Building Energy Consumption Data Research Articles

Reducing uncertainty of building shape information in urban building energy modeling using Bayesian calibration

This study proposes urban building energy modeling that extends beyond single-building-level models to the urban level. However, most urban building energy models use representative buildings that may not accurately reflect the diversity of building shapes, systems, and envelope performance when conducting building energy evaluations at the urban scale. To address this issue, previous studies have utilized representative buildings and Bayesian calibration to estimate uncertain building information parameters without considering building shape information. Therefore, the primary objective of this study is to estimate building shape information using artificial neural networks and Bayesian calibration based on building energy consumption data to identify the shape information uncertainty of representative buildings. The results indicate that some shape information can be estimated by comparing the overall distribution of the building stock using the two-sample Kolmogorov–Smirnov test. Furthermore, we found that the proposed energy modeling methodology yields energy consumption patterns similar to those of the target building stock. This preliminary investigation addresses the uncertainty of representative buildings in urban-scale modeling, elucidates the relationship between building form and energy consumption, and introduces a method for inferring shape information from energy consumption data.

Analysis and prediction of energy consumption in office buildings with variable refrigerant flow systems: A case study

Accurately predicting building air-conditioning energy consumption is particularly important for energy management. However, for small datasets, there is a lack of simple and effective prediction methods applicable to multiple buildings. In order to fully explore the energy consumption mode of Variable Refrigerant Flow (VRF) air conditioning system and improve the accuracy of the model. This study uses the weather data of an outdoor meteorological data collection system and the air conditioning system and energy consumption data of an office building in Jinan. Initial analysis revealed that occupant behavior and daily variations in outdoor air temperatures significantly impact the energy consumption of air conditioning systems. Then, by applying the random forest algorithm and Pearson correlation analysis, the study identified daily average outdoor air temperature and VRF indoor unit running time as critical features for prediction. Subsequently, three machine learning models—Multiple Linear Regression (MLR), Back Propagation Neural Network (BPNN), and Long Short-Term Memory (LSTM) neural network—were developed and assessed using a 4:1 training-to-testing set ratio. Comparative analysis revealed that all models performed robustly, with the MLR model exhibited superior predictive accuracy (R2 (Coefficient of Determination) = 0.98, MAE (Mean Absolute Error) = 0.005 kWh/(m2·day), and RMSE (Root Mean Square Error) = 0.007 kWh/(m2·day) for the testing set) and simplicity. Furthermore, the effectiveness of the MLR model in handling small sample sizes was confirmed through cross-validation of data from two additional office buildings, highlighting its suitability for predicting energy consumption of VRF air conditioning systems in office buildings.

K-PCD: A new clustering algorithm for building energy consumption time series analysis and predicting model accuracy improvement

Clustering algorithms are often applied to building energy consumption data analysis to mine representative patterns of building energy usage. This paper proposes a new clustering algorithm: K-PCD, which is particularly suitable for building energy consumption time series. K-PCD utilizes a Pearson correlation coefficient-based distance measure (PCD), and a novel centroid calculation method that takes into account both the PCD and the traditional Euclidean distance (ED) between time series. This study also proposes a new clustering validity index (CVI) tailored to the predictability of building energy consumption: Energy Prediction Clustering Performance Index (EPCPI). The K-PCD and EPCPI are practically analyzed and validated using one year of data from 29 real buildings. The results indicate that comparing with traditional clustering algorithms, the K-PCD achieves better clustering results. This EPCPI index thoroughly explores building operation patterns, and after adding clustering labels, it can maximize the prediction accuracy of the Back Propagation Neural Network (BPNN) model for building energy consumption.

Clustering Method Comparison for Rural Occupant’s Behavior Based on Building Time-Series Energy Data

The purpose of this research is to compare clustering methods and pick up the optimal clustered approach for rural building energy consumption data. Research undertaken so far has mainly focused on solving specific issues when employing the clustered method. This paper concerns Yushan island resident’s time-series electricity usage data as a database for analysis. Fourteen algorithms in five categories were used for cluster analysis of the basic data sets. The result shows that Km_Euclidean and Km_shape present better clustering effects and fitting performance on continuous data than other algorithms, with a high accuracy rate of 67.05% and 65.09%. Km_DTW is applicable to intermittent curves instead of continuous data with a low precision rate of 35.29% for line curves. The final conclusion indicates that the K-means algorithm with Euclidean distance calculation and the k-shape algorithm are the two best clustering algorithms for building time-series energy curves. The deep learning algorithm can not cluster time-series-building electricity usage data under default parameters in high precision.

