Energy and Buildings最新文献

{"title":"Good practices for documenting AI-based studies on energy and buildings","authors":"Tianzhen Hong,&nbsp;Han Li","doi":"10.1016/j.enbuild.2026.117043","DOIUrl":"10.1016/j.enbuild.2026.117043","url":null,"abstract":"<div><div>Artificial intelligence has transformed building science research over the past decade, with applications spanning energy modeling, energy prediction, HVAC optimization and controls, fault detection, and occupancy modeling. However, many studies lack adequate documentation of datasets, algorithms, training procedures, and validation methods. Building science research faces additional challenges including inconsistent evaluation metrics, limited generalizability across building types, climates, and significant gaps between experimental studies and deployed systems. This communication provides practical guidance for good practices in documenting and publishing AI-based research following established standards from the computer science and machine learning communities. By adopting frameworks such as Datasheets for Datasets, Model Cards, and standardized reproducibility checklists, researchers can ensure their work meets the rigorous documentation standards necessary for reproducible, comparable, and impactful building science research.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 117043"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146014897","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A machine learning based rapid thermal performance modeling method for modular buildings with BIPV: A novel decomposition strategy with real-time prediction capabilities","authors":"Yiqian Zheng ,&nbsp;Biao Yang ,&nbsp;Miaomiao Hou ,&nbsp;Yi Zhang ,&nbsp;Yuekuan Zhou ,&nbsp;Xing Zheng ,&nbsp;Pengyuan Shen","doi":"10.1016/j.enbuild.2026.117063","DOIUrl":"10.1016/j.enbuild.2026.117063","url":null,"abstract":"<div><div>The global push for carbon neutrality has intensified the need for rapid and accurate energy prediction methods for BIPV-integrated modular buildings. Traditional physics-based simulation approaches suffer from excessive computational burden. This study presents a novel machine learning-based rapid energy prediction methodology specifically designed for modular buildings with building-integrated photovoltaics. A comprehensive feature engineering framework captures the unique thermal and geometric characteristics of modular construction through six-surface property encoding, geometric parameters, and solar irradiance calculations. The methodology employs a modular building decomposition strategy that enables individual module analysis while maintaining system-level accuracy. An XGBoost-based prediction model achieves superior performance across four representative climate zones. The model achieves R² values exceeding 0.93 for heating loads, cooling loads, and total energy consumption. Experimental validation using a real-world BIPV-integrated modular building demonstrates prediction accuracy within industry-acceptable limits, with mean absolute errors below 1.5°C. The computational efficiency assessment shows prediction speeds over 2,000 × faster than traditional simulation approaches, enabling real-time design iteration. Successful integration with Grasshopper parametric design platforms facilitates immediate energy feedback during conceptual design phases. This advancement removes computational barriers to energy performance optimization and supports the broader adoption of sustainable modular construction practices by providing practical tools for energy-informed design decision-making.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 117063"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146048122","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Adaptive thermostat preference learning using behaviour nudging and multi-armed bandits: A field implementation","authors":"Hussein Elehwany ,&nbsp;Andre Markus ,&nbsp;Burak Gunay ,&nbsp;Mohamed Ouf ,&nbsp;Nunzio Cotrufo ,&nbsp;Jean-Simon Venne ,&nbsp;Junfeng Wen","doi":"10.1016/j.enbuild.2026.117030","DOIUrl":"10.1016/j.enbuild.2026.117030","url":null,"abstract":"<div><div>Occupant behaviour (OB) centric controls have significant potential in advancing next-generation HVAC systems. Many OB-centric control studies solicit feedback from occupants to tackle the thermal preference learning problem. Behaviour nudging was also implemented in various systems to influence occupant behaviour to be more energy efficient. This study addresses the gap of using behaviour nudging and unsolicited occupant thermostat overrides to learn their thermal preferences. A multi-armed bandit (MAB) reinforcement learning (RL) was used to learn occupant thermal preferences from their thermostat interactions. The reward signal of the algorithm was designed to reward energy savings and penalize discomfort. The occupants were continuously nudged by slowly reducing the zone setpoint during the heating season, to encourage them to override the thermostats. The algorithm was implemented in two zones with multiple occupants in an academic facility in Ottawa, Canada, achieving energy savings of up to 12.7% compared to static setpoints.