Modeling Diesel Demand in Iran Land Transport Sector Using GMDH Neural Network

Diesel is one of the most important energy carriers in Iran and the land transport is the largest consumer of diesel. According to the low price and the important role of this fuel in carrying passengers and goods in Iran, Checking and identification of variables have impressive on demand that is important. In this paper from the GMDH neural network is been used as a tool for high ability in routing and detection of complex non-linear trends, identification effective factors and modeling whit limit the number of observations. Also for choice of effective variables is used from the two fundamental and technical analysis. In This study fundamental analysis system has been evaluated in during three steps impact of 10 the system variables inside and outside the on diesel demand. results is indicates in the final stage of fundamental analysis that the system variables inside, GDP per capita, the number of diesel-powered vehicles and the system output variables of allocated to diesel subsidies and informal market exchange rate have doubly effect and liquidity has identical and ordinary effect on the diesel demand. In addition, the results are show that entry of variables of informal market exchange rate, liquidity and subsidy variables besides variables that have been used in previous research, in rise authenticity and precision evaluation criteria of forecast of the model has a significant impact. Also the results of modeling and forecasting diesel demand using both fundamental and technical analysis shows low error and the accuracy and high performance of the forecast.