TY - CHAP
T1 - Deep Learning-Based Approach for Short-Term Solar Power Forecasting
AU - Carrera, Berny
AU - Kim, Kwanho
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.
PY - 2024
Y1 - 2024
N2 - In recent years, there has been a significant increase in the importance of solar energy as a clean and renewable energy source. Solar energy has become a clean, widely available, and eco-friendly option in the field of electricity generation. Solar power generation is not constrained by the negative effects of global warming and pollution, in contrast to traditional energy sources. The intermittent and erratic nature of solar energy, however, presents a serious obstacle. Accurate forecasting is crucial for fully using the potential of solar energy since it is closely linked to different weather conditions. Forecasting energy generation is crucial for smart grid operators and solar electricity suppliers since it is necessary to assure power continuity to dispatch and store the energy appropriately. In this study, we present a system for predicting solar power production 36 h in advance at the Yeongam solar power facility in South Jeolla Province, South Korea. The results demonstrate that the suggested method of utilizing several deep neural networks is superior to existing machine learning forecasting models.
AB - In recent years, there has been a significant increase in the importance of solar energy as a clean and renewable energy source. Solar energy has become a clean, widely available, and eco-friendly option in the field of electricity generation. Solar power generation is not constrained by the negative effects of global warming and pollution, in contrast to traditional energy sources. The intermittent and erratic nature of solar energy, however, presents a serious obstacle. Accurate forecasting is crucial for fully using the potential of solar energy since it is closely linked to different weather conditions. Forecasting energy generation is crucial for smart grid operators and solar electricity suppliers since it is necessary to assure power continuity to dispatch and store the energy appropriately. In this study, we present a system for predicting solar power production 36 h in advance at the Yeongam solar power facility in South Jeolla Province, South Korea. The results demonstrate that the suggested method of utilizing several deep neural networks is superior to existing machine learning forecasting models.
KW - Deep Learning
KW - Renewable Energy Forecasting
KW - Sustainable Urban Energy
UR - http://www.scopus.com/inward/record.url?scp=85183446646&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-52330-4_10
DO - 10.1007/978-3-031-52330-4_10
M3 - Chapter
AN - SCOPUS:85183446646
T3 - Springer Proceedings in Earth and Environmental Sciences
SP - 119
EP - 127
BT - Springer Proceedings in Earth and Environmental Sciences
PB - Springer Nature
ER -