Data mining techniques for electricity customer characterization

Abstract The liberalization of electricity markets has been resulted in the emergence of new players, increasing the competitiveness in the markets, standing those can provide better services for better prices. The knowledge of energy consumers’ profile has been an important tool to help players to make decisions in the electrical sectors. In this paper, a characterization model of typical load curves for Low Voltage (LV) customers is proposed and evaluated. The identification of consumption patterns is based on clustering analysis. The clustering methodology is based on seven algorithms, partitional and hierarchical. Also, five clustering validity indices are used to identify the best data partition. With the knowledge obtained in clustering analysis, a classification model is used to classify new customers according to their consumption data. The classification model is used to select the correct class for each customer. To make the model simple, each load curve is represented by three indices which represent load curves shape. The methodology used in this work demonstrates to be an effective tool and can be used in most diverse sectors, highlighting the use of knowledge in the optimization of the energy contracting for low voltage customers. The energy consumption data can be constantly updated to improve the model precision, finding estimates that can better represent consumers and their consumption habits.

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

[2]  Z. Vale,et al.  An electric energy consumer characterization framework based on data mining techniques , 2005, IEEE Transactions on Power Systems.

[3]  Jie Lu,et al.  A New Index and Classification Approach for Load Pattern Analysis of Large Electricity Customers , 2012, IEEE Transactions on Power Systems.

[4]  Sarvapali D. Ramchurn,et al.  On the distinctiveness of the electricity load profile , 2018, Pattern Recognit..

[5]  J. Dunn.  Well-Separated Clusters and Optimal Fuzzy Partitions , 1974 .

[6]  Alfonso Capozzoli,et al.  Mining typical load profiles in buildings to support energy management in the smart city context , 2017 .

[7]  Sanjay Lall,et al.  Shape-Based Approach to Household Electric Load Curve Clustering and Prediction , 2017, IEEE Transactions on Smart Grid.

[8]  T. Caliński,et al.  A dendrite method for cluster analysis , 1974 .

[9]  A. Goia,et al.  Functional clustering and linear regression for peak load forecasting , 2010 .

[10]  N.D. Hatziargyriou,et al.  Two-Stage Pattern Recognition of Load Curves for Classification of Electricity Customers , 2007, IEEE Transactions on Power Systems.

[11]  H. Edelsbrunner,et al.  Efficient algorithms for agglomerative hierarchical clustering methods , 1984 .

[12]  Fábio Gonçalves Jota,et al.  Building load management using cluster and statistical analyses , 2011 .

[13]  Greg Hamerly,et al.  Learning the k in k-means , 2003, NIPS.

[14]  Khaled M. Abo-Al-Ez,et al.  A data mining based load forecasting strategy for smart electrical grids , 2016, Adv. Eng. Informatics.

[15]  Zita Vale,et al.  USE OF DATA MINING TECHNIQUES TO CHARACTERIZE MV CONSUMERS AND TO SUPPORT THE CONSUMER- SUPPLIER RELATIONSHIP , 2006 .

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

[17]  Uwe Aickelin,et al.  Variability of Behaviour in Electricity Load Profile Clustering; Who Does Things at the Same Time Each Day? , 2014, ICDM.

[18]  J. A. Hartigan,et al.  A k-means clustering algorithm , 1979 .

[19]  David C. H. Wallom,et al.  Impacts of Raw Data Temporal Resolution Using Selected Clustering Methods on Residential Electricity Load Profiles , 2015, IEEE Transactions on Power Systems.

[20]  Víctor Manuel Fernandes Mendes,et al.  A fuzzy clustering approach to a demand response model , 2016 .

[21]  Carlos Arthur Mattos Teixeira Cavalcante,et al.  A new method for pattern recognition in load profiles to support decision-making in the management of the electric sector , 2013 .

[22]  Furong Li,et al.  Multi-resolution load profile clustering for smart metering data , 2016, 2017 IEEE Power & Energy Society General Meeting.

[23]  Xiuwen Cai,et al.  Clustering-based short-term load forecasting for residential electricity under the increasing-block pricing tariffs in China , 2018, Energy.

[24]  Gianfranco Chicco,et al.  Emergent electricity customer classification , 2005 .

[25]  Z. Vale,et al.  Data mining techniques application in power distribution utilities , 2008, 2008 IEEE/PES Transmission and Distribution Conference and Exposition.

[26]  Michael Conlon,et al.  A clustering approach to domestic electricity load profile characterisation using smart metering data , 2015 .

[27]  Zita Vale,et al.  A data-mining-based methodology to support MV electricity customers’ characterization , 2015 .

[28]  G. N. Lance,et al.  A General Theory of Classificatory Sorting Strategies: 1. Hierarchical Systems , 1967, Comput. J..

[29]  Cheng Hao Jin,et al.  Short‐term electricity load and price forecasting based on clustering and next symbol prediction , 2015 .

[30]  Peter J. Rousseeuw,et al.  Clustering by means of medoids , 1987 .

[31]  Keun Ho Ryu,et al.  Local characterization‐based load shape factor definition for electricity customer classification , 2017 .

[32]  Benjamin C. M. Fung,et al.  Development of building energy saving advisory: A data mining approach , 2018, Energy and Buildings.

[33]  Carlos León,et al.  Electricity clustering framework for automatic classification of customer loads , 2017, Expert Syst. Appl..