Modeling and forecasting building energy consumption: A review of data-driven techniques

Bourdeau, Mathieu; Zhai, Xiao qiang; Nefzaoui, Elyes; Guo, Xiao‐Feng; Chatellier, Patrice

doi:10.1016/j.scs.2019.101533

Cited by 539 publications

(291 citation statements)

References 115 publications

Supporting

Mentioning

221

Contrasting

Unclassified

Order By: Relevance

“…Here, four different supervised learning algorithms have been selected as representative of ML algorithms to model daily energy consumption: support vector regression (SVR), linear model stepwise regression (LMSR), distance weighted K-nearest neighbours (KNN) and naive bayes (NB). SVR and LMSR are regression models and they have been used in energy modelling [35,36]. NB and KNN are classification algorithms and they have been used for water demand forecasting [37,38], however few papers discuss their application in energy consumption.…”

Section: Machine-learning Modelsmentioning

confidence: 99%

Demand Forecasting for a Mixed-Use Building Using Agent-Schedule Information with a Data-Driven Model

Friedrich

Harrison

2020

Energies

View full text Add to dashboard Cite

There is great interest in data-driven modelling for the forecasting of building energy consumption while using machine learning (ML) modelling. However, little research considers classification-based ML models. This paper compares the regression and classification ML models for daily electricity and thermal load modelling in a large, mixed-use, university building. The independent feature variables of the model include outdoor temperature, historical energy consumption data sets, and several types of 'agent schedules' that provide proxy information that is based on broad classes of activity undertaken by the building's inhabitants. The case study compares four different ML models testing three different feature sets with a genetic algorithm (GA) used to optimize the feature sets for those ML models without an embedded feature selection process. The results show that the regression models perform significantly better than classification models for the prediction of electricity demand and slightly better for the prediction of heat demand. The GA feature selection improves the performance of all models and demonstrates that historical heat demand, temperature, and the 'agent schedules', which derive from large occupancy fluctuations in the building, are the main factors influencing the heat demand prediction. For electricity demand prediction, feature selection picks almost all 'agent schedule' features that are available and the historical electricity demand. Historical heat demand is not picked as a feature for electricity demand prediction by the GA feature selection and vice versa. However, the exclusion of historical heat/electricity demand from the selected features significantly reduces the performance of the demand prediction.

show abstract

Section: Machine-learning Modelsmentioning

confidence: 99%

Demand Forecasting for a Mixed-Use Building Using Agent-Schedule Information with a Data-Driven Model

Friedrich

Harrison

2020

Energies

View full text Add to dashboard Cite

show abstract

“…Our literature search found nine review papers published in 2019 alone. Runge & Zmeureanu (2019) [2] and Bourdeau et al (2019) [3] reviewed studies on using machine learning for building energy consumption forecasting. Qolomany et al (2019) [4] and Djenouri et al (2019) [5] reviewed how machine learning and big data could be applied to smart buildings.…”

Section: Introductionmentioning

confidence: 99%

“…The literature search result shown here is not mutually exclusive. A paper might be counted twice if it covers different stages of building life cycle; for instance, design and construction 3. We acknowledge that these selection criteria might be subjective to a certain degree.…”

mentioning

confidence: 99%

State-of-the-art on research and applications of machine learning in the building life cycle

Hong

Wang

Luo

et al. 2020

Energy and Buildings

232

View full text Add to dashboard Cite

Fueled by big data, powerful and affordable computing resources, and advanced algorithms, machine learning has been explored and applied to buildings research for the past decades and has demonstrated its potential to enhance building performance. This study systematically surveyed how machine learning has been applied at different stages of building life cycle. By conducting a literature search on the Web of Knowledge platform, we found 9579 papers in this field and selected 153 papers for an in-depth review. The number of published papers is increasing year by year, with a focus on building design, operation, and control. However, no study was found using machine learning in building commissioning. There are successful pilot studies on fault detection and diagnosis of HVAC equipment and systems, load prediction, energy baseline estimate, load shape clustering, occupancy prediction, and learning occupant behaviors and energy use patterns. None of the existing studies were adopted broadly by the building industry, due to common challenges including (1) lack of large scale labeled data to train and validate the model, (2) lack of model transferability, which limits a model trained with one data-rich building to be used in another building with limited data, (3) lack of strong justification of costs and benefits of deploying machine learning, and (4) the performance might not be reliable and robust for the stated goals, as the method might work for some buildings but could not be generalized to others. Findings from the study can inform future machine learning research to improve occupant comfort, energy efficiency, demand flexibility, and resilience of buildings, as well as to inspire young researchers in the field to explore multidisciplinary approaches that integrate building science, computing science, data science, and social science.