Office building energy consumption forecast: Adaptive long short term memory networks driven by improved beluga whale optimization algorithm

With the development of urbanization, buildings have become a major source of energy consumption. This research uses a data-driven approach to achieve accurate building energy consumption prediction by analyzing and modeling building energy consumption data. The model proposed in this paper uses improved beluga whale optimization algorithm (IBWO) to optimize long short-term memory networks (LSTM) for accurate energy consumption prediction. In order to enhance the ability of BWO in global search and local exploitation, a new method of dynamic adjustment of step factor as well as strategies such as nonlinear decreasing are introduced to improve BWO. For the first time, it is proposed to explore the accuracy of the number of hyper-parameters of LSTM on the prediction of energy consumption, and the improved beluga whale optimization algorithm is used to optimize the two, three, and four hyper-parameters of LSTM respectively. Then short-term prediction of historical energy consumption data of an office building in Xi'a is performed. Experiments show that the optimization of the four hyperparameters of LSTM using the IBWO of this paper can reduce the mean absolute error (MAE) of the pre-improvement model from 830.71 KW to 128.28 KW, the mean absolute percentage error (MAPE) from 12.32 % to 1.38 %, and the coefficient of variation (CV) from 7.5 % to 1.2 %.

A Filling Method Based on K-Singular Value Decomposition (K-SVD) for Missing and Abnormal Energy Consumption Data of Buildings

Massive data can be collected from meters to analyze the energy use behavior and detect the operation problems of buildings. However, missing and abnormal data often occur for the raw data. Effective data filling and smoothing methods are required to improve data quality before conducting the analysis. This paper introduces a data filling method based on K-SVD. The complete dictionary is trained and then utilized to reconstruct incomplete samples to fill the missing or abnormal data. The impacts of the dictionary size, the data missing continuity, and the sample size on the performance of the proposed method are studied. The results show that a smaller dictionary size is recommended considering the computational complexity and accuracy. The K-SVD method outperforms traditional methods, showing a reduction in the MAPE and CVRMSE by 3.8–5.4% and 6.7–87.8%. The proposed K-SVD filling method performs better for non-consecutive missing data, with an improvement in the MAPE and CVRMSE by 0.1–4% and 5.1–6.7%. Smaller training samples are recommended. The method proposed in this study would provide an effective solution for data preprocessing in building and energy systems.

Thermal Modeling and Electric Space Heating of a University Building in Newfoundland

Buildings play a substantial role in global energy consumption, constituting a considerable share of the overall energy use. In Canada, they contribute to around 25% of the total final energy consumption. Notably, space heating emerges as the primary energy consumer, accounting for approximately 57% of energy utilization in institutional and commercial buildings. This paper presents a feasibility analysis of converting the space heating system of the Core Science Facility (CSF) building of Memorial University of Newfoundland (MUN). Analysis is done using RETScreen Clean Energy Management Software, known as RETScreen Expert, a software package developed by the Government of Canada, and the thermal modeling of the building using Energy3D, developed by the National Renewable Energy Laboratory (NREL). The feasibility study indicates that significant savings can be achieved if space heating is switched to electric resistive heating. The results indicate a 24.2% savings in annual energy costs, with a simple payback period of 10.5 years. The simulation results from Energy3D are compared with the measured building energy consumption data provided by the MUN Facilities Management Department. The thermal model indicates less energy consumption than the actual measured values, which is a result of transmission losses, the interconnection between the CSF building and the University Center, building occupancy, the ventilation system, and degradation of equipment that are not considered in the model.

Prediction and Management of Building Energy Consumption Based on Building Environment Simulation Design Platform DeST and Meteorological Data Analysis Algorithm

Currently, the carbon emissions of building energy consumption account for a significant portion of all carbon emissions. How to reduce carbon emissions to achieve carbon neutrality is an important current research direction. Therefore this research builds a predictive algorithm model for analyzing energy consumption data of meteorological buildings using DeST platform for energy saving and emission reduction to achieve carbon neutrality. The new model uses Internet of Things and cloud platform technology to build a simulation building platform, and uses the support vector machine algorithm in the analysis algorithm to vectorize building energy consumption data, which can achieve normalization processing of building energy consumption and meteorological data. By processing building energy consumption data, prediction of building energy consumption at the next moment can be achieved. The experimental results show that the precision and accuracy of the new algorithm are higher than genetic algorithm 1 and 0.15 respectively, and 0.6 and 0.07 higher than clustering analysis algorithm respectively. Therefore, applying this algorithm model to building energy consumption prediction can significantly improve the accuracy and precision of the algorithm.

