Electric energy consumption is becoming a relevant topic in the last years due to environmental and economic reasons. One of the field of interests in this energy problem is the reduction of the consumption in a house environment. A practical method to measure each appliance consumption is necessary to achieve this purpose. The non-intrusive appliance load monitor (NALM) approach measures aggregate power energy use as power enters the home with simple hardware but complex software to disaggregate the overall measured data in the single appliance. The aim of the project is the design of a self-learning algorithm that automatically identifies the appliances in a NALM system, determining itself the significant signatures and the associated appliances without any external information. The problem is divided into two parts: event detection and cluster analysis. The first part extracts from the overall data significant signatures that characterise the appliances. The extracted data are called events. The second part analyses the extracted events to find frequent patterns which identify the appliances
Self-learning algorithm for energy disaggregation
De Poli, Giulia
2012/2013
Abstract
Electric energy consumption is becoming a relevant topic in the last years due to environmental and economic reasons. One of the field of interests in this energy problem is the reduction of the consumption in a house environment. A practical method to measure each appliance consumption is necessary to achieve this purpose. The non-intrusive appliance load monitor (NALM) approach measures aggregate power energy use as power enters the home with simple hardware but complex software to disaggregate the overall measured data in the single appliance. The aim of the project is the design of a self-learning algorithm that automatically identifies the appliances in a NALM system, determining itself the significant signatures and the associated appliances without any external information. The problem is divided into two parts: event detection and cluster analysis. The first part extracts from the overall data significant signatures that characterise the appliances. The extracted data are called events. The second part analyses the extracted events to find frequent patterns which identify the appliancesFile | Dimensione | Formato | |
---|---|---|---|
tesi.pdf
accesso aperto
Dimensione
2.51 MB
Formato
Adobe PDF
|
2.51 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/15808