Katyukha I. Prediction of electrical load distribution networks under uncertainty of the initial information

Object of research: the process of forecasting electrical loads of distribution networks in the conditions of uncertainty of the source information. Subject of research: - methods and models of forecasting of electrical loads of distribution networks taking into account the characteristics of individual electric consumers in conditions of uncertainty of the initial information. The purpose of the dissertation work: to increase the efficiency of consumption and conservation of electric energy in distribution networks by improving the forecast of electricity consumption. Research methods: for studying loading schedules - elements of mathematical statistics and regression analysis; for forecasting of electrical loads - methods of fuzzy regression analysis; for the development of algorithms and software for predicting power consumption - simulation modeling, mathematical programming methods. consists in the developed method of forecasting of electric consumption on the basis of fuzzy regression analysis, which includes the principles of constructing predictive models, algorithmic support, as well as software, implemented in a convenient for integration in ASKOE form. The method of fuzzy regression analysis for the construction of long-term predictive models of electrical loads in distribution networks has been improved. The parity participation of two criteria of fuzzy model efficiency: the degree of alignment and the degree of fuzziness when constructing forecast models are taken into account. The method of correction of long-term forecasting models for short-term forecast is developed. The approach to construction of the kind of forecasting models for any types of loads is proposed. The connection of fuzzy indicators of accuracy of the forecast with the relative median-modulus error is established analytically. The developed method is tested in the development of predictive models of electric load of a number of consumers with different types of load schedules. The scientific novelty of the obtained results: the method of obtaining predictive models of electrical loads, which differs by constructing a fuzzy regression compatibility criterion based on the intersection of fuzzy numbers, has been improved, which makes it possible to reveal the uncertainty of the output data and improve the quality of the prediction of electrical loads; received further development of the method of unification of the type of forecast models, which differs in that in the daily schedule of electricity consumption allocated functional areas, with the separate use for them of fuzzy regression analysis, which allows you to get the kind of forecast models for different loads; For the first time analytical method is given to determine the efficiency of the forecast of electricity consumption in electric networks, which allows to perform a comparative analysis of fuzzy regression models of the forecast with models obtained by other methods. The main results of the research carried out in the dissertation are implemented in PE "Milk Factory-OLCOM", Priazovsky REM of JSC "Zaporizzhyaoblenergo" and used in the educational process of the Department of Electrical Engineering and Electromechanics of the Tavria State Agrotechnological University.