Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Building thermal load prediction through shallow machine learning and deep learning

Journal Article · · Applied Energy

Building thermal load prediction informs the optimization of cooling plant and thermal energy storage. Physics-based prediction models of building thermal load are constrained by the model and input complexity. In this study, we developed 12 data-driven models (7 shallow learning, 2 deep learning, and 3 heuristic methods) to predict building thermal load and compared shallow machine learning and deep learning. The 12 prediction models were compared with the measured cooling demand. It was found XGBoost (Extreme Gradient Boost) and LSTM (Long Short Term Memory) provided the most accurate load prediction in the shallow and deep learning category, and both outperformed the best baseline model, which uses the previous day's data for prediction. Then, we discussed how the prediction horizon and input uncertainty would influence the load prediction accuracy. Major conclusions are twofold: first, LSTM performs well in short-term prediction (1 h ahead) but not in long term prediction (24 h ahead), because the sequential information becomes less relevant and accordingly not so useful when the prediction horizon is long. Second, the presence of weather forecast uncertainty deteriorates XGBoost's accuracy and favors LSTM, because the sequential information makes the model more robust to input uncertainty. Training the model with the uncertain rather than accurate weather data could enhance the model's robustness. Finally, our findings have two implications for practice. First, LSTM is recommended for short-term load prediction given that weather forecast uncertainty is unavoidable. Second, XGBoost is recommended for long term prediction, and the model should be trained with the presence of input uncertainty.

Research Organization:
Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
Sponsoring Organization:
USDOE Office of Energy Efficiency and Renewable Energy (EERE), Building Technologies Office (EE-5B)
Grant/Contract Number:
AC02-05CH11231
OSTI ID:
1607416
Alternate ID(s):
OSTI ID: 1600838
Journal Information:
Applied Energy, Journal Name: Applied Energy Journal Issue: C Vol. 263; ISSN 0306-2619
Publisher:
ElsevierCopyright Statement
Country of Publication:
United States
Language:
English

References (33)

Cooling-load prediction by the combination of rough set theory and an artificial neural-network based on data-fusion technique journal September 2006
Applying support vector machine to predict hourly cooling load in the building journal October 2009
Cooling output optimization of an air handling unit journal March 2010
An interactive building power demand management strategy for facilitating smart grid optimization journal March 2014
Development of prediction models for next-day building energy consumption and peak power demand using data mining techniques journal August 2014
A short-term building cooling load prediction method using deep learning algorithms journal June 2017
Data analytics and optimization of an ice-based energy storage system for commercial buildings journal October 2017
Machine learning-based thermal response time ahead energy demand prediction for building heating systems journal July 2018
Practical factors of envelope model setup and their effects on the performance of model predictive control for building heating, ventilating, and air conditioning systems journal February 2019
Prediction of occupancy level and energy consumption in office building using blind system identification and neural networks journal April 2019
Data fusion in predicting internal heat gains for office buildings through a deep learning approach journal April 2019
Forecasting district-scale energy dynamics through integrating building network and long short-term memory learning algorithm journal August 2019
Review of energy conservation technologies for fresh air supply in zero energy buildings journal February 2019
Modeling and measurement study on an intermittent heating system of a residence in Cambridgeshire journal October 2015
Load prediction method for heat and electricity demand in buildings for the purpose of planning for mixed energy distribution systems journal January 2008
Predicting future hourly residential electrical consumption: A machine learning case study journal June 2012
EKF based self-adaptive thermal model for a passive house journal January 2014
Modeling heating and cooling loads by artificial intelligence for energy-efficient building design journal October 2014
Short-term load forecasting in a non-residential building contrasting models and attributes journal April 2015
Development and validation of grey-box models for forecasting the thermal response of occupied buildings journal April 2016
Predicting plug loads with occupant count data through a deep learning approach journal August 2019
Simplified building model for transient thermal performance estimation using GA-based parameter identification journal April 2006
A review on the prediction of building energy consumption journal August 2012
Review of building energy modeling for control and operation journal September 2014
Heating and cooling energy trends and drivers in buildings journal January 2015
A review of data-driven building energy consumption prediction studies journal January 2018
An Inverse Gray-Box Model for Transient Building Load Prediction journal January 2002
XGBoost: A Scalable Tree Boosting System conference January 2016
Long Short-Term Memory journal November 1997
The building performance gap: Are modellers literate? journal January 2017
Asymptotic Theory of Certain "Goodness of Fit" Criteria Based on Stochastic Processes journal June 1952
Parallel Support Vector Machines Applied to the Prediction of Multiple Buildings Energy Consumption journal June 2010
An Analysis of Variance Test for Normality (Complete Samples) journal December 1965

Similar Records

Wattile: Probabilistic Deep Learning-based Forecasting of Building Energy Consumption [SWR-20-94]
Software · Tue Aug 25 20:00:00 EDT 2020 · OSTI ID:code-103525

Building thermal dynamics modeling with deep transfer learning using a large residential smart thermostat dataset
Journal Article · Tue Dec 19 23:00:00 EST 2023 · Engineering Applications of Artificial Intelligence · OSTI ID:2377667

Data-Driven Day-Ahead PV Estimation Using Autoencoder-LSTM and Persistence Model
Journal Article · Tue Sep 22 00:00:00 EDT 2020 · IEEE Transactions on Industry Applications · OSTI ID:1810249