
Results in Engineering, Год журнала: 2024, Номер unknown, С. 103434 - 103434
Опубликована: Ноя. 1, 2024
Язык: Английский
Results in Engineering, Год журнала: 2024, Номер unknown, С. 103434 - 103434
Опубликована: Ноя. 1, 2024
Язык: Английский
Results in Engineering, Год журнала: 2025, Номер unknown, С. 103838 - 103838
Опубликована: Янв. 1, 2025
Язык: Английский
Процитировано
4Results in Engineering, Год журнала: 2025, Номер unknown, С. 105046 - 105046
Опубликована: Апрель 1, 2025
Язык: Английский
Процитировано
0Expert Systems with Applications, Год журнала: 2025, Номер unknown, С. 127812 - 127812
Опубликована: Май 1, 2025
Язык: Английский
Процитировано
0EPJ Web of Conferences, Год журнала: 2025, Номер 326, С. 05002 - 05002
Опубликована: Янв. 1, 2025
Accurate forecasting of Global Horizontal Irradiance (GHI) is critical for enhancing both grid stability and the efficiency solar energy systems. A comparative assessment several deep learning models presented in this study real-time GHI forecasting, specifically Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), a hybrid LSTM-GRU architecture. Approach performance evaluated using standard metrics, including MAE, RMSE, R². Findings indicate that while GRUs are computationally efficient, they struggle to maintain long-term temporal dependencies. In contrast, LSTMs effectively capture these dependencies, resulting improved accuracy. Notably, model outperforms individual architectures, achieving lowest MAE (12.931), RMSE (21.825), highest R² (0.996), thereby demonstrating superior predictive performance. These results highlight potential applications, improving forecast reliability stability. This advances irradiance methodologies, facilitating integration renewable sources effectiveness operations.
Язык: Английский
Процитировано
0Applied Soft Computing, Год журнала: 2025, Номер unknown, С. 113208 - 113208
Опубликована: Май 1, 2025
Язык: Английский
Процитировано
0Energies, Год журнала: 2024, Номер 17(22), С. 5767 - 5767
Опубликована: Ноя. 18, 2024
The global horizontal irradiance (GHI) is the most important metric for evaluating solar resources. accurate prediction of GHI great significance effectively assessing energy resources and selecting photovoltaic power stations. Considering time series nature monitoring sites dispersed over different latitudes, longitudes, altitudes, this study proposes a model combining deep neural networks convolutional multi-step GHI. utilizes parallel temporal gate recurrent unit attention prediction, final result obtained by multilayer perceptron. results show that, compared to second-ranked algorithm, proposed improves evaluation metrics mean absolute error, percentage root square error 24.4%, 33.33%, 24.3%, respectively.
Язык: Английский
Процитировано
1Results in Engineering, Год журнала: 2024, Номер unknown, С. 103767 - 103767
Опубликована: Дек. 1, 2024
Язык: Английский
Процитировано
1Results in Engineering, Год журнала: 2024, Номер unknown, С. 103434 - 103434
Опубликована: Ноя. 1, 2024
Язык: Английский
Процитировано
0