Forecasting Natural Gas Consumption in Iran; Using a Combined Mathematical Approach

In this paper, an efficient approach for modeling the natural gas consumption in Iran is presented. The main objective is to forecast future natural gas consumption in Iran. First, a comprehensive literature review is done, concerning different mathematical techniques in order to project. Then, a combined mathematical approach is applied so as to forecast the future natural gas demand in Iran. This combined approach includes linear regression and meta-heuristic algorithms. In fact, the meta-heuristic algorithms are applied to correct the deficiency of regression-based techniques for such fields in which historical data are scarce. To serve the above purpose, a linear model is considered based on socio-economic indicators. These indicators include natural gas consumption, GDP, population, alternative and nuclear energy (Percent of total energy use), electricity production from natural gas sources and average domestic gas price. The model is developed using linear regression and GA (Genetic Algorithm) technique with insignificant error. Eventually, natural gas consumption in Iran is forecasted over the next 14years.

Keywords:

Natural gas consumption in Iran , Projection , Genetic algorithm , Linear regression Introduction