Forecasting electricity consumption by clustering data in order to decline the periodic variable’s affects and simplification the pattern

Electricity demand forecasting is known as one of the most important challenges in managing supply and demand of electricity. Consumption pattern of electricity has been affected by some social, economical and environmental factors by which the pattern will form various seasonal, monthly, daily and hourly complex variations. Diversity and complexity in consumption pattern of electricity have been leading to the extension of the complicated models. Many attempts have been made to find the best estimation for electricity consumption. These studies have been tried to forecast the demand in two levels: (1) Macro economic decision making and (2) Engineering and middle management. In this research an attempt has been made to introduce a method for pre-preparing data and for developing a model that could be applied in both the mentioned levels. By clustering primary data and by eliminating the periodic variance in our study, the complicated pattern is decomposed to a set of simple patterns which could be easily analyzed with conventional tools in both the levels.