A method for monitoring energy consumption data of near zero energy buildings based on BIM technology

A method for monitoring energy consumption data of near zero energy buildings based on BIM technology

An evolutionary deep learning model based on EWKM, random forest algorithm, SSA and BiLSTM for building energy consumption prediction

Accurate prediction of building energy consumption is crucial to the rational scheme of building energy. Combining the entropy-weighted K-means (EWKM) with the random forest (RF) method, a feature selection (EWKM-RF) method is proposed in this paper. Based on the proposed EWKM-RF method, the classification and feature selection of the energy consumption influencing factors can be achieved exclusively. Meanwhile, based on the EWKM-RF method and the bi-directional long-short-term memory neural network (BiLSTM) optimized by the sparrow search algorithm (SSA), an RF-SSA-BiLSTM prediction model for building energy consumption is established in this paper. As the weight, learning rate, and hidden layer node parameters of the BiLSTM neural network are optimized with the SSA, the constraints for manually adjusting parameters are avoided in the proposed prediction model. To examine the accuracy of the proposed model, energy consumption data of a civil public building in Dalian city are collected and tested. Results show the prediction error of RF-SSA-BiLSTM after feature selection is reduced by 24.55 % in high and low energy consumption months. Compared with RF-BiLSTM, RF-PSO-BiLSTM, and RF-CNN-BiLSTM, the RF-SSA-BiLSTM has strong robustness. The average MAE, RMSE and MAPE values of energy consumption prediction in the four seasons are 1.30, 1.63 and 0.02.

A review of prediction models of total carbon emission for civil buildings in China

The carbon emissions from the building sector are one of the major sources of carbon emissions globally. In order to address global climate change, the Chinese government has proposed the 3,060 dual carbon goals. In this context, the government urgently needs a predictive model for calculating and forecasting the energy consumption and carbon emissions in the construction industry to help formulate decarbonization strategies. The review and analysis of a predictive model for the current total carbon dioxide emissions in the Chinese construction sector can provide a basis for calculating and predicting carbon emissions, as well as for formulating corresponding emission reduction policies. This article analyzes the Carbon emission factor and the methods of obtaining building energy consumption data, which are crucial for predicting carbon emissions. Furthermore, it examines the predictive models for total CO2 emissions in the Chinese construction sector and summarizes their respective advantages and limitations. Finally, it highlights the shortcomings of existing research in terms of carbon emission factors, energy consumption data, and accounting scope, while suggesting future research directions.

Optimization of building demand flexibility using reinforcement learning and rule-based expert systems

The increasing use of renewable energy in buildings requires optimization of building demand flexibility to reduce energy costs and carbon emissions. Nevertheless, the optimization process is generally challenging that needs to consider the on-site intermittent energy supply, dynamic building energy demand, and proper utilization of energy storage systems. Leveraging the growing availability of operational data in buildings, data-driven strategies such as reinforcement learning (RL) have emerged as effective approaches to optimizing building demand flexibility. However, training a reliable RL agent is practically data-demanding and time-consuming, limiting its practical applicability. This study proposes a new strategy that integrates a rule-based expert system (RBES) and RL agents into the decision-making process to jointly reduce building energy costs, minimize the peak-to-average ratio (PAR) of grid power, and maximize PV self-consumption. In this strategy, the RBES determines system operation directly in less complex decision-making scenarios, while, in more intricate decision-making environments, it provides a reference decision for RL to explore optimal solutions further. This integration empowers RL agents to avoid unnecessary exploration and significantly enhance learning efficiency. The proposed strategy was tested using PV generation data and energy consumption data of a low energy office building. The results demonstrated an 85.7% improvement in RL learning efficiency and this strategy can successfully avoid sub-optimal convergence during policy learning. Compared to relying solely on the RBES, the proposed strategy led to 5.4% and 19.2% reductions in the electricity costs and daily PAR of grid power at peak hours, respectively. The strategy also achieved a satisfying PV self-consumption ratio of 62.4%, which is merely 0.4% lower than the optimal value determined by the RBES strategy that prioritized maximizing PV self-consumption. Additionally, compared with a model predictive control method developed for cost reduction, the strategy achieved similar cost savings while significantly reducing the decision time.

