Підвищення ефективності короткострокового прогнозування електричного навантаження енергооб'єднання

Abstract

This paper deals with the improvement of hierarchical multivariate mathematical model of the total electrical load (TEL) for solving short-term load forecasting. In the mathematical model of the TEL additionally introduced an astronomical component of the electrical load associated with the reaction of TEL on changes in the intensity of natural light. The algorithm for determining the astronomical component of TEL, and the results of calculations according to the JSC "Kyivenergo" are described. Additionally investigated questions about automatically choosing the optimal length of history of information for short-term load forecasting on different intervals of the year. The correctness of the proposed mathematical models applied to data of united power system of Ukraine using a three-level program developed by the multivariate forecasting of electrical load.