Cross-Station Solar Power Prediction: A Transfer Learning Approach with Deep Learning Models

Pandith, Savan Kumar (2024) Cross-Station Solar Power Prediction: A Transfer Learning Approach with Deep Learning Models. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (2MB) | Preview

Abstract

This research investigates the effectiveness of hybrid deep learning models in predicting solar power generation using a transfer learning approach. Three models combining Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Transformer (TF) architectures were developed and evaluated CNN-LSTM-TF, CNNTF, and LSTM-TF. The models were trained on data from a lower capacity solar station and tested on a higher capacity station and the results showed a good performance by all three models in solar power forecasting, in which CNN-TF models emerged as the top performer with R2 score of 95% and lowest error in MSE of 35.63. The superior performance of CNN-TF models suggests that the combination of extracting features by CNN and the transformer-based attention mechanism has a good capability in capturing complex patterns in solar power generation data when compared to the LSTM-TF model, the CNN-LSTM-TF model performed better, with an R2 score of 92%. Also, the comparison study showed that CNN-TF models performed better than the existing studies.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Mulwa, Catherine UNSPECIFIED
Uncontrolled Keywords:	Solar power forecasting; Transfer learning; CNN; LSTM; Transformer-attention mechanism; Hybrid models; Deep learning; Cross-station prediction
Subjects:	G Geography. Anthropology. Recreation > GE Environmental Sciences Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HC Economic History and Conditions > Natural resources > Power resources > Energy consumption Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	25 Aug 2025 09:08
Last Modified:	25 Aug 2025 09:08
URI:	https://norma.ncirl.ie/id/eprint/8605

Actions (login required)

View Item