Modern capabilities of intelligent control systems are increasingly being used in areas previously considered the exclusive work of people — experts with relevant experience in a particular field. Machine learning capabilities in the field of electric power industry, obtaining forecasts based on the data of intelligent sensors of various purposes are not an exception. At present the Russian Government has adopted a program for the development of the manufacturing industry until the end of 2035: during this time manufacturing output should grow by 192 %. It is obvious that this program should also meet the requirements of the modern scientific concept of industrial revolution "Industry 4.0", when manufacturing enterprises and corporations begin to develop and apply subsystems and elements of "smart manufacturing", which help to build intelligent communications between individual tasks and operations during the entire life cycle of production, in accordance with the principles and methods of systems engineering. It is important to note that the issues of intelligent management in the subject-oriented area of electric power industry (in our case — energy saving), as the basis of any industrial production in modern conditions, require the development and implementation, first of all, of new solutions based on modern IT-technologies. It is known that energy intensity in Russia, according to the World Bank, is 3-4 times lower than in European countries. It is also known that in connection with the new provisions in the field of housing and communal services, aimed at improving economic efficiency in terms of electricity consumption, it becomes very important to ensure its accurate and operational accounting with the possibility of further forecasting of electricity consumption and the state of power grid facilities, which will allow specialized organizations and services, as well as the managing bodies in the shortest time to make a balanced specialized decisions This paper proposes the concept of intelligent control system to manage the process of condition monitoring based on data from intelligent sensors. The novelty of the concept is to consider a variant of solving the problem of integration of information systems associated with weakly structured subject-oriented information flows in the electric power industry enterprise by using methods of set theory and category theory.