Smart meters as well as distributed power metering equipment (often referred to as smart plugs) have become widely available in recent years. These devices allow for the measurement of electric appliance power consumptions at high sampling frequencies, and also allow for the transmission of the collected power traces to a centralized processing platform. In general, the collected power traces do not only contain useful information about the energy consumption of the measured device, but also about its identity and the usage behavior of its owner. These valuable features are, however, hidden within the power traces, and need to be extracted.
We have thus developed the Smart Metering Toolkit SMTK to extract this information and prepare it for its presentation to the user. The SMTK allows for a batch processing of collected power consumption traces and helps the user to:
There are two main use cases for the Smart Metering Toolkit:
1. To collect power traces of different electrical appliances, we use distributed smart meters. Then we apply the SMTK to inspect and cleanse the collected measurements. Afterwards we export the power traces to the Tracebase.
2. To expose information hidden in the power trace the smart metering toolkit uses the Weka machine learning framework. Our SMTK provides the infrastructure for training and testing machine learning models that extract information from the given power traces. For example one can use the SMTK to classify the device based on the power consumption of that device.
For questions regarding SMTK please contact Frank Englert or Andreas Reinhardt.