Modelling of Turkish electricity consumption using genetic algorithm and ANFIS

Electric energy is one of the most important means of contemporary civilization that enables the development of societies from economic, social and cultural aspects.Electrical energy is the most popular in terms of comfortable living standards and other forms of energy. Electrical energy, which is indispensable to the type of energy today is an increasing need every day. The provision of electricity supply security requires accurate, reliable and meaningful estimations must be made for efficient electricity demand forecasts. There are many methods to model electricity forecasting. In this study, between the years 1995-2017 in order to model the electrical energy consumption in Turkey genetic algorithm for linear and quadratic forms and has been tested Anfis triangles and Gaussian membership functions. A hypothesis test was then conducted to determine whether there was a significant difference between the predicted values determined by the GA and ANFIS methods and the actual electricity consumption values. Linear model for GA, triangular membership function for Anfis model yielded successful results. In the linear model R2 value for the method GA %99,2 values and mape 3.2%; ANFIS method for triangular Membership function in the model the R2 value is %95,5 MAPE value and 3.9% have been identified as. As a result, GA and ANFIS established models can be used to estimate electricity consumption. Keywords: Genetic algorithm, ANFIS, Forecasting, Electricity Consumption