Researchers from CEST (Skoltech) have patented a novel method for disaggregating cumulative energy consumption signals, which will help improve energy metering in industries, commercial buildings, and households. The patent covers the team’s research supported by the megagrant “Advanced methods for monitoring, protecting, and controlling future electrical systems” within the national project “Science and Universities.”
The novelty of the method lies in the application of statistical methods to split the overall data on power consumption into several components without adding new sensors. This is achieved by analyzing data from existing electricity meters to accurately identify the equipment consuming energy at any given moment.
While conventional techniques use several meters for different power phases, Skoltech’s method simplifies the process and enhances measurement accuracy. Leveraging advanced machine learning techniques, such as device clustering and neural network tuning, the researchers have made considerable progress in enhancing signal disaggregation accuracy.
The team’s findings offer significant benefits for the power industry by enabling more efficient energy consumption management, peak load reduction, and optimized resource allocation. Additionally, this research opens new horizons for enhancing energy efficiency for both individual enterprises and entire cities and regions.