| Title |
Application of Recurrent Neural Network Model for Short Term Photovoltaic Generation Forecasting |
| ID_Doc |
29960 |
| Authors |
Dinesh, LP; Al Khafaf, N; McGrath, B |
| Title |
Application of Recurrent Neural Network Model for Short Term Photovoltaic Generation Forecasting |
| Year |
2022 |
| Published |
|
| DOI |
10.1109/AUPEC58309.2022.10216175 |
| Abstract |
This paper presents the application of a Recurrent Neural Network (RNN) Model for one-step forecasting of Photovoltaic (PV) generation using exogenous weather variables with a 15-minute resolution. The paper finds application in Australian urban environments, with an RMIT University campus building based in the City of Melbourne presented as a representative example for the time period of January to December 2020. The model accuracy is expressed in terms of Normalised Mean Absolute Error (nMAE) normalised by the range of the actual generation values. The nMAE values are below 0.5 in all the cases, suggesting the model has a reasonable accuracy. This model underlines the importance of very short term forecasting in the context of energy trading, when the PV panels export power to the grid and when backup and storage assets such as batteries form part of the grid. |
| Author Keywords |
RNN; PV generation forecasting; machine learning; short term forecasting |
| Index Keywords |
Index Keywords |
| Document Type |
Other |
| Open Access |
Open Access |
| Source |
Conference Proceedings Citation Index - Science (CPCI-S) |
| EID |
WOS:001058521300029 |
| WoS Category |
Energy & Fuels; Engineering, Electrical & Electronic |
| Research Area |
Energy & Fuels; Engineering |
| PDF |
|