Transfer Learning for Photovoltaic Power Forecasting Across Regions Using Large-Scale Datasets

Accurate photovoltaic (PV) power forecasting enables stable grid operation; however, acquiring sufficient and diverse data remains challenging. Although numerous deep learning models have been employed, most rely on proprietary datasets or data spanning only a few years, limiting the generalizabilit...

Full description

Saved in:
Bibliographic Details
Main Authors: Seongho Bak, Sowon Choi, Donguk Yang, Doyoon Kim, Heeseon Rho, Kyoobin Lee
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11086606/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849391584563429376
author Seongho Bak
Sowon Choi
Donguk Yang
Doyoon Kim
Heeseon Rho
Kyoobin Lee
author_facet Seongho Bak
Sowon Choi
Donguk Yang
Doyoon Kim
Heeseon Rho
Kyoobin Lee
author_sort Seongho Bak
collection DOAJ
description Accurate photovoltaic (PV) power forecasting enables stable grid operation; however, acquiring sufficient and diverse data remains challenging. Although numerous deep learning models have been employed, most rely on proprietary datasets or data spanning only a few years, limiting the generalizability of forecasting applications. This study proposes a cross-continental transfer learning framework for PV power forecasting using a large-scale dataset with 4.5 million data points, including a source domain dataset with over 3.5 million data points. The prediction model, which is a transformer-based approach, compares the results of zero-shot learning, linear probing, fine-tuning, and training on the target dataset alone. The target datasets are located on different continents with different meteorological characteristics. Results demonstrate that the proposed transfer learning approach significantly improves prediction accuracy. It reduces the mean absolute percentage error by up to 38.8% compared with models trained solely on the target data. Further analysis reveals that freezing a few transformer blocks is beneficial when the target dataset is sufficiently large, whereas freezing most layers is effective for smaller datasets. This study analyzes the impact of source and target datasets on prediction performance, demonstrating that larger datasets in both domains enhance forecasting accuracy. These findings offer a robust approach for cross-continental knowledge transfer in renewable energy forecasting, particularly benefiting regions with limited historical data or newly installed systems. The source code is available on <uri>https://github.com/gist-ailab/transfer-learning-PV-forecasting</uri>
format Article
id doaj-art-78151015332f486fa1fff604bafaee29
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-78151015332f486fa1fff604bafaee292025-08-20T03:41:01ZengIEEEIEEE Access2169-35362025-01-011313617513619010.1109/ACCESS.2025.359104011086606Transfer Learning for Photovoltaic Power Forecasting Across Regions Using Large-Scale DatasetsSeongho Bak0https://orcid.org/0000-0001-6032-9279Sowon Choi1Donguk Yang2https://orcid.org/0000-0002-4235-3264Doyoon Kim3Heeseon Rho4Kyoobin Lee5https://orcid.org/0000-0003-4299-4923Department of AI Convergence, Gwangju Institute of Science and Technology (GIST), Gwangju, South KoreaDepartment of AI Convergence, Gwangju Institute of Science and Technology (GIST), Gwangju, South KoreaNew Energy Business Department, Hydrogen Business Team, Korea Electric Power Corporation (KEPCO), Naju, South KoreaDepartment of AI Convergence, Gwangju Institute of Science and Technology (GIST), Gwangju, South KoreaDepartment of AI Convergence, Gwangju Institute of Science and Technology (GIST), Gwangju, South KoreaDepartment of AI Convergence, Gwangju Institute of Science and Technology (GIST), Gwangju, South KoreaAccurate photovoltaic (PV) power forecasting enables stable grid operation; however, acquiring sufficient and diverse data remains challenging. Although numerous deep learning models have been employed, most rely on proprietary datasets or data spanning only a few years, limiting the generalizability of forecasting applications. This study proposes a cross-continental transfer learning framework for PV power forecasting using a large-scale dataset with 4.5 million data points, including a source domain dataset with over 3.5 million data points. The prediction model, which is a transformer-based approach, compares the results of zero-shot learning, linear probing, fine-tuning, and training on the target dataset alone. The target datasets are located on different continents with different meteorological characteristics. Results demonstrate that the proposed transfer learning approach significantly improves prediction accuracy. It reduces the mean absolute percentage error by up to 38.8% compared with models trained solely on the target data. Further analysis reveals that freezing a few transformer blocks is beneficial when the target dataset is sufficiently large, whereas freezing most layers is effective for smaller datasets. This study analyzes the impact of source and target datasets on prediction performance, demonstrating that larger datasets in both domains enhance forecasting accuracy. These findings offer a robust approach for cross-continental knowledge transfer in renewable energy forecasting, particularly benefiting regions with limited historical data or newly installed systems. The source code is available on <uri>https://github.com/gist-ailab/transfer-learning-PV-forecasting</uri>https://ieeexplore.ieee.org/document/11086606/Photovoltaic power forecastinglarge-scale datasettransfer learningdeep learning
spellingShingle Seongho Bak
Sowon Choi
Donguk Yang
Doyoon Kim
Heeseon Rho
Kyoobin Lee
Transfer Learning for Photovoltaic Power Forecasting Across Regions Using Large-Scale Datasets
IEEE Access
Photovoltaic power forecasting
large-scale dataset
transfer learning
deep learning
title Transfer Learning for Photovoltaic Power Forecasting Across Regions Using Large-Scale Datasets
title_full Transfer Learning for Photovoltaic Power Forecasting Across Regions Using Large-Scale Datasets
title_fullStr Transfer Learning for Photovoltaic Power Forecasting Across Regions Using Large-Scale Datasets
title_full_unstemmed Transfer Learning for Photovoltaic Power Forecasting Across Regions Using Large-Scale Datasets
title_short Transfer Learning for Photovoltaic Power Forecasting Across Regions Using Large-Scale Datasets
title_sort transfer learning for photovoltaic power forecasting across regions using large scale datasets
topic Photovoltaic power forecasting
large-scale dataset
transfer learning
deep learning
url https://ieeexplore.ieee.org/document/11086606/
work_keys_str_mv AT seonghobak transferlearningforphotovoltaicpowerforecastingacrossregionsusinglargescaledatasets
AT sowonchoi transferlearningforphotovoltaicpowerforecastingacrossregionsusinglargescaledatasets
AT dongukyang transferlearningforphotovoltaicpowerforecastingacrossregionsusinglargescaledatasets
AT doyoonkim transferlearningforphotovoltaicpowerforecastingacrossregionsusinglargescaledatasets
AT heeseonrho transferlearningforphotovoltaicpowerforecastingacrossregionsusinglargescaledatasets
AT kyoobinlee transferlearningforphotovoltaicpowerforecastingacrossregionsusinglargescaledatasets