Optimal Photovoltaic Panel Direction and Tilt Angle Prediction Using Stacking Ensemble Learningopen access
- Authors
- Khan, Prince Waqas; Byun, Yung-Cheol; Lee, Sang-Joon
- Issue Date
- 27-Apr-2022
- Publisher
- FRONTIERS MEDIA SA
- Keywords
- machine learning; data curation; tilt prediction; energy forecasting; direction prediction; solar panels; RE100; solar energy
- Citation
- FRONTIERS IN ENERGY RESEARCH, v.10
- Journal Title
- FRONTIERS IN ENERGY RESEARCH
- Volume
- 10
- URI
- https://scholarworks.bwise.kr/gachon/handle/2020.sw.gachon/88436
- DOI
- 10.3389/fenrg.2022.865413
- ISSN
- 2296-598X
- Abstract
- Renewable energy sources produce electricity without causing increment in pollution, and solar energy is one of the primary renewable sources. Switching to renewable electricity is particularly impactful for companies whose emissions from purchased energy are the primary source. The Renewable Energy (RE100) initiative provides awareness to governments and the general public. Therefore, organizations must now move from renewable energy sources to clean energy sources. Solar panels are the primary source of renewable energy. However, a harsh environment or solar panel malfunction can lead to missing data, which causes various problems, such as data processing complexity, severe biases, and commitment to data quality. Optimal orientation and tilt angle for solar panels effectively get more energy from the solar panels. We have used machine learning to predict the optimal angle for a solar panel according to the season and time. This article studies solar panel data's photovoltaic energy generation value and proposes a machine learning model based on the stacking ensemble learning technique. Three ML models, including catboost, XGboost, and random forest, are ensebmled. Experimental data are obtained by setting up sixteen solar panels with different combinations of tilt and direction. The performance of the proposed method is compared with other ML and statistical models. We obtained a regression score (R-2) of 0.86 and a mean absolute percentage error (MAPE) of 2.54%.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - ETC > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/gachon/handle/2020.sw.gachon/88436)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.