A Multi-level Artificial Neural Network for Gasoline Demand Forecasting of Iran

Abstract

This paper presents a neuro-based approach for Iran annual gasoline demand forecasting by several socio-economic indicators. In order to analyze the influence of economic and social indicators on the gasoline demand, the gross domestic product (GDP), the population and the total number of vehicles are selected. This approach is structured as a multi-level artificial neural network (ANN) based on supervised multi-layer perceptron (MLP), trained with the backpropagation (BP) algorithm. This multi-level ANN is designed properly. Actual data of Iran from 1968-2006 is used to train the multi-level ANN and illustrate capability of the approach in this regard. Comparison of the model predictions with data of the evaluating period shows validity of the model. Furthermore, the demand for the period of 2007 to 2030 is estimated.