An Experimental Study on Electrical Signature Identification of Non-Intrusive Load Monitoring (NILM) Systems

Electrical load disambiguation for end-use recognition in the residential sector has become an area of study of its own right. Several works have shown that individual loads can be detected (and separated) from sampling of the power at a single point (e.g. the electrical service entrance for the house) using a non-intrusive load monitoring (NILM) approach. This work presents the development of an algorithm for electrical feature extraction and pattern recognition, capable of determining the individual consumption of each device from the aggregate electric signal of the home. Namely, the idea consists of analyzing the electrical signal and identifying the unique patterns that occur whenever a device is turned on or off by applying signal processing techniques. We further describe our technique for distinguishing loads by matching different signal parameters (step-changes in active and reactive powers and power factor) to known patterns. Computational experiments show the effectiveness of the proposed approach.

[1]  Thorsten Joachims,et al.  Making large scale SVM learning practical , 1998 .

[2]  Xiaohua Xia,et al.  Active power residential non-intrusive appliance load monitoring system , 2009, AFRICON 2009.

[3]  A. Albicki,et al.  Algorithm for nonintrusive identification of residential appliances , 1998, ISCAS '98. Proceedings of the 1998 IEEE International Symposium on Circuits and Systems (Cat. No.98CH36187).

[4]  Lucio Soibelman,et al.  Learning Systems for Electric Consumption of Buildings , 2009 .

[5]  Steven B. Leeb,et al.  A conjoint pattern recognition approach to nonintrusive load monitoring , 1993 .

[6]  Thomas Hofmann,et al.  Support vector machine learning for interdependent and structured output spaces , 2004, ICML.

[7]  Jon Atli Benediktsson,et al.  Evaluation of Kernels for Multiclass Classification of Hyperspectral Remote Sensing Data , 2006, 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings.

[8]  F. Sultanem,et al.  Using appliance signatures for monitoring residential loads at meter panel level , 1991 .

[9]  Pat Langley,et al.  Editorial: On Machine Learning , 1986, Machine Learning.

[10]  Hujun Yin,et al.  Intelligent Data Engineering and Automated Learning - IDEAL 2010, 11th International Conference, Paisley, UK, September 1-3, 2010. Proceedings , 2010, IDEAL.

[11]  Bernardete Ribeiro,et al.  Extracting Features from an Electrical Signal of a Non-Intrusive Load Monitoring System , 2010, IDEAL.

[12]  G. W. Hart,et al.  Nonintrusive appliance load monitoring , 1992, Proc. IEEE.

[13]  Hsueh-Hsien Chang,et al.  Load identification in nonintrusive load monitoring using steady-state and turn-on transient energy algorithms , 2010, The 2010 14th International Conference on Computer Supported Cooperative Work in Design.

[14]  A. Albicki,et al.  Data extraction for effective non-intrusive identification of residential power loads , 1998, IMTC/98 Conference Proceedings. IEEE Instrumentation and Measurement Technology Conference. Where Instrumentation is Going (Cat. No.98CH36222).

[16]  Koby Crammer,et al.  On the Algorithmic Implementation of Multiclass Kernel-based Vector Machines , 2002, J. Mach. Learn. Res..

[17]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.