Philippe RAVIER, Tayeb MOHAMADI, Fateh GHAZALI, Abdenour HACINE-GHARBI

Extraction and selection of statistical harmonics features for electrical appliances identification using k-NN classifier combined with voting rules method

In this paper, we propose a novel framework for electrical appliances identification using statistical harmonicfeatures of current signals and the use of the k-NN classifier combined with a voting rule strategy. Harmonic coefficientsare computed over time using short-time Fourier series of the current signals. From these sequences of coefficients, themean, standard deviation, skewness, and kurtosis are computed, which provide the statistical harmonic features. Thisframework has three novelties: (i) selecting the best combination of statistical measures in the sense of classificationrate (CR); (ii) combining the k-NN classifier with the voting rule method in order to search for the best number ofvoting vectors; and (iii) selecting relevant features for the task of appliances identification by using one of the relevantfeature selection algorithms based on mutual information. Results evaluated on the Plaid dataset clearly show that themean and standard deviation statistics combination gives the best CR of 92% with 500 features and gives the minimalcomputing time compared to the system based on HMM models. Moreover, combining the k-NN classifier with thevoting rule using the above features increases the CR up to 95%. Using this combination, the results also show that anincrease of the training dataset size further improves identification performance results in terms of precision, sensitivity,and F-score. A feature selection procedure based on joint mutual information strategy shows that using a selected subsetof five features is sufficient, giving similar CR results to those obtained using the total number of features, whatever thetraining dataset size.

