Published Fast: - If it's accepted, We aim to get your article published online in 48 hours.

Home / Articles

No Article found
Deep Learning Paradigms for Solar Radiation Prediction: A Review of Recent Advances
Author Name

K.C. Jayasankar, G. Anandhakumar

Abstract

Deep learning algorithms present a promising approach to solar energy prediction with the ability to produce accurate forecasts. Amidst the realms of scholarly exploration, an exhaustive examination delves into myriad intricate learning frameworks harnessed for the analysis of temporal data, aimed at prognosticating solar radiation and the consumption of energy via photovoltaic (PV) means. The focal points of scrutiny encompass a spectrum of architectural constructs, including the recurrent neural network (RNN), the long short-term memory (LSTM), the gated recurrent unit (GRU), and the convolutional neural network to LSTM (CNN-LSTM) amalgamation. The assessment of these edifices revolves around sundry determinants, such as precision, the composite nature of input data, forecasting duration, seasonal undulations, meteorological parameters, and the temporal requisites for training. The research highlights the unique benefits and limitations that are present in each architectural design across various contexts. The performance of LSTM is remarkable, since it outperforms alternative independent designs, particularly in terms of the root-mean-square error (RMSE) statistic. On the other hand, the combined CNN-LSTM design demonstrates more effectiveness compared to individual architectures, however it requires a longer training period. An important observation suggests that deep learning architectures are becoming more effective than traditional machine learning models in predicting solar irradiance and PV power. Furthermore, the utilization of relative root mean square error (RMSE) arises as a relevant assessment measure that enables accurate comparisons among diverse investigations.

Keywords: Recurrent neural network (RNN), Long short-term memory (LSTM), Gated recurrent unit (GRU), Convolutional neural network-LSTM (CNN-LSTM) and Deep Belief Networks (DBN)



Published On :
2024-05-31

Article Download :
Publish your academic thesis as a book with ISBN Contact – connectirj@gmail.com
Visiters Count :