Database and image preprocessing for solar energy forecasting
Published in Senior Project 499 at Chulalongkorn University by Kanawut Suwandee and Kongpob In-odd, 2024
The study involves training and evaluating three deep learning models SUNSET, Unet, and SolarNet for forecasting solar irradiance up to 15 future timestamps (with 1–2 minute resolution), followed by model improvement. The preprocessing pipeline includes fisheye distortion correction and region of interest (RoI) extraction from sky images, followed by cloud mask generation.
Authors:
- Kanawut Suwandee
- Kongpob In-odd
You can cite their work with the following bibtex
@techreport{suwandee2026solar,
author = {Kanawut Suwandee and Kongpob In-odd},
title = {Database and image preprocessing for solar energy forecasting},
institution = {Chulalongkorn University},
year = {2024},
type = {Senior Project Report},
note = {Advisor: Dr. Suwichaya Suwanwimolkul}
}
or
Suwandee, Kanawut, and Kongpob In-odd. Database and Image Preprocessing for Solar Energy Forecasting. Advisor: Dr. Suwichaya Suwanwimolkul. Chulalongkorn University, 2024. Senior Project Report.
Year: SeniorY2024