Short-Term Power Production Forecasting in Smart Grid Based on Solar Power Plants

Qudsia Memon, Nurettin Cetinkaya

Since the world is moving towards the modernization so the smart grid idea is one of the smart idea leads to the modernization. One of the most important factors for the smart grid is the optimal production-commutation balance. Due to the lacking capabilities of accomplishing the increasing needs of the power with normal procedures, the world is moving towards the power production from the renewable energy sources. To get the efficient power production, the world is making the grids which generate power from the renewable sources, smart. Since solar is one of the important renewable energy sources, hence the changing climatic conditions affect heavily the ratio of power production from the solar sources. In this research, some of these climatic factors are considered to predict the solar power production by using the real-time data of a solar power plant located at Konya, Turkey. The inputs factors in consideration are on the daily basis which includes the average humidity, the minimum, average and the maximum temperatures, the solar irradiance, average and the maximum wind speed and the power generation values. The behavior of this solar power plant along with the prediction of the power production is carried out by using an Artificial Neural Network (ANN) in Matrix Laboratory (MATLAB) software’s built-in toolbox named as Neural Net Fitting toolbox. In ANNs, three different built-in learning algorithms in this toolbox named as Levenberg-Marquardt, Bayesian Regularization, and Scaled Conjugate Gradient are used to compare the prediction results, finally to get good and accurate results.