Математическая модель от НИТУ «МИСИС» повысит КПД горнодобывающей отрасли

Researchers from NUST MISIS have developed a mathematical model that can increase the energy efficiency of mining enterprises by 20%. The development aims to accurately assess the operation of industrial equipment and reduce unjustified energy costs.

The new methodology is based on an econometric model of stochastic frontier analysis. Unlike existing expert assessments, the algorithm analyzes accurate data on equipment operation, identifying parameters at which the equipment consumes more energy than necessary.

What the new model (stochastic analysis) does:

Collects accurate data: It connects a "smart meter" to each "device" — recording thousands of parameters of equipment operation in different modes (loaded/empty, uphill/downhill, etc.).
Builds an ideal standard: The algorithm calculates how much energy a machine should consume ideally to perform a specific job.
Finds discrepancies: Then the program compares the ideal indicator with actual consumption and accurately determines the difference — those same "hidden losses" of 4.5-20.7%.
Separates random from systematic: An important point — the model "understands" which losses are caused by random factors (for example, bad weather), and which are a systemic problem (worn parts, suboptimal settings, operator errors) that can and should be fixed.

The technology translates the company's energy consumption from a mode of "guesses and approximate estimates" to a mode of precise digital diagnostics. It shows the management not just the fact of overspending, but its exact size and, most importantly, the source of the problem that needs to be eliminated.

The model was tested on data collected in 2021–2023 at a coal mine in Siberia. The study analyzed nearly five thousand work shifts. Calculations showed that previously unaccounted energy losses during the operation of heavy equipment, such as excavators and dump trucks, range from 4.5% to 20.7%.

According to scientists' estimates, these losses are equivalent to millions of kilowatt-hours and tens of millions of rubles annually for one enterprise. As the developers explained, the methodology is applicable not only in the mining industry, but also for other types of technological production, which can contribute to increasing their profitability and sustainable development.