Shulle Y. Efficiently prognostication of the electric loadings in electroconsumption systems by the means of their fractal properties

A research object is the the process of electric energy industry, electrical equipment and systems; dissertation work is devoted development of new method and models of operative electric loadings prognostication at the the systems of electroconsumption by the means of their fractal properties, that will allow to promote efficiency of management of electro-consumption. Тheoretical results - were was first discovered fractal properties of electrical loads charts and the availability of short-term and long-term of their memory, allowing the use of fractal analysis to mathematical modeling and prediction of future values depending on their schedules loads of past changes or "initial conditions", were was first developed a method for operational forecasting of electricity load in real time, taking into account of its fractal nature, based on neural network technology which can improve prediction accuracy compared to existing methods, improved mathematical model to describe the characteristics of graphs loads as a random process, and classification scenarios electrical loads systems power consumption based on the selection of stochastic component of the dynamic range of power consumption, allowing for operational forecasting electrical loads. Practical results - the algorithm for operational forecasting electrical loads of industrial enterprises with regard to its fractal nature, based on neural network technology, which can detect critical patterns of electricity consumption and reduce the risk of accidental off; developed hardware-software complex to automate the operational forecasting of electric load systems electricity, which can increase accuracy and speed. The degree of implementation - the results have been implemented in Ltd. "Vinnitsa bakery number 2" (the act of 03.09.2012), and the learning process of Vinnytsia National Technical University (the act of 04.09.2012). Usage field - power engineering.