Russian Scientists Patent Digital Tool to Prevent Utility Network Failures
Model improves forecasting, reduces energy losses, and optimizes system management
Researchers at the Artificial Intelligence Center of Novosibirsk State University have developed a mathematical model of heating networks designed to analyze and optimize district heating systems, the university’s press service said.
The digital model allows engineers to test different calculation and control algorithms without risking real infrastructure. While not tied to a specific city, the tool replicates the key characteristics of real-world heating systems.
A Real-World Use Case
The model network is also well suited for training neural networks, as it includes datasets that reflect system behavior under different scenarios. The tool can be used to monitor operational parameters in real time, detect leaks, and predict potential failures.
The next stage of the project involves developing digital twins tailored to specific heating networks, taking into account their unique characteristics. In the future, the technology could be scaled to other cities.
