Scientists from Perm National Research Polytechnic University (PNRPU) have created an artificial intelligence-based system that can reduce urban heating costs by 10–12% per year. The development takes into account the weather forecast and the condition of the equipment, optimizing the supply of heat to homes.
The new system uses a neural network algorithm that analyzes data from temperature and pressure sensors in heating networks, and also takes into account weather forecast data. This allows you to automatically adjust the temperature of the coolant, avoiding overheating or insufficient heating of rooms. Testing showed the accuracy of the algorithm at 97.9%.
According to Vladimir Oniskiv, Associate Professor at PNRPU, the development was trained on a virtual stand simulating the operation of heating networks with different parameters. After additional training in real conditions, the system adapts to the characteristics of a specific network.
When testing the algorithm on test indicators, its accuracy was 97.9%. Our development accurately predicts what the temperature of the coolant should be at the outlet of the boiler room so that it is comfortable in the houses. The main advantage is that the system quickly adapts to changes in the weather and adjusts the temperature so that it reaches the consumer close to the standard values. This eliminates the so-called "overheating" of the system and leads to savings in resource costs by approximately 10-12% during the heating period.
Centralized heat supply provides heating and hot water for about 70% of the Russian population. However, due to equipment wear and manual control, heat losses can reach 30%. The introduction of an AI solution will help reduce costs and increase energy efficiency.
The study was conducted as part of the Priority-2030 program, and its results are published in the collection "XIV All-Russian Meeting on Management Problems".
Subscribe to the news of "First Technical" on Google News to stay up to date with the latest events from the world of science and technology.
Read more materials on the topic:
A new voice fake recognition system "Sipuha" was created by MEPhI scientists