Current

# Semi-Automated Training of Non-Intrusive Load Monitoring Systems

### Faculty & Collaborators

Anthony Rowe , Mario Bergés

Suman Giri

### Project Description

Research has shown that providing meaningful feedback about electricity consumption patterns to consumers can reduce usage by 15-20%. One form of such meaningful feedback about electricity consumption is the appliance level breakdown of electrical energy usage in a building. This can be achieved through an approach pioneered in the 1980s called Non-Intrusive Load Monitoring (NILM). NILM typically includes an energy meter at the circuit panel level that can be used to monitor changes in voltage and/or current that are produced when appliances being fed through that circuit change states. Based on features of the signal changes (like change in magnitude, harmonics), the system identifies which appliances are turned ON at any given moment, and their respective power consumption levels.

CMU and Samsung are collaborating on this project to study the feasibility of sensor-aided NILM platforms. An EMF sensor has been designed at CMU (syntonistor) with hopes of training the central NILM algorithm on appliance signatures. The goals of this project are to perform extensive data collection from the syntonistor, study the feasibility of event detection and classification using the syntonistor, develop algorithms for automated training of NILM systems based on local EMF data, perform initial feasibility analysis of Fault Detection and Diagnosis (FDD) algorithms, and to create a test bed for experimentation.

### Selected Publications

1. S. Giri and M. Berges, "Towards automatic classification of appliances: Tackling cross talk in EMF sensors using blind source separation techniques," in Proceedings of the 2013 International EG-ICE Workshop on Intelligent Computing, Vienna, Austria.
bibtex
@inproceedings{giri_study_2013, address = {Vienna, Austria},  title = {Towards automatic classification of appliances: Tackling cross talk in EMF sensors using blind source separation techniques, booktitle = {Proceedings of the 2013 International {EG-ICE} Workshop on Intelligent Computing},  author = {Giri, Suman and Berges, Mario}, }
2. S. Giri, P. Lai, and M. Berges, "Novel Techniques For ON and OFF states detection of appliances for Power Estimation in Non-Intrusive Load Monitoring," in ISARC 2013 The 30th International Symposium on Automation and Robotics in Construction and Mining, Montreal Canada.
bibtex
@inproceedings{giri_transients_2013, address = {Montreal Canada},  title = {Novel Techniques For ON and OFF states detection of appliances for Power Estimation in Non-Intrusive Load Monitoring},  booktitle = {{ISARC 2013} The 30th International Symposium on Automation and Robotics in Construction and Mining},  author = {Giri, Suman and Lai, Po-Hsiang and Berges, Mario}, }
3. S. Giri, M. Bergés, and A. Rowe, "Towards automated appliance recognition using an EMF sensor in NILM platforms," Advanced Engineering Informatics, 2013.
bibtex
@article{Giri2013, title = {Towards automated appliance recognition using an {EMF} sensor in {NILM} platforms },  journal = {Advanced Engineering Informatics },  year = {2013},  issn = {1474-0346},  doi = {10.1016/j.aei.2013.03.004},  url = {http://www.sciencedirect.com/science/article/pii/S1474034613000268},  author = {Suman Giri and Mario Bergés and Anthony Rowe},  } }
4. N. Rajagopal, S. Giri, A. Rowe, and M. Berges, "A Magnetic Field-based Appliance Metering System," in ACM/IEEE Third International Conference on Cyber-Physical Systems, Philadelphia, PA.
bibtex
@inproceedings{rajagopal_magnetic_2013, address = {Philadelphia, PA},  title = {A Magnetic Field-based Appliance Metering System},  booktitle = {{ACM/IEEE} Third International Conference on Cyber-Physical Systems},  author = {Rajagopal, Niranjini and Giri, Suman and Rowe, Anthony and Berges, Mario}, }
5. A. Rowe, M. Berges, and R. Rajkumar, "Contactless sensing of appliance state transitions through variations in electromagnetic fields," in Proceedings of the 2nd ACM Workshop on Embedded Sensing Systems for Energy-Efficiency in Building, New York, NY, USA, 2010, p. 19.
bibtex
@inproceedings{rowe_contactless_2010, address = {New York, {NY},  {USA}},  series = {{BuildSys} '10},  title = {Contactless sensing of appliance state transitions through variations in electromagnetic fields},  isbn = {978-1-4503-0458-0},  location = {Zurich, Switzerland},  doi = {10.1145/1878431.1878437},  booktitle = {Proceedings of the 2nd {ACM} Workshop on Embedded Sensing Systems for {Energy-Efficiency} in Building},  publisher = {{ACM}},  author = {Rowe, Anthony and Berges, Mario and Rajkumar, Raj},  year = {2010},  note = {{ACM} {ID:} 1878437},  keywords = {experimentation, load tree analysis, measurement, {Non-Intrusive} Load Monitoring, performance, physical sciences and engineering},  pages = {19–24}, }
6. S. Giri and M. Berges, "A study on the feasibility of automated data labeling and training using an EMF sensor in NILM platforms. Best Paper award," in Proceedings of the 2012 International EG-ICE Workshop on Intelligent Computing., Herrsching, Germany.
bibtex
@inproceedings{giri_study_2012, address = {Herrsching, Germany},  title = {A study on the feasibility of automated data labeling and training using an {EMF} sensor in {NILM} platforms. {Best Paper award}},  booktitle = {Proceedings of the 2012 International {EG-ICE} Workshop on Intelligent Computing.},  author = {Giri, Suman and Berges, Mario}, }