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 117030"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146001008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Surrogate model evaluation and building energy benchmarking for commercial buildings","authors":"Venkatesh Chinde ,&nbsp;Rohit Chintala,&nbsp;Janghyun Kim,&nbsp;Alex Chapin,&nbsp;Jie Xiong,&nbsp;Katherine Fleming,&nbsp;Brian L. Ball","doi":"10.1016/j.enbuild.2026.117033","DOIUrl":"10.1016/j.enbuild.2026.117033","url":null,"abstract":"<div><div>Building energy consumption benchmarking involves challenges associated with various energy patterns for different building types; heating, ventilating, and air-conditioning (HVAC) system types; and climates. Given significant variation in energy use patterns, accurate prediction of long-term energy use using surrogate models remains challenging. Multiple linear regression (MLR) is commonly used for building energy benchmarking because of its simple structure; however, it lacks accuracy compared to other black-box models. Although many studies have compared surrogate models and offer guidance on model selection based on metrics, they do not provide detailed analysis on improving the surrogate model accuracy. In this paper, we implement a surrogate model using polynomial ridge regression (i.e., MLR with interaction terms combined with ridge regularization) for small office and retail strip mall buildings across six HVAC system types and all climate zones, for electricity and natural gas in baseline and proposed scenarios. A simulation workflow is developed using OpenStudio™/EnergyPlus™ to generate simulation data using measures over a wide range of efficiency inputs. Enhancements based on statistical insights are used for improving the model accuracy using filters, input transformations, and change points. Surrogate models achieved average coefficient of variation of the root mean squared error (CVRMSE) values of 2.17, 1.06, 2.05, and 3.26 for proposed electricity, proposed natural gas, baseline electricity, and baseline natural gas, respectively, with enhancements reducing CVRMSE by an average of 14.9% across all combinations. We provide model interpretation via Shapley additive explanations to determine which input variables most influence energy consumption and provide supportive arguments for enhancements.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 117033"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146048181","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Effect of local heating on thermal responses of cleaning staff when working in an unheated space in winter and comparisons with the college students","authors":"Yicheng Ren ,&nbsp;Yuxin Wu ,&nbsp;Yujie Zhou ,&nbsp;Yuting Li ,&nbsp;Shuang Zheng ,&nbsp;Yonghong Wu","doi":"10.1016/j.enbuild.2026.117039","DOIUrl":"10.1016/j.enbuild.2026.117039","url":null,"abstract":"<div><div>The cleaning staff plays a crucial role in maintaining the good condition of buildings. In southern China, they were frequently exposed to non-heated cold environments in winter, where a personal comfort system is needed. However, the study about their thermal comfort is insufficient. This study seeks to examine the impact of cold stress and local heating on the thermal responses of cleaning staff and college students during cleaning work under cold winter conditions. Thirty-two participants (16 cleaning staff and 16 college students) were recruited to perform cleaning activities in a semi-open corridor (9.2 °C, 52.5% RH) in winter (outside temperature: 5.4 °C). Three heating cases (head, hands, and feet heating modes) using heating sheets with adjustable power levels (max to 5.5 W / 2 sheets) were tested to compare with the no heating case. Each case lasted for 60 min and included three 20-minute periods: windless cleaning activities (initial period), windless rest (second period), and wind-exposed cleaning activities (third period). The results indicated that the cleaning staff group felt satisfied with the cold environments under the no heating case and insensitive to local heating. Head heating was unwanted in short-term (within 20 min) cold exposure because of the higher sensitivity of the head, and overheating at the head causes thermoregulation disorder. For students, feet heating resulted in significantly lower thermal sensation vote (TSV) compared to head or hands heating during the wind-exposed cleaning activities period (p &lt; 0.01). For cleaning staff, feet heating was found to have no significant influence on overall TSV, while hands heating maintained the lowest blood pressure throughout the experiment. Hands heating consistently resulted in the highest average thermal pleasure throughout the experiment, with mean values of 0.47 for students and 0.97 for cleaning staff. Thus, hands heating with gloves was recommended for local heating of cleaning staff in cold, windy environments.