Title: Oil demand forecasting for India using artificial neural network

Authors: S. Jebaraj; S. Iniyan

Addresses: Department of Mechanical Engineering, Universiti Teknologi PETRONAS, Perak, Malaysia ' Department of Mechanical Engineering, Anna University, Chennai, Tamil Nadu, India

Abstract: Energy is a vital input for the growth of any nation. Since oil resource has become a vital factor for future developments of a country, a system of models has to be developed to provide forecasts of oil demands in various sectors. This analysis utilises regression techniques, double moving average method, double exponential smoothing method, triple exponential smoothing method, Autoregressive Integrated Moving Average (ARIMA) model and Artificial Neural Network (ANN) model (univariate and multivariate) for oil demand forecasts in India. Model validation is done to select the best forecasting model. It is found that the ANN model gives better results in most of the cases. Hence, it is suggested that the ANN model can be used for forecasting oil demands in India. It is also predicted that the total oil demand for the years 2020 and 2030 will be 415,373 and 720,688 thousand tonnes, respectively.

Keywords: oil consumption; oil demand forecasting; forecasting models; ANNs; artificial neural networks; modelling; simulation.

DOI: 10.1504/IJGEI.2015.070280

International Journal of Global Energy Issues, 2015 Vol.38 No.4/5/6, pp.322 - 341

Available online: 20 Jun 2015 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article