MACROMODELING FOR THE FORECASTING OF LOCAL ENERGY SUPPLY SYSTEMS BALANCE, APPLYING FRACTAL PROPERTIES OF LOADING AND GENERATION GRAPHS

Abstract

Method for the forecasting of the energy consumption balance of the urban civil development objects using discrete macro-models has been suggested. Energy supply system (ESS) of the district is considered, characteristic feature of the system is energy consumption from the electric public grids and which has own source of electric energy generation from the renewable sources of energy (RSE). Such local electric energy system (LES) at certain conditions can operate as an independent balanced energy object. For optimal operation of LES in such conditions the graphs of its energy consumption must be forecast.

The suggested method of macro modeling enables to develop with the needed accuracy the determined models of energy consumption, based on the retrospective information without using the procedures of the preliminary data processing. The solution of the problem of energy consumption graphs forecast is simplified as a result of using in the process of model construction only the basic or determining characteristics. They comprise fractal properties of the graphs of ESS loading. The essence of the alternative forecasting on the base of macro modeling is the process of construction of energy consumption model by stages. At the first stage of the analysis procedure realization the collection and processing of the data regarding energy consumption of the investigated real object is provided, the second stage includes the selection of mathematical model, further improvement of the mathematical model parameters and testing.

The performance of the suggested method was verified by means of construction the macro model of daily energy consumption of 9-story 216-apartment residential building. For this purpose input data of average daily energy consumption and week average consumption were used. The results obtained showed that using АSCAЕ (Automatic system for commercial accounting of power consumption) information it is possible to obtain a number of obvious daily graphs of electric loading and generalize them for further usage. As a result of the application in the process of macro modeling the fractal properties of LES loading graphs the accuracy of the energy consumption forecast improves and the term of the satisfactory forecast increases.