show abstract

“…It means that any buildings in such countries are required to increase energy efficiency when using heat pumps and thermal storage. In this regard, 4th generation DH integrating smart thermal grids [12], thermal energy storage strategies for solar heating systems, and data-driven models for building scale applications are suggested [13].This has made the conflict even more profound on the comparative advantages of energy efficiency between business operators of the CHP-DH system and the SHP system [14][15][16]. Accordingly, several comparative studies have been published claiming comparative advantage of energy efficiency between the CHP-DH and SHP systems [17,18].…”

mentioning

confidence: 99%

mentioning

confidence: 99%

Energy Efficiency for Supplier and Sustainability for Demand: A Case of Heating Systems in South Korea

et al. 2019

View full text Add to dashboard Cite

This paper addresses an analysis of a recent debate in South Korea on energy efficiencies between a combined heat and power system with district heating (CHP-DH) and separate heat and power system (SHP) by applying a scenario analysis technique. In accordance with the Low Carbon Green Growth strategy where the rise of social awareness for energy savings is imminent, various scenarios were considered including CHP-DH (excluding waste heat recovery) versus SHP and CHP-DH (including waste heat recovery) versus SHP. To analyze the satisfaction on energy efficiency of each heating system, nationwide average field data were acquired via a survey with the help of the Korea District Heating Corporation (KDHC), and a comprehensive model for measuring sustainable value related to energy consumption, including economic (ECON), non-economic (NECON), and environmental (ENV) values were presented based on customer satisfaction (SAT). As a result of the structural equation model (SEM) and ordinary least squares (OLS) regression, variations of energy efficiencies of each heat and power delivery system are rather distinctive among different scenarios due to the performance characteristics of each system in the field. The energy savings of CHP-DH systems owned by KDHC, based on nationwide average performance, was found to be higher than the SHP system by 3.25% to 18.1%. While the effect of NECON and ENV on SAT was not significant in SHP, it was significant in CHP-DH. Based on these results, rational decisions can be made to lead the government and energy consumers to maximize energy efficiency, especially in South Korea, where the energy market heavily relies on foreign countries.Sustainability 2019, 11, 4216 2 of 21 treaty in Kyoto, parties to the UNFCCC met at conferences to discuss how to achieve the treaty's aims. At the 3rd Conference of the Parties (COP-3), the parties paved the way to reduce greenhouse gas emission effectively and support sustainable development of developing countries by defining three "flexibility mechanisms" which are International Emissions Trading (IET), the Clean Development Mechanism (CDM), and Joint Implementation (JI); the Kyoto Protocol [2].Although South Korea is not bound by the U.N.'s Kyoto Protocol as a first developing (non-Annex I) country, the Korean government voluntarily set a 2020 emissions reduction target to head off an estimated 30 percent rise in emissions compared to the 2020 Business-as-Usual (BAU) emission projection, considering increased national competitiveness and expectations from international society [3]. This goal reaches the highest level of recommendations from the IPCC, showing the strong commitment of the Korean government to join the global movement for coping with climate change. However, the heavy-duty climate change policy of Korean government re-ignited the debate over the effectiveness between combined heat and power systems with district heating (CHP-DH) and separate heat and power systems (SHP), which is a deep-rooted debate in the domestic heating energy indus...

show abstract

Modeling and forecasting building energy consumption: A review of data-driven techniques

Cited by 539 publications

References 115 publications

Demand Forecasting for a Mixed-Use Building Using Agent-Schedule Information with a Data-Driven Model

Demand Forecasting for a Mixed-Use Building Using Agent-Schedule Information with a Data-Driven Model

State-of-the-art on research and applications of machine learning in the building life cycle

Energy Efficiency for Supplier and Sustainability for Demand: A Case of Heating Systems in South Korea

Contact Info

Product

Resources

About