Аннотация
This study is important for the applied testing of big data processing methods using clustering machine learning methods coming from an industrial facility that generates electricity, heat, and industrial steam. A Pavlodar combined heat and power plant (CHPP) is special in that it is located in the northern zone of Kazakhstan, and has two different consumers, the aluminium plant and the medium-sized city heating system. The complexity in the variability of the consumption profile requires the use of modern methods of predictive planning of the composition of generating equipment for a day, a week, a year in advance. The deterioration due to long service life of the units, the hours worked require constant adjustment of the repair plan for the year ahead. It is necessary to take into account the curve of dynamic demand for heat and electricity using forecasting methods depending on the weather, the variability of heat, and electricity consumption changes by probabilistic laws based on historical data of population consumption. The correlation dependences between electricity consumption and ambient temperature could be helpful for predictive planning of the unit commitment, for optimal operations of the power plant, for meeting demand of each type of commodities using machine learning technologies.