Prediction of Renewable Electricity Output of Middle East and North Africa Countries Using Artificial Neural Networks

Considering that renewable energy sources depend on environmental conditions, their integration in the electrical grid introduces some uncertainty into the power generation process. For this reason, accurate prediction of electrical energy generation from renewable sources has a vital impact on taking countermeasures against power outage and providing sustainability. Although there are numerous conventional but sophisticated approaches that can be utilized in energy forecasting, the simplicity of recently developed nature-inspired algorithms makes them an ideal option for prediction. In this study a bio-inspired algorithm, artificial neural network, is examined for the prediction of renewable electricity output of Middle East and North Africa countries. Electrical power consumption, energy intensity level of primary energy, GDP per unit of energy use and energy imports are used as input parameters of the artificial neural network and renewable electricity output is assigned as output parameter. The statistical data of world development indicators between 1990 and 2014 are processed. The proposed method provided accurate and reliable results in prediction.