Electricity customer classification using frequency–domain load pattern data

Abstract In competitive electricity markets, electricity customer classification is becoming increasingly important, due to new degrees of freedom the electricity providers have been given in formulating dedicated tariff options for different customer classes. Several customer classification techniques have been proposed in the literature, in which the load patterns are typically represented by time–domain data. However, a good load pattern representation requires using several data for each customer, causing possible difficulties in storing a large amount of data in the electricity company's databases. In order to reduce the number of data to be stored for each customer, an original solution is proposed in this paper, based on post-processing the results of time–domain measurements to obtain a reduced set of data defined in the frequency domain. The new set of data is successively used in a customer classification procedure, e.g. a suitable clustering technique, whose adequacy can be assessed by means of properly defined indicators. This paper provides the mathematical background for the frequency–domain data definition and investigates on the effectiveness of the customer classification for different choices of the number of data to be stored. Results obtained on a set of customers belonging to a real distribution system are presented and discussed. These results show that the proposed representation is effective in reducing the number of data stored while maintaining a satisfactory level of classification adequacy.

[1]  Dejan J. Sobajic,et al.  Combined use of unsupervised and supervised learning for dynamic security assessment , 1991 .

[2]  Enrico Carpaneto,et al.  Stratified sampling of the electricity customers for setting up a load profile survey , 2000 .

[3]  Enrico Carpaneto,et al.  Customer classification by means of harmonic representation of distinguishing features , 2003, 2003 IEEE Bologna Power Tech Conference Proceedings,.

[4]  C. S. Chen,et al.  Synthesis of power system load profiles by class load study , 2000 .

[5]  J. C. Hwang,et al.  Determination of customer load characteristics by load survey system at Taipower , 1996 .

[6]  P. Postolache,et al.  Load pattern-based classification of electricity customers , 2004, IEEE Transactions on Power Systems.

[7]  J. Neyman.  On the Two Different Aspects of the Representative Method: the Method of Stratified Sampling and the Method of Purposive Selection , 1934 .

[8]  Donald W. Bouldin,et al.  A Cluster Separation Measure , 1979, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[9]  Gianfranco Chicco,et al.  Application of clustering algorithms and self organising maps to classify electricity customers , 2003, 2003 IEEE Bologna Power Tech Conference Proceedings,.

[10]  Regina Lamedica,et al.  A bottom-up approach to residential load modeling , 1994 .

[11]  P. Stephenson,et al.  Tariff development for consumer groups in internal European electricity markets , 2001 .

[12]  P. Postolache,et al.  Customer Characterization Options for Improving the Tariff Offer , 2002, IEEE Power Engineering Review.

[13]  B.D. Pitt,et al.  Application of data mining techniques to load profiling , 1999, Proceedings of the 21st International Conference on Power Industry Computer Applications. Connecting Utilities. PICA 99. To the Millennium and Beyond (Cat. No.99CH36351).

[14]  G. Chicco,et al.  Characterisation of the aggregated load patterns for extraurban residential customer groups , 2004, Proceedings of the 12th IEEE Mediterranean Electrotechnical Conference (IEEE Cat. No.04CH37521).

[15]  C. F. Walker,et al.  Residential Load Shape Modelling Based on Customer Behavior , 1985, IEEE Transactions on Power Apparatus and Systems.

[16]  C. S. Chen,et al.  Application of load survey systems to proper tariff design , 1997 .