Voronenko D. Predictive models of connected consumption of power resources for automated management of regional power network

A research object - are processes of automated control for regional energy systems. A research aim is a development of forecasting models set for coherent consumption of energy resources in public sector. Research methods and apparatus are system analysis methods, neural network technologies, local optimization methods, cluster analysis methods, imitation modeling methods and methods of statistical analysis of experimental data. Theoretical results - development of a set of forecasting models for coherent consumption of energy in public sector during heating season, that allows improving efficiency of automated control for electricity supply systems. The practical value of the work results consists in creation of algorithms and software for synthesis and analysis of forecasting models for daily electricity consumption as well as for meteorological data transformation into weather scenarios for usage in the set of forecasting models. The novelty of work consists in conducting of analysis of the current state and development tendencies for automated control process of regional energy systems as well as substantiation of necessity for development of a forecasting models set. A degree of introduction is the results of scientific developments are implemented in JSC "Kharkivoblenergo" (Kharkov) and used in lecture courses of the electrical stations department of NTU "KhPI". Sphere of the use - in practice of creation of automated control systems for large cities with centralized heating system to support regional decisions for energy resources distribution during heating season.