Author(s)

S. A. Eroshenko, K. V. Vinokurov & A. Y. Smolina

Abstract

Some long-term design activities are based on electrical load forecasting. Forecasting errors result in the wrong decisions being made in the future. At present there are three widely used methods of forecasting: the method based on enlarged specific indexes (ESI), the econometric method and the self-sufficient method. All of these methods have their own disadvantages and errors in estimation. The main reasons for errors in areas such as objective reasoning, changes in production technology, the lack of reliable long-term forecasting for economy development and consumer infraction of load schedule are determined in this paper and an objective estimation for the forecasted error of different spheres of application is made. This work aims to determine a sphere of application for each of these methods and to develop a principal of combined application, so that they may decrease forecasting error. The purpose of this work consists of the creation of an electrical load forecasting method to obtain more reliable data for the different time periods and territory. A new method is presented, based on calculations carried out. Recommendations for all existing methods are given with reference to the calculations and the way that the combined application of several methods was introduced to increase the reliability of a forecast. Keywords: electric load, electricity demand forecasting.

Keywords

electric load, electricity demand forecasting.