Accurate forecasting of solar irradiance is crucial for minimizing uncertainty of the solar energy output from PV plants, as solar irradiance can be utilized to determine the solar energy produced. The proposed models use the historical solar irradiance data along with 8 features namely temperature, dew, humidity, wind gust, wind speed, cloud cover, pressure, and visibility. The primary objective is to create a model that forecasts solar irradiance accurately. LSTM, GRU, SVR, and BiLSTM models are used and they are trained and tested on a dataset of 7 months with the frequency of one hour collected from a site in Tamil Nadu. By calculating the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE), the performance of the models are evaluated. The LSTM model has performed better than other models and gave better accuracy in terms of RMSE and MAE.
Paper
Full text
Solar Irradiance Forecasting Using Machine Learning
Semantic Scholar · Engineering · 2023
Abstract
Accurate forecasting of solar irradiance is crucial for minimizing uncertainty of the solar energy output from PV plants, as solar irradiance can be utilized to determine the solar energy produced. The proposed models use the historical solar irradiance data along with 8 features namely temperature, dew, humidity, wind gust, wind speed, cloud cover, pressure, and visibility. The primary objective is to create a model that forecasts solar irradiance accurately. LSTM, GRU, SVR, and BiLSTM models are used and they are trained and tested on a dataset of 7 months with the frequency of one hour collected from a site in Tamil Nadu. By calculating the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE), the performance of the models are evaluated. The LSTM model has performed better than other models and gave better accuracy in terms of RMSE and MAE.