Forecasting of petroleum-based energy consumption in Iran using Artificial Neural Network (ANN)

Petroleum based resources provide different products including transportation fuels, fuel oils for heating and electricity generation, asphalt and road oil, and the feedstocks used to make chemicals, plastics, and synthetic materials found in nearly everything we use today. Optimal prediction and forecasting of petroleum-based resource consumption would help energy policy makers and practitioners to take appropriate strategic decisions in this sector. In this paper, an Artificial Neural network approach is proposed to optimum forecasting of petroleum-based energy consumption in Iran. Oil and natural gas together make petroleum. The most important and effective environmental and economic factors in petroleum resource consumption are considered in the applied ANN. The ANN trains and tests data with Multi-Layer Perceptron approach which has the lowest mean absolute percentage error. The obtained results justify the efficiency of the proposed approach in prediction of petroleum based energy consumption.