The purpose of the project is to apply, improve, and develop new methods and algorithms for predictive data analytics in order to increase energy efficiency in heating systems and reduce heat waste.
Sensor data collected from substations, for example with the help of Individual Metering and Debiting (IMD) sensors, could reveal important properties of the heat delivery system from which the data was recorded. A key challenge, however, is understanding the potential of the data. In this presentation, Gideon will share his experiences from efforts aimed at the detection of anomalies when the underlying data is highly dimensional.
This webinar is part of a series of breakfast webinars arranged by Energiforsk's Värmekluster for the academics within district heating. The programme is supported by the Swedish Energy Agency and a number of companies working to accelerate innovation in the district heating sector.