mining Approach

Climate change and technological development are pushing buildings to become more sophisticated. The installation of modern building automation systems, smart meters, and IoT devices is increasing the amount of available building operational data. The common term for this kind of building is a smart building but producing large amounts of raw data does not automatically offer intelligence that would offer new insights to the building’s operation. Smart meters are mainly used only for tracking the energy or water consumption in the building. On the other hand, building occupancy is usually not monitored in the building at all, even though it is one of the main influencing factors of consumption and indoor climate parameters. This paper is bringing the true smart building closer to practice by using machine learning methods with sub-metered electricity and water consumptions to predict the building occupancy. In the first approach, the number of occupants was predicted in an office floor using a supervised data mining method Random Forest. The model performed the best with the use of all predictors available, while from individual predictors, the sub-metered electricity used for office equipment showed the best performance. Since the supervised approach requires the continuous long-term collection of ground truth reference data (between one to three months, by this study), an unsupervised data mining method k-means clustering was tested in the second approach. With the unsupervised method, this study was able to predict the level of occupancy in a day as zero, medium, or high in a case study office floor using the equipment electricity consumption.

A Pattern-Recognition-Based Ensemble Data Imputation Framework for Sensors from Building Energy Systems

Building operation data are important for monitoring, analysis, modeling, and control of building energy systems. However, missing data is one of the major data quality issues, making data imputation techniques become increasingly important. There are two key research gaps for missing sensor data imputation in buildings: the lack of customized and automated imputation methodology, and the difficulty of the validation of data imputation methods. In this paper, a framework is developed to address these two gaps. First, a validation data generation module is developed based on pattern recognition to create a validation dataset to quantify the performance of data imputation methods. Second, a pool of data imputation methods is tested under the validation dataset to find an optimal single imputation method for each sensor, which is termed as an ensemble method. The method can reflect the specific mechanism and randomness of missing data from each sensor. The effectiveness of the framework is demonstrated by 18 sensors from a real campus building. The overall accuracy of data imputation for those sensors improves by 18.2% on average compared with the best single data imputation method.

Development of an ANN-based building energy model for information-poor buildings using transfer learning

Accurate building energy prediction is vital to develop optimal control strategies to enhance building energy efficiency and energy flexibility. In recent years, the data-driven approach based on machine learning algorithms has been widely adopted for building energy prediction due to the availability of massive data in building automation systems (BASs), which automatically collect and store real-time building operational data. For new buildings and most existing buildings without installing advanced BASs, there is a lack of sufficient data to train data-driven predictive models. Transfer learning is a promising method to develop accurate and reliable data-driven building energy prediction models with limited training data by taking advantage of the rich data/knowledge obtained from other buildings. Few studies focused on the influences of source building datasets, pre-training data volume, and training data volume on the performance of the transfer learning method. The present study aims to develop a transfer learning-based ANN model for one-hour ahead building energy prediction to fill this research gap. Around 400 non-residential buildings’ data from the open-source Building Genome Project are used to test the proposed method. Extensive analysis demonstrates that transfer learning can effectively improve the accuracy of BPNN-based building energy models for information-poor buildings with very limited training data. The most influential building features which influence the effectiveness of transfer learning are found to be building usage and industry. The research outcomes can provide guidance for implementation of transfer learning, especially in selecting appropriate source buildings and datasets for developing accurate building energy prediction models.

An Edge-Cloud Integrated Solution for Buildings Demand Response Using Reinforcement Learning

Buildings, as major energy consumers, can provide great untapped demand response (DR) resources for grid services. However, their participation remains low in real-life. One major impediment for popularizing DR in buildings is the lack of cost-effective automation systems that can be widely adopted. Existing optimization-based smart building control algorithms suffer from high costs on both building-specific modeling and on-demand computing resources. To tackle these issues, this paper proposes a cost-effective edge-cloud integrated solution using reinforcement learning (RL). Beside RL's ability to solve sequential optimal decision-making problems, its adaptability to easy-to-obtain building models and the off-line learning feature are likely to reduce the controller's implementation cost. Using a surrogate building model learned automatically from building operation data, an RL agent learns an optimal control policy on cloud infrastructure, and the policy is then distributed to edge devices for execution. Simulation results demonstrate the control efficacy and the learning efficiency in buildings of different sizes. A preliminary cost analysis on a 4-zone commercial building shows the annual cost for optimal policy training is only 2.25% of the DR incentive received. Results of this study show a possible approach with higher return on investment for buildings to participate in DR programs.

CU-BEMS, smart building electricity consumption and indoor environmental sensor datasets

This paper describes the release of the detailed building operation data, including electricity consumption and indoor environmental measurements, of the seven-story 11,700-m2 office building located in Bangkok, Thailand. The electricity consumption data (kW) are that of individual air conditioning units, lighting, and plug loads in each of the 33 zones of the building. The indoor environmental sensor data comprise temperature (°C), relative humidity (%), and ambient light (lux) measurements of the same zones. The entire datasets are available at one-minute intervals for the period of 18 months from July 1, 2018, to December 31, 2019. Such datasets can be used to support a wide range of applications, such as zone-level, floor-level, and building-level load forecasting, indoor thermal model development, validation of building simulation models, development of demand response algorithms by load type, anomaly detection methods, and reinforcement learning algorithms for control of multiple AC units.