Fraunhofer Building Innovation Alliance
An innovative meter system which can detect different electrical appliances from the overall power consumption – the affiliated partners Discovergy GmbH, EasyMeter GmbH, GreenPocket GmbH and RWE GBS GmbH have applied themselves with this challenging task under the lead of the Fraunhofer IMS. In the two year lasting research project NILM (Nonintrusive Load Monitoring) consumption patterns of individual appliances are extracted via data mining methods (machine learning techniques) out of high-frequency measuring data. The result of such a system is lower installation, maintenance and hardware costs compared to comparable sub-metering solutions. Furthermore, for the first time, the nonintrusive load monitoring system gives the building or facility operator broad access to device-specific analyses of energy consumption.
The guiding principle of the project is that every system or every appliance influence current and voltage in a characteristic way and therefore leaves a “fingerprint” in the power supply system. This fingerprint is recorded by measurement technology, like smart meters, as aggregated overall power consumption. By using these data mining methods patterns in the overall power consumption are detected and can be assigned to individual appliances. In a laboratory test environment appliances with a constant effective power consumption or a characteristic switch-on current pattern could easily be identified and their device-specific power consumption could be assigned.
