Implementation of GSA (Gravitation Search Algorithm) and IWO (Invasive Weed Optimization) for The Prediction of The Energy Demand in Turkey Using Linear Form

This paper deals with energy demand forecast based on economic indicators in Turkey. Two different models based on the Gravity Search Algorithm (GSA) and Invasive Weed Optimization Algorithm (IWO) techniques are proposed to estimate energy demand. GSA is heuristic optimization algorithm inspired by Newton s laws of motion and gravity. The IWO algorithm is an evolutionary optimization algorithm inspired by the invasive characters of weeds in the wild. Energy demand models based on GSA and IWO methods are proposed using gross domestic product (GDP), population, import and export data as input parameters. Proposed methods are developed using linear regression model. Turkey s future energy demand is estimated under three different scenarios. The experimental results obtained by prediction models are given comparatively. In the prediction model using data between 1979 and 2005, IWO is compared with other methods in the literature and IWO method shows the highest performance. However, in the forecasting model obtained using the entire data set between 1979 and 2011, GSA is compared with the IWO method and GSA achieves better performance than IWO.