Vector Method for Singular Spectrum Analysis (SSA) of Electricity Consumption

Russian Federation's newly revised Energy Strategy 2035 implies shifting from relying on raw materials to innovative use of resources; it prioritizes qualitative change in the structure of energy consumption over quantitative increase, while also focusing on improving the quality of energy services, upgrades, and energy saving; the document emphasizes the need for technological innovation and more energy-efficient energy production. One aspect of the strategy concerns the monitoring and prediction of electricity consumption at various facilities, which requires qualitative analysis of inputs, data processing and further use. There is no single universal data processing method; the existing ones have many drawbacks, being unable to adjust for the specifics of electricity consumption on the site, not sufficiently accurate, computationally intensive, etc. Singular spectrum analysis has significant advantages and only minor flaws. SSA as a method for short-term prediction of electricity consumption at mining facilities is in line with Russia's energy strategy. Besides, accurate short-term predictions and the adaptability of mathematical modeling to change in electricity consumption trends are critical for predictive electricity management in the mining industry. However, papers covering the issue are non-existent, a situation that calls for research. Developing and implementing novel prediction methods as part of automated energy management systems will enable operators to accurately predict the emergence of problematic points of energy use to manage the load schedule more efficiently.