Improving the accuracy of multi-step prediction of building energy consumption based on EEMD-PSO-Informer and long-time series

Accurate multi-step forecasting of building energy consumption is an essential tool for effective planning and plays a vital role in building energy management systems. With the advent of big data, many artificial intelligence techniques require long time series to predict multi-step energy consumption. However, energy consumption data of buildings are often nonlinear and non-stationary in the state as well as considerable period, making model prediction more difficult. In this research, we propose a hybrid algorithm that combines the ensemble empirical modal decomposition (EEMD) and informer, where the parameters of the informer are optimized by the particle swarm optimization algorithm (PSO). At the beginning of our improved method, we use EEMD to break down the raw data into several intrinsic mode functions(IMF) components. Informer is then used to make predictions, and PSO is used to tune hyper-parameters during prediction. In the final step, the final prediction result is obtained by combining the prediction results of all IMF components. The hourly electricity consumption of five buildings in the BDG2 dataset is used to evaluate the effectiveness of the proposed method. Five existing models are compared with this method and evaluated by different performance metrics. From the perspective of five cases, the accuracy rate has increased by up to 78.68% compared with the existing methods. Compared with the original model, the accuracy rate has increased by up to 56.11%.

An Integrated Decision-Making Framework for Mitigating the Impact of Urban Heat Islands on Energy Consumption and Thermal Comfort of Residential Buildings

Urban heat island (UHI) is a zone that is significantly warmer than its surrounding rural zones as a result of human activities and rapid and dense urbanization. Excessive air temperature due to the UHI phenomenon affects the energy performance of buildings and human health and contributes to global warming. Knowing that most of the building energy is consumed by residential buildings, therefore, developing a framework to mitigate the impact of the UHI on residential building energy performance is vital. This study develops an integrated framework that combines hybrid micro-climate and building energy performance simulations and multi-criteria decision-making techniques. As a case study, an urban area is analyzed under the Urban GreenUP project funded by the European Union’s Horizon 2020 Programme. Four different strategies to mitigate the UHI effect, including the current situation, changing the low-albedo materials with high-albedo ones, nature-based solutions, and changing building façade materials, are investigated with a micro-climatic simulation tool. Then, the output of the strategies, which is potential air temperature, is used in a dynamic building energy simulation software to obtain energy consumption and thermal comfort data of the residential buildings in the case area. Finally, a multi-criteria decision-making model, using real-life criteria, such as total energy consumption, thermal comfort, capital cost, lifetime and installation flexibility, is used to make a decision for decreasing the UHI effect on residential energy performance of buildings. The results showed that applying NBSs, such as green roofs and changing existing trees with high leaf area density ones, have the highest ranking among all mitigation strategies. The output of this study may help urban planners, architects, and engineers in the decision-making processes during the design phase of urban planning.

A hybrid WOA-SVM based on CI for improving the accuracy of shopping mall air conditioning system energy consumption prediction

Energy consumption prediction is of great practical significance for the study of energy-saving scheduling of building air-conditioning systems. However, there are two problems in prediction models: (1) A single prediction model is difficult to solve the nonlinear problem of building energy consumption data. (2) A single prediction result can hardly reflect the uncertainty in the building operation process. Therefore, in this study, a hybrid prediction model with an improved whale optimization algorithm- optimization support vector machine (IWOA-SVM) is proposed based on the air conditioning system of a shopping mall in Xi'an. The WOA is improved using the opposition-based learning strategy (OBL), differential evolution algorithm (DE), and introducing a control parameter λ. The Confidence Interval (CI) of the model fitting error is used to represent the prediction results. The results show that the root mean square error (RMSE) of the optimized model is reduced by 9.5, and the (RMSE) of the total energy consumption prediction value after adding CI is >1.68. The method achieves higher model accuracy, reflecting the operational fluctuations of the building.

국공립 어린이집 그린리모델링 사업의 리트로핏 대안별 에너지 성능 및 경제성 비교 분석 : 단일 사례의 시뮬레이션 기반 분석 중심으로

Purpose: The purpose of this study was to present an optimized green remodeling retrofit plan through energy performance and economic analysis of the public daycare center for which green remodeling has been completed. Method: By collecting data on public building green remodeling project management in 2020 from the Korea Authority of Land & Infrastructure Safety, target daycare centers that are single-use and can collect monthly building energy consumption data through a public data open system were extracted. The building outline and operating conditions were investigated, and energy performance was analyzed using Design Builder, a dynamic building energy analysis program. After calibration, an alternative model analysis for each technical element was performed, including the basic base model. Then, economic analysis was performed using the net present value method. Result: The results of this study were as follows. After comprehensively analyzing energy performance and economic feasibility, ALT-4(heat recovery ventilation system), ALT-9 (heat recovery ventilation system+high-efficiency air conditioning+LED lighting replacement), ALT-6(LED lighting replacement) models were determined by the final priority. ALT-4(heat recovery ventilation system) as a single element model and ALT-9(heat recovery ventilation system+high-efficiency air conditioner+LED lighting replacement) as a combined element model were the best alternative when comprehensively considering energy performance and economic feasibility in green remodeling.

