Promoting solar energy utilization: Prediction, analysis and evaluation of solar radiation on building surfaces at city scale
Release time:2024-12-11
Hits:
- Affiliation of Author(s):
- 建筑学院
- Journal:
- ENERGY AND BUILDINGS
- Abstract:
- Evaluating the solar potential in urban areas is crucial for the low-carbon transition of city energy systems. However, the complex urban environment presents challenges for the accurate and efficient prediction of solar radiation on building surfaces at the city scale. To address this challenge, this study integrates Geographic In formation System (GIS), Ladybug Tools (LBT), and Machine Learning (ML) to establish a model for predicting solar radiation on building surfaces at city scale. Using Zhengzhou, a typical city in China, as a case study, the model’s accuracy and efficiency were validated using the coefficient of determination (R 2 ), root mean square error (RMSE), and time taken as indicators. The SHapley value was then employed to identify key features influencing solar radiation on building surfaces and their impact trends. Further analysis and evaluation were conducted on the distribution and variation patterns of solar radiation across different building surfaces, building types, radiation types, and temporal dimensions. The results indicate that: (1) Among 17 common machine learning models, XGBoost performed the best. After hyperparameter optimization, it achieved an R 2 of 0.969 and an RMSE of 627,453.36, demonstrating satisfactory predictive performance on unseen datasets. Key factors affecting solar radiation include geometric features of building clusters, temporal parameters, and radiation types. (2) Roofs account for 27 % of the total building surface area but receive 54 % of the total solar radiation, with an annual reception of 64.66 TWh, making them a priority for utilization. Residential buildings, with the largest building surface area (1.61 × 10 8 m 2 ) and annual solar radiation reception (87.66 TWh) among the ten building types, are a focal point for solar radiation utilization. (3) Utilizing all building surfaces in Zhengzhou could generate 19.04 TWh of photovoltaic power, meeting over one-third of the city’s annual electricity demand. The proposed framework and findings enhance the utilization of solar radiation on urban building surfaces.
- Co-author:
- 秦国晋
- First Author:
- yanzengfeng
- Indexed by:
- Journal paper
- Correspondence Author:
- 倪平安,雷馥铭
- Volume:
- 319
- ISSN No.:
- 0378-7788
- Translation or Not:
- no
- Date of Publication:
- 2024-09-15