It has been predicted that in the coming decades a massive amount of new photovoltaic (PV) solar systems will be installed. This presents a significant challenge for the grid integration. To obtain additional reserves or change the output of conventional generators, accurate information on the solar power generation for the power systems is required, to preserve the balance between supply and demand.
As PV penetration increases, reliable meteorological and climatic data are becoming increasingly important. In the last 10 years, there has been a substantial increase in the research and interest in solar irradiance forecasting. For predicting solar radiation time series as well as probabilistic forecasting, multiple forecasting horizons and applications are being presented. The adopted forecasting strategy or modelling approach relies on the predictions and data input to the model. The relationship between solar irradiance forecasts and PV modelling is paramount, as solar forecasting for grid management, including storage, is required for high PV penetration. In particular, for grid-connected PV facilities, the ability to estimate solar irradiance and power can be critical in developing power dispatching plans, assuring grid stability, and enabling optimal unit commitment and economic dispatch. Furthermore, PV forecasts might be a valuable reference for refining charge controller control algorithms for stand-alone and hybrid systems.
In this context, this Research Topic welcome the recent advances covering the following themes;
• Machine learning and Deep learning techniques for photovoltaic and solar radiation forecasting to provide academics and practitioners with a wide view of current advanced methodologies.
• Forecasting contributions that employ data-driven methodologies based on new timestamps and weather condition features that affect the PV systems.
• Substantial advancements in the field of applications or suggest methodologies that have never been used in the field of renewable forecasting.
• Address the evaluation of different Machine learning and Deep learning techniques for PV forecasting for the benefits obtained by the blended models.
It has been predicted that in the coming decades a massive amount of new photovoltaic (PV) solar systems will be installed. This presents a significant challenge for the grid integration. To obtain additional reserves or change the output of conventional generators, accurate information on the solar power generation for the power systems is required, to preserve the balance between supply and demand.
As PV penetration increases, reliable meteorological and climatic data are becoming increasingly important. In the last 10 years, there has been a substantial increase in the research and interest in solar irradiance forecasting. For predicting solar radiation time series as well as probabilistic forecasting, multiple forecasting horizons and applications are being presented. The adopted forecasting strategy or modelling approach relies on the predictions and data input to the model. The relationship between solar irradiance forecasts and PV modelling is paramount, as solar forecasting for grid management, including storage, is required for high PV penetration. In particular, for grid-connected PV facilities, the ability to estimate solar irradiance and power can be critical in developing power dispatching plans, assuring grid stability, and enabling optimal unit commitment and economic dispatch. Furthermore, PV forecasts might be a valuable reference for refining charge controller control algorithms for stand-alone and hybrid systems.
In this context, this Research Topic welcome the recent advances covering the following themes;
• Machine learning and Deep learning techniques for photovoltaic and solar radiation forecasting to provide academics and practitioners with a wide view of current advanced methodologies.
• Forecasting contributions that employ data-driven methodologies based on new timestamps and weather condition features that affect the PV systems.
• Substantial advancements in the field of applications or suggest methodologies that have never been used in the field of renewable forecasting.
• Address the evaluation of different Machine learning and Deep learning techniques for PV forecasting for the benefits obtained by the blended models.
