Data-driven optimization for mitigating energy consumption and GHG emissions in buildings
To cope with rapid urbanization and achieve urban sustainable development, both energy efficiency and GHG emissions in the building sector are considered as the main challenges in recent years. Multi-objective optimization will be a useful tool in energy saving and low carbon for town planning polic...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/179170 |
| _version_ | 1811696508630728704 |
|---|---|
| author | Zhang, Yan Teoh, Bak Koon Zhang, Limao |
| author2 | School of Civil and Environmental Engineering |
| author_facet | School of Civil and Environmental Engineering Zhang, Yan Teoh, Bak Koon Zhang, Limao |
| author_sort | Zhang, Yan |
| collection | NTU |
| description | To cope with rapid urbanization and achieve urban sustainable development, both energy efficiency and GHG emissions in the building sector are considered as the main challenges in recent years. Multi-objective optimization will be a useful tool in energy saving and low carbon for town planning policy making. This study incorporates geographical weighting (GW) with the Light Gradient Boosting Machine (LGBM), namely the GW-LGBM method, to analyze the impact of the surrounding environment on building energy performance. Besides, a genetic algorithm-based approach is applied in this research to achieve a multi-objective optimization solution for buildings' energy performance and GHG emissions. A Pareto front of the optimal trade-off solution with different influential variables and multi-objectives can be determined. Several scenarios incorporating various percentages of constraints are performed, aiming to provide more strategies for decision-makers under different situations. The main findings are summarized as: (1) The proposed GW-LGBM shows superior predictability for assessing the buildings' energy performance than the traditional LGBM. The value of the indices R2 is 0.91 in site EUWN (weather normalized energy use) and 0.90 in GHG emissions, which have respectively 6.14% and 9.22% improvement compared with the LGBM; (2) Four common factors, including the Natural gas, total Gross floor area, Energy star score, and shape form, are identified as the most important factors for both site EUWN and GHG emissions; (3) The change of the three influence factors, such as the natural gas, vertical to horizontal ratio, and greenery density, is expected to achieve a 37.77% improvement for mitigating energy consumption and GHG emissions given a 10% change in the adjustable factors. The novelty lies in the development of GW-LGBM by adding geographical weight to learning energy patterns for achieving more accurate results in building performance estimation and optimization. |
| first_indexed | 2024-10-01T07:40:29Z |
| format | Journal Article |
| id | ntu-10356/179170 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T07:40:29Z |
| publishDate | 2024 |
| record_format | dspace |
| spelling | ntu-10356/1791702024-07-22T05:33:33Z Data-driven optimization for mitigating energy consumption and GHG emissions in buildings Zhang, Yan Teoh, Bak Koon Zhang, Limao School of Civil and Environmental Engineering Engineering Machine learning Building energy performance To cope with rapid urbanization and achieve urban sustainable development, both energy efficiency and GHG emissions in the building sector are considered as the main challenges in recent years. Multi-objective optimization will be a useful tool in energy saving and low carbon for town planning policy making. This study incorporates geographical weighting (GW) with the Light Gradient Boosting Machine (LGBM), namely the GW-LGBM method, to analyze the impact of the surrounding environment on building energy performance. Besides, a genetic algorithm-based approach is applied in this research to achieve a multi-objective optimization solution for buildings' energy performance and GHG emissions. A Pareto front of the optimal trade-off solution with different influential variables and multi-objectives can be determined. Several scenarios incorporating various percentages of constraints are performed, aiming to provide more strategies for decision-makers under different situations. The main findings are summarized as: (1) The proposed GW-LGBM shows superior predictability for assessing the buildings' energy performance than the traditional LGBM. The value of the indices R2 is 0.91 in site EUWN (weather normalized energy use) and 0.90 in GHG emissions, which have respectively 6.14% and 9.22% improvement compared with the LGBM; (2) Four common factors, including the Natural gas, total Gross floor area, Energy star score, and shape form, are identified as the most important factors for both site EUWN and GHG emissions; (3) The change of the three influence factors, such as the natural gas, vertical to horizontal ratio, and greenery density, is expected to achieve a 37.77% improvement for mitigating energy consumption and GHG emissions given a 10% change in the adjustable factors. The novelty lies in the development of GW-LGBM by adding geographical weight to learning energy patterns for achieving more accurate results in building performance estimation and optimization. 2024-07-22T05:33:33Z 2024-07-22T05:33:33Z 2024 Journal Article Zhang, Y., Teoh, B. K. & Zhang, L. (2024). Data-driven optimization for mitigating energy consumption and GHG emissions in buildings. Environmental Impact Assessment Review, 107, 107571-. https://dx.doi.org/10.1016/j.eiar.2024.107571 0195-9255 https://hdl.handle.net/10356/179170 10.1016/j.eiar.2024.107571 2-s2.0-85195321033 107 107571 en Environmental Impact Assessment Review © 2024 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies. |
| spellingShingle | Engineering Machine learning Building energy performance Zhang, Yan Teoh, Bak Koon Zhang, Limao Data-driven optimization for mitigating energy consumption and GHG emissions in buildings |
| title | Data-driven optimization for mitigating energy consumption and GHG emissions in buildings |
| title_full | Data-driven optimization for mitigating energy consumption and GHG emissions in buildings |
| title_fullStr | Data-driven optimization for mitigating energy consumption and GHG emissions in buildings |
| title_full_unstemmed | Data-driven optimization for mitigating energy consumption and GHG emissions in buildings |
| title_short | Data-driven optimization for mitigating energy consumption and GHG emissions in buildings |
| title_sort | data driven optimization for mitigating energy consumption and ghg emissions in buildings |
| topic | Engineering Machine learning Building energy performance |
| url | https://hdl.handle.net/10356/179170 |
| work_keys_str_mv | AT zhangyan datadrivenoptimizationformitigatingenergyconsumptionandghgemissionsinbuildings AT teohbakkoon datadrivenoptimizationformitigatingenergyconsumptionandghgemissionsinbuildings AT zhanglimao datadrivenoptimizationformitigatingenergyconsumptionandghgemissionsinbuildings |