Creating a Large-Scale National Residential Building Energy Dataset Using a Two-Stage Machine Learning Approach DOI
Sorena Vosoughkhosravi, Amirhosein Jafari

Construction Research Congress 2022, Journal Year: 2024, Volume and Issue: unknown, P. 305 - 315

Published: March 18, 2024

Buildings account for 40% of total energy demand in the US. Consequently, there is a pressing need dataset that provides comprehensive information on consumption household units The current practice large-scale simulations may not reflect actual patterns. Additionally, existing national building datasets, such as RECS, have limited number datapoint and do social aspects households. This study aimed to create residential using two-stage machine learning approach, combining two datasets RECS AHS. outcome this contains about well their detailed features. Three algorithms, including artificial neural networks (ANN), random forest (RF), gradient boosting regression (GBR), were used develop data-integration framework. results showed RF had best performance predicting end-use consumption. predicted generated an accuracy over 80%. These findings significant implications energy-efficient design operation.

Language: Английский

Predicting occupant energy consumption in different indoor layout configurations using a hybrid agent-based modeling and machine learning approach DOI
Mohammad Nyme Uddin, Minhyun Lee, Xuerong Cui

et al.

Energy and Buildings, Journal Year: 2024, Volume and Issue: unknown, P. 115102 - 115102

Published: Nov. 1, 2024

Language: Английский

Citations

2
Calibrating thermal sensation vote scales for different short-term thermal histories using ensemble learning DOI
Yuan Liang, Rong Qu,

Tianyu Chen

et al.

Building and Environment, Journal Year: 2023, Volume and Issue: 246, P. 110998 - 110998

Published: Oct. 31, 2023

Language: Английский

Citations

5
Performance comparison on improved data-driven building energy prediction under data shortage scenarios in four perspectives: data generation, incremental learning, transfer learning, and physics-informed DOI
Guannan Li, Lei Zhan, Xi Fang

et al.

Energy, Journal Year: 2024, Volume and Issue: unknown, P. 133640 - 133640

Published: Oct. 1, 2024

Language: Английский

Citations

1
An Explainable Prediction Model for Aerodynamic Noise of an Engine Turbocharger Compressor Using an Ensemble Learning and Shapley Additive Explanations Approach DOI Open Access
Rong Fung Huang,

Jimin Ni,

Pengli Qiao

et al.

Sustainability, Journal Year: 2023, Volume and Issue: 15(18), P. 13405 - 13405

Published: Sept. 7, 2023

In the fields of environment and transportation, aerodynamic noise emissions emitted from heavy-duty diesel engine turbocharger compressors are great harm to human health, which needs be addressed urgently. However, for study compressor noise, particularly at full operating range, experimental or numerical simulation methods costly long-period, do not match engineering requirements. To fill this gap, a method based on ensemble learning is proposed predict noise. study, 10,773 datasets were collected establish normalize an dataset. Four algorithms (random forest, extreme gradient boosting, categorical boosting (CatBoost) light machine) applied mapping functions between total sound pressure level (SPL) speed, mass flow rate, ratio frequency compressor. The results showed that, among four models, CatBoost model had best prediction performance with correlation coefficient root mean square error 0.984798 0.000628, respectively. addition, predicted SPL observed value was smallest, only 0.37%. Therefore, algorithm proposed. For different points compressor, high accuracy. contour cloud in MAP better characterizing variation SPL. maximum minimum SPLs 122.53 dB 115.42 dB, further interpret model, analysis conducted by applying Shapley Additive Explanation that significantly affected SPL, while rate little effect could well

Language: Английский

Citations

2
Creating a Large-Scale National Residential Building Energy Dataset Using a Two-Stage Machine Learning Approach DOI
Sorena Vosoughkhosravi, Amirhosein Jafari

Construction Research Congress 2022, Journal Year: 2024, Volume and Issue: unknown, P. 305 - 315

Published: March 18, 2024

Buildings account for 40% of total energy demand in the US. Consequently, there is a pressing need dataset that provides comprehensive information on consumption household units The current practice large-scale simulations may not reflect actual patterns. Additionally, existing national building datasets, such as RECS, have limited number datapoint and do social aspects households. This study aimed to create residential using two-stage machine learning approach, combining two datasets RECS AHS. outcome this contains about well their detailed features. Three algorithms, including artificial neural networks (ANN), random forest (RF), gradient boosting regression (GBR), were used develop data-integration framework. results showed RF had best performance predicting end-use consumption. predicted generated an accuracy over 80%. These findings significant implications energy-efficient design operation.

Language: Английский

Citations

0