Building energy consumption prediction based on VMD-CNN-BILSTM-Attention

Accurate building energy consumption prediction can help buildings optimize energy allocation and save energy. To achieve accurate and reliable building energy consumption prediction, this paper proposes a building energy consumption prediction method based on VMD decomposition and noise reduction, i.e., CNN-BILSTM-Attention hybrid prediction. In this paper, we first use the VMD method to decompose the historical energy consumption series to obtain relatively smooth multiple components, which lays the foundation for accurate prediction. Then we combine Convolutional Neural Network (CNN) and Bi-directional Long and Short Term Memory (BILSTM) network to fully extract the Spatio-temporal characteristics of energy consumption data and then introduce the Attention mechanism to automatically assign corresponding weights to the BILSTM hidden layer states to distinguish the importance of different influencing factors on energy consumption and improve the prediction accuracy. To verify the effectiveness of the model, the most later analyzed with real building energy consumption data of a region, and the prediction results were shown by python, and the prediction results were improved by 40% compared with the traditional machine learning algorithm.

Energy Consumption Forecasting of Urban Residential Buildings in Cold Regions of China

Long-term predictions of the energy consumption of a building can be a reference in formulating energy conservation policies to achieve carbon neutrality. However, existing research on the prediction of energy consumption of urban buildings mainly adopts top-down or bottom-up single models that do not consider the coupling effect of macro and micro factors. Therefore, a coupled top-down and bottom-up prediction model has been proposed in this paper. The applicability of the proposed method was investigated based on residential buildings in Beijing, China. First, based on a top-down methodology, the energy consumption data of residential buildings in Beijing were investigated using an urban statistical yearbook, and the energy consumption of residential buildings under different energy-saving policies was analyzed. Second, micro factors, such as envelope parameters, personnel behavior, air conditioning, and electrical usage of typical residential buildings, were investigated. Subsequently, a simulation model of residential building energy consumption was constructed according to the survey data. The actual and simulated values for 2017 were 0.264 GJ/m2 and 0.252 GJ/m2, respectively. Finally, pessimistic and optimistic scenarios are proposed using the Human Impact, Population, Affluence, Technology (IPAT) model. The energy consumption of residential buildings in Beijing for the next decade was predicted under different scenarios by adopting the gray prediction method and multiple regression analysis, which was verified using the back-propagation neural network algorithm. In the baseline scenario, the projected 2021 heating energy consumption value was 13.4 kgce/m2 with a relative error of 6.52%.

Analysis and prediction of carbon emission in the large green commercial building: A case study in Dalian, China

Reducing carbon emissions from the construction industry has been vital to addressing the growing global environmental change challenge. Building energy consumption data is crucial to many applications, such as carbon emission auditing, energy efficiency improvement, etc. This study aims to mine the energy consumption patterns of commercial buildings and explore the applicability of a data-driven model in building carbon emissions prediction. This research selected Dalian's large-scale green commercial building as a case study. The electrical load data for the past four years were collected, and the indoor and outdoor environmental data were monitored under different seasons. Machine learning was used to develop building carbon emissions forecasting models. The average annual increase rate of building electricity consumption before the pandemic was 5.9%. Miscellaneous electric loads (MELs) is the largest electricity consumer in the target building. On typical days, indoor illuminance and CO2 are highly correlated under different seasons. A forecasting model based on ensemble learning is found to have certain advantages in building carbon emissions prediction.

Estimating building energy efficiency from street view imagery, aerial imagery, and land surface temperature data

Current methods to determine the energy efficiency of buildings generally require on-site visits of certified energy auditors which makes the process slow, costly, and geographically incomplete. To accelerate the identification of promising retrofit targets on a large scale, we propose to estimate building energy efficiency from widely available and remotely sensed data sources only, namely street view, aerial view, footprint, and satellite-borne land surface temperature (LST) data. After collecting data for almost 40,000 buildings in the United Kingdom, we combine these data sources by training multiple end-to-end deep learning models with the objective to classify buildings as energy efficient (EU rating A–D) or inefficient (EU rating E–G). After evaluating the trained models quantitatively as well as qualitatively, we extend our analysis by studying the predictive power of each data source in an ablation study. We find that the end-to-end deep learning model trained on all four data sources achieves a macro-averaged F1 score of 64.64% and outperforms models trained on building energy consumption data by 9.78%. As industry experts use building energy consumption data to approximate a building’s energy efficiency prior to a potential on-site inspection, this work shows the potential and complementary nature of remotely sensed data in predicting energy efficiency and opens up new opportunities for future work to integrate additional data sources.