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 117039"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146014894","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Surrogate models for evaluating HVAC retrofit options in multi-unit residential buildings","authors":"Harry Vallianos,&nbsp;Costa Kapsis","doi":"10.1016/j.enbuild.2026.117060","DOIUrl":"10.1016/j.enbuild.2026.117060","url":null,"abstract":"<div><div>This study investigates the use of surrogate modeling to optimize retrofit strategies in multi-unit residential buildings (MURBs), including HVAC systems. A comprehensive synthetic dataset was generated using EnergyPlus simulations, parameterized across a wide range of building and system variables, including ten distinct HVAC configurations. Multiple surrogate modeling approaches were evaluated, including single-output and multi-output artificial neural networks (ANNs) as well as Light Gradient Boosting Machine (LGBM) models. The models were trained to predict key performance metrics: Energy Use Intensity (EUI), Thermal Energy Demand Intensity (TEDI), and Cooling Energy Demand Intensity (CEDI). Results show that multi-output ANN models, with HVAC system as a categorical input, achieved high accuracy (R² &gt; 0.997 and RMSE&lt;2.2kWh/m²/year) and superior generalization compared to both single-output ANNs and LGBM models, while also reducing computational effort. The findings underscore the effectiveness of ANN-based surrogate models for rapid and accurate evaluation of retrofit scenarios involving diverse HVAC systems, supporting more efficient decision-making in building energy retrofits.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 117060"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146048127","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Indoor thermal comfort and energy–saving opportunities in university classrooms: a field study across heating and cooling seasons","authors":"Qiong He,&nbsp;Lu Han,&nbsp;Yayun Gan","doi":"10.1016/j.enbuild.2026.117005","DOIUrl":"10.1016/j.enbuild.2026.117005","url":null,"abstract":"<div><div>Classroom thermal comfort is often compromised by a lack of localized data, leading to excessive cooling in summer and overheating in winter, resulting in huge energy waste. To meet students’ real thermal comfort needs and achieve the goal of energy saving in university classrooms, this study monitored the indoor thermal parameters in 35 different types of university classrooms and collected 1618 valid questionnaires in energy-consuming seasons in Nanjing with hot summer and cold winter climate. Key findings are as follows:(1) 77.8% and 12% of classrooms have above 60% humidity in summer and winter but the percentage of votes on “normal humidity” in these investigated classrooms is 67% and 68% in the same seasons, implying that most students generally prefer a more humid indoor environment in the area. (2) The gaps between <span><math><mrow><msub><mi>T</mi><mi>n</mi></msub></mrow></math></span> and <span><math><mrow><msub><mi>T</mi><mi>p</mi></msub></mrow></math></span> in summer and winter are 2.62°C and 2.0556°C, respectively, indicating that increase and decrease in classroom temperature can still ensure thermal comfort during summer and winter, respectively (3) Applying thermal comfort parameters <span><math><mrow><msub><mi>T</mi><mi>p</mi></msub></mrow></math></span>, <span><math><mrow><msub><mi>T</mi><mi>n</mi></msub></mrow></math></span>, <span><math><mrow><msubsup><mi>T</mi><mrow><mi>A</mi></mrow><mi>s</mi></msubsup></mrow></math></span> , <span><math><mrow><msubsup><mi>T</mi><mrow><mi>A</mi></mrow><mi>a</mi></msubsup></mrow></math></span> to maintain indoor environments can significantly reduce energy consumption based on values (19°C-25°C in winter and 22°C-26°C in summer) suggested by standards: the maximum energy savings can reach 4.5%, 13.7%, 15% and 18.1% in summer but up to 24.6%, 31.4%, 33.5% and 33.7% in winter, respectively. Thus, maintaining indoor thermal comfort according to real local demands rather than general specifications has great potential to save energy in university classrooms in Nanjing.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 117005"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145995553","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Two-stage MLP-lookup table model for predicting heat pump power in greenhouses","authors":"Eun Jung Choi ,&nbsp;Doyun Lee ,&nbsp;Sang Min Lee ,&nbsp;Sungil Lim","doi":"10.1016/j.enbuild.2026.117027","DOIUrl":"10.1016/j.enbuild.2026.117027","url":null,"abstract":"<div><div>Energy costs account for a significant proportion of greenhouse operating expenses; thus, high-fidelity predictive tools are increasingly important for optimizing energy consumption. Although machine learning models demonstrate high accuracy within training ranges, their applicability to diverse operational conditions remains limited. This study developed and compared two prediction approaches: a two-stage multilayer perceptron-lookup table (MLP-LUT) model and a standalone multilayer perceptron (s-MLP) model for forecasting electrical heat pumps (EHP) energy consumption. The MLP-LUT model first predicts greenhouse temperature and humidity and then estimates power consumption through manufacturer performance mapping, whereas the s-MLP model directly predicts consumption. Bayesian optimization was used for hyperparameter tuning.