Comparative Study of Univariate and Multivariate Long Short-Term Memory for Very Short-Term Forecasting of Global Horizontal Irradiance

However, developing a reliable GHI forecasting model is challenging due to the spatial, temporal, and meteorological variability of GHI. Recently, the long short-term memory (LSTM) deep learning network has become a powerful tool for modeling complex time series problems. This study aims to develop and compare univariate and multivariate LSTM models that can predict GHI in Guntur, India on a very short-term basis. To build the multivariate time series models, all possible combinations of temperature, humidity, and wind direction variables along with GHI as inputs were considered, resulting in seven multivariate models. In contrast, the univariate model considered only GHI variability. Meteorological data for Guntur from 1 January 2016 to 31 December 2016 was collected and used to build 12 datasets, each containing variability of GHI, temperature, humidity, and wind direction of a month. The models were constructed to measure up to 2 h ahead of forecasting of GHI. To evaluate the performances of the prediction models, root mean square error (RMSE) and mean absolute error (MAE) were used. The results indicate that compared to the univariate method, each multivariate LSTM performs better in the very short-term GHI prediction task. Among the multivariate LSTM models, the model that incorporates the temperature variable with GHI as input has outweighed others, achieving average RMSE values of 0.74 W/m2–1.5 W/m2. The study employed a one-year weather observation from Guntur, India to build the models and observe their performances. The models were forecasted GHI up to 2 h ahead and analyzed the effect of different input variables in the forecasting task. The main contributions of the paper include the development of two categories of models, univariate LSTM and multivariate LSTM, to predict GHI one to 24 steps ahead. The study compared the performance of all models in very short-term GHI forecasting. Experimental results demonstrate the effectiveness of the multivariate LSTM models over the univariate model, meaning that inclusion of additional meteorological variables can improve prediction models. In addition, among the multivariate models, two models have far outperformed others. The study highlights the importance of accurate GHI forecasting for the efficient management of PV power plants. The results of this study can be used to improve the performance of GHI forecasting models, which can lead to better management of PV power plants and increased efficiency in energy production. The study also emphasizes the need for further research on GHI forecasting using deep learning approaches, particularly multivariate LSTM models. The study employed several statistical models, machine learning algorithms, and deep learning approaches for GHI forecasting. The results of this study can be used to improve the performance of GHI forecasting models, which can lead to better management of PV power plants and increased efficiency in energy production. The study also highlights the importance of considering multiple meteorological variables in GHI forecasting models to improve their performance. The study's findings have significant implications for the development of efficient GHI forecasting models. The results of this study can be used to improve the performance of GHI forecasting models, which can lead to better management of PV power plants and increased efficiency in energy production. The study also emphasizes the need for further research on GHI forecasting using deep learning approaches, particularly multivariate LSTM models. The study's methodology involved collecting meteorological data for Guntur from 1 January 2016 to 31 December 2016 and building 12 datasets, each containing variability of GHI, temperature, humidity, and wind direction of a month. The models were constructed to measure up to 2 h ahead of forecasting of GHI. To evaluate the performances of the prediction models, root mean square error (RMSE) and mean absolute error (MAE) were used. The study's results demonstrate the effectiveness of the multivariate LSTM models over the univariate model, meaning that inclusion of additional meteorological variables can improve prediction models. In addition, among the multivariate models, two models have far outperformed others. The study's findings have significant implications for the development of efficient GHI forecasting models. The study's results can be used to improve the performance of GHI forecasting models, which can lead to better management of PV power plants and increased efficiency in energy production. The study also emphasizes the need for further research on GHI forecasting using deep learning approaches, particularly multivariate LSTM models. The study's methodology involved collecting meteorological data for Guntur from 1 January 2016 to 31 December 2016 and building 12 datasets, each containing variability of GHI, temperature, humidity, and wind direction of a month. The models were constructed to measure up to 2 h ahead of forecasting of GHI. To evaluate the performances of the prediction models, root mean square error (RMSE) and mean absolute error (MAE) were