Long-term electrical energy demand forecasting by using artificial intelligence techniques – a case study of Turkey

Abstract

The article provides a long term forecast for gross annual electrical energy demand of Turkey by using socio-economic indicators such as gross domestic product, population, import and export values. The technics artificial neural networks (ANNs), adaptive neuro fuzzy inference system (ANFIS), and quadratic models among metaheuristic algorithms which are genetic algorithms (GA), differential evolution (DE) and particle swarm optimization (PSO) are applied to precisely formulate the relationship between the historical data and the forecasted data. The simulations are done on the actual energy demand data between the years 1975 through 2017. The performances provided by each model are evaluated over the training period. Future estimations of annual gross electrical energy demand of Turkey are projected up to 2030 under different scenarios. The methods are compared with five error metrics to determine the best model for forecasting. The mean average percentage error of all models is obtained within the range of high forecasting quality. Overall the best performance is obtained by the quadratic model with GA. Simulations are done on MATLAB™ platform.