</div><div>The robust generalization performance of both models underwent evaluation across diverse operating conditions, including variations in the setpoint temperature, location, control strategies, and equipment model. At baseline, both models achieved comparable accuracy, with <span><math><mrow><mi>C</mi><mi>V</mi><msub><mrow><mfenced><mrow><mi>R</mi><mi>M</mi><mi>S</mi><mi>E</mi></mrow></mfenced></mrow><mrow><mi>h</mi><mi>r</mi></mrow></msub></mrow></math></span> values of 4.05% and 4.04%, respectively, satisfying the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) criteria. For generalization testing across various setpoint temperatures, locations, and control strategies, both models retained stable prediction accuracy. However, when the EHP units are replaced, the s-MLP model exhibits severe degradation, whereas the MLP-LUT model maintains <span><math><mrow><mi>C</mi><mi>V</mi><msub><mrow><mfenced><mrow><mi>R</mi><mi>M</mi><mi>S</mi><mi>E</mi></mrow></mfenced></mrow><mrow><mi>h</mi><mi>r</mi></mrow></msub></mrow></math></span> of less than 4.0%. The MLP-LUT framework offers resilience to hardware substitution by separating environmental predictions from equipment-specific performance mapping. In contrast, the s-MLP approach is constrained to static configurations. The present work establishes practical guidelines for the system selection and provides a foundation for the development of optimal greenhouse control strategies.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 117027"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145995556","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimal ventilation systems via MILP: Duct sizing, fan placement, control strategies","authors":"Julius H.P. Breuer,&nbsp;Peter F. Pelz","doi":"10.1016/j.enbuild.2026.117016","DOIUrl":"10.1016/j.enbuild.2026.117016","url":null,"abstract":"<div><div>Ventilation systems in buildings account for a substantial share of overall energy consumption, with fans representing one of the largest contributors. Improving energy efficiency requires considering the interaction of system components. Novel topologies, such as distributed fans integrated into the central duct network, offer promising potential for efficiency gains. At the same time, building owners demand cost-effective solutions, which depend strongly on a well-designed duct network. Meeting these requirements calls for a life-cycle-oriented planning approach with integrated component selection and duct sizing. Existing planning algorithms, however, have several limitations: they often assume single load cases, rely on overly simplified fan models, neglect novel, distributed topologies, and lack guarantees of global optimality. This paper addresses these shortcomings by presenting a novel optimisation problem formulation that jointly considers topological decisions (e.g., fan and volume flow controller placement, duct sizing) and system operation under multiple load cases. The methodology enables systematic comparison of control strategies, duct limitations – in velocity and height – and analysis of cost-energy trade-offs. To reduce computation times, the non-linear optimisation problem is relaxed to a Mixed-Integer Linear Program (MILP), with proven error bounds that quantify the distance to the global optimum. The methodology is demonstrated on a case study building, showing 14 % reduced LCC compared to the existing system. Six different central or distributed control strategies and duct constraints are optimised within seconds of computation time. This makes the method suitable for practical planning processes, providing transparent decision support, e.g. through Pareto front analyses.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 117016"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145995564","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Adoption drivers and barriers of Building Information Modelling (BIM) in Europe","authors":"Martin Burgess ,&nbsp;Charlie Wilson ,&nbsp;Yee Van Fan","doi":"10.1016/j.enbuild.2026.116953","DOIUrl":"10.1016/j.enbuild.2026.116953","url":null,"abstract":"<div><div>Building Information Modelling (BIM) enables time, cost, and materials savings in building design and construction. However, the promise of BIM is yet to be realised. We assessed the current state-of-the-art in BIM adoption and use, identifying both barriers and opportunities across six dimensions defined by the PESTLE framework (political, economic, social, technical, legal, environmental). We combined market survey, literature review, and new insights from 41 expert interviews with architects, consultants and constructors across 11 European countries. We find BIM is used principally by larger firms as a design and data-processing tool to enable collaboration between project partners. BIM’s value proposition is primarily to streamline construction processes not improve resource efficiencies. Barriers to BIM adoption include interoperability issues, split incentives and value chain fragmentation, and weak economic incentives particularly for small firms. In the medium-term we find two important drivers of more widespread BIM adoption. First, institutional investors in the commercial buildings sector are increasingly pushing green certification standards for which compliance is demonstrated by BIM. Second, whole building lifecycle emission regulations for buildings mandated by the EU from 2030 will require BIM calculations. Four pioneer Northern European countries already have similar emission limits in place. Under realistic assumptions, we estimate material savings enabled by BIM in new building construction could deliver 21–31% embodied emission reductions after 10 years.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"355 ","pages":"Article 116953"},"PeriodicalIF":7.1,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146075624","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}