The Solar Energy Forecasting Using LSTM Deep Learning Technique

The future power systems will be characterized by high penetration or even dominance of renewable energy sources. This will present a challenge for the power system operators responsible for the coordination and distribution of electricity due to the direct dependence of the renewable energy sources on climatic changes. For its unpredictable and intermittent nature, the integration of the renewable energy sources imposes significant problems on the management of the grid and the balance between electricity consumption and production. Deep learning techniques are used to predict an accurate generated power of solar photovoltaic (PV) power plants. The various meteorological conditions are related to each other in terms of influence. In this paper, a wide range of features are considered in the forecasting process. From the results, the effect of the atmospheric factors on the forecasting process is not the same. Attention should be paid to these differences to increase the accuracy of the prediction. The root mean squared error (RMSE) and mean square error (MSE) are used for evaluation.

I. Introduction

In the last decade, there has been a huge increase on the use of renewable energies such as the PV energy due to the high consumption of energy. Despite the advantages of these types of energies, they suffer from the problem of intermittence. An accurate prediction of the power generated from PV panels is very essential for successful integration of into the grid. The unpredicted behavior of solar energy causes some problems, such as stability and voltage variation. Moreover, PV panels may cause imbalance in power dispatching in power systems. Therefore, good prediction methods are increasingly become very important measures. These methods are constantly being developed with the aim of achieving an accurate prediction [1]. Deep learning techniques, in general, are used to overcome the problems associated with the integration of renewable sources into the national electric grid [2].