PDF

___

[1] Nait-Meziane M, Hacine-Gharbi A, Ravier P, Lamarque G, Le Bunetel JC et al. HMM-based transient and steadystate current signals modeling for electrical appliances identification. In: 5th International Conference on Pattern Recognition Applications and Methods; Rome, Italy; 2016. pp. 670-677.
[2] Hacine-Gharbi A, Ravier P. Wavelet cepstral coefficients for electrical appliances identification using hidden Markov models. In: 7th International Conference on Pattern Recognition Applications and Methods; Funchal, Portugal; 2018. pp. 541-549.
[3] Nait-Meziane M, Hacine-Gharbi A, Ravier P, Lamarque G, Le Bunetel JC et al. Electrical appliances identification and clustering using novel turn-on transient features. In: 6th International Conference on Pattern Recognition Applications and Methods; Porto, Portugal; 2017. pp. 647-654.
[4] Thiruvaran T, Phung T, Ambikairajah E. Automatic identification of electric loads using switching transient current signals. In: IEEE TENCON Spring Conference; Sydney, Australia; 2013. pp. 252–256.
[5] Hsueh-Hsien C, Nguyen VL. Statistical feature extraction for fault locations in nonintrusive fault detection of low voltage distribution systems. Energies 2017; 10: 611.
[6] Esmael B, Arnaout A, Fruhwirth RK, Thonhauser G. A statistical feature-based approach for operations recognition in drilling time series. International Journal of Computer Information Systems and Industrial Management Applications 2013; 5: 454-461.
[7] Hacine-Gharbi A, Ravier P, Nemo F. Local and global feature selection for prosodic classification of the word’s uses. In: 6th International Conference on Pattern Recognition Applications and Methods; Porto, Portugal; 2017. pp. 711-717.
[8] Nilanon T, Yao J, Hao J, Purushotham S, Liu Y. Normal/abnormal heart sound recordings classification using convolutional neural network. In: Computing in Cardiology Conference; Vancouver, Canada; 2016. pp. 585-588.
[9] Grofman B, Owen G, Feld SL. Thirteen theorems in search of truth. Theory and Decision 1983; 15: 261-278.
[10] Xu L, Krzyzak A, Suen CY. Method of combining multiple classifiers and their application to handwriting recognition. IEEE Transactions on Systems, Man, and Cybernetics 1992; 22 (3): 418-435.
[11] Ho TK, Hull JJ, Srihari SN. Decision combination in multiple classifier systems. IEEE Transactions on Pattern Analysis and Machine Intelligence 1994; 16 (1): 66-75.
[12] Lam L, Suen CY. Application of majority voting to pattern recognition: an analysis of the behavior and performance. IEEE Transactions on Systems, Man, and Cybernetics 1997; 27 (5): 553-567.
[13] Zou J, Nagy G. A comparative study of local matching approach for face recognition. IEEE Transactions on Image Processing 2007; 16 (10): 2617-2628.
[14] Ogawara K, Fukutomi M, Uchida S, Feng Y. A voting-based sequential pattern recognition method. PLoS ONE 2013; 8 (10): e76980. doi: 10.1371/ journal.pone.0076980
[15] Ridi A, Hennebert J. Hidden Markov models for ilm appliance identification. Procedia Computer Science 2014; 32: 1010–1015.
[16] Mittelsdorf M, Hüwel A, Klingenberg T, Sonnenschein M. Submeter based training of multi-class support vector machines for appliance recognition in home electricity consumption data. In: Proceedings of SmartGreens; Aachen, Germany; 2013. pp. 151-158.
[17] De Paiva Penha D, Castro ARG. Convolutional neural network applied to the identification of residential equipment in nonintrusive load monitoring systems. In: 3rd International Conference on Artificial Intelligence and Applications; Chennai, India; 2017. pp. 11–21.
[18] Dat Nguyen T, Dong Do T, Ha Le M, Le NT, Benjapolakul W. Appliance classification method based on k-nearest neighbors for home energy management system. In: First International Symposium on Instrumentation, Control, Artificial Intelligence, and Robotics; Bangkok, Thailand; 2019. pp. 53-56.
[19] Xue B. Particle swarm optimisation for feature selection in classification. PhD, Victoria University, Wellington, New Zealand, 2014.
[20] Fleuret F. Fast binary feature selection with conditional mutual information. Journal of Machine Learning Research 2004; 5: 1531-1555.
[21] Duch W. Feature extraction: foundations and applications. In: Guyon I, Gunn S, Nikravesh M, Zadeh L (editors). Studies in Fuzziness & Soft Computing. Berlin, Germany: Springer, 2006, pp. 89–117.
[22] Hacine-Gharbi A, Deriche M, Ravier P, Harba R, Mohamadi T. A new histogram-based estimation technique of entropy and mutual information using mean squared error minimization. Computers and Electrical Engineering 2013; 39 (3): 918–933.
[23] Brown G, Pocock A, Lujan MJ, Zhao M. Conditional likelihood maximisation: a unifying framework for information theoretic feature selection. Journal of Machine Learning Research 2012; 13 (1): 27–66.
[24] Liu H, Motoda H. Feature Selection for Knowledge Discovery and Data Mining. Dordrecht, the Netherlands: Kluwer Academic Publishers, 1998.
[25] Hart GW. Residential energy monitoring and computerized surveillance via utility power flows. IEEE Technology and Society Magazine 1989; 8 (2): 12-16.
[26] Hart GW. Nonintrusive appliance load monitoring. Proceedings of the IEEE 1992; 80 (12): 1870–1891.
[27] Sultanem F. Procédé et appareil d’analyse de signaux de courant et de tension en vue du repérage de charges à usage domestique. Brevet Français FR 2645968, 1990 (in French).
[28] Bons M. Modèles à sources markovienne cachée appliqués à l’analyse non-intrusive des principaux usages domestiques de l’électricité. PhD, Université de Rennes I, Rennes, France, 1996 (in French).
[29] Onada T, Nakano Y, Yoshimoto K. System and method for estimating power consumption of electric apparatus, and abnormality alarm system utilizing the same. US patent 6,816,078 B2, 2004.
[30] Leeb S. A conjoint pattern recognition approach to nonintrusive load monitoring. PhD, Massachusetts Institute of Technology, Cambridge, MA, USA, 1993.
[31] Cole AI, Albicki A. Data extraction for effective nonintrusive identification of residential power loads. In: IEEE Instrumentation and Measurement Technology Conference; St. Paul, MN, USA; 1998. pp. 812-815.
[32] Drenker S, Kader A. Nonintrusive monitoring of electric loads. IEEE Computer Applications in Power 1999; 12 (4): 47-51.
[33] Chan WL, So ATP, Lai LL. Harmonics load signature recognition by wavelets transforms. In: International Conference on Electric Utility Deregulation and Restructuring and Power Technologies; London, UK; 2000. pp. 666-671.
[34] Baranski M, Voss J. Nonintrusive appliance load monitoring based on an optical sensor. In: IEEE Bologna Power Tech Conference; Bologna, Italy; 2003. p. 8.
[35] Patel SN, Robertson T, Kientz JA, Reynolds MS, Abowd GD. At the flick of a switch: detecting and classifying unique electrical events on the residential power line. In: International Conference on Ubiquitous Computing; Innsbruck, Austria; 2007. pp. 271–288.
[36] Chang HH, Lin CL, Yang HT. Load recognition for different loads with the same real power and reactive power in a nonintrusive load monitoring system. In: 12th IEEE International Conference on Computer Supported Cooperative Work in Design; Xi’an, China; 2008. pp. 1122–1127.
[37] Du Y, Du L, Lu B, Harley R, Habetler T. A review of identification and monitoring methods for electric loads in commercial and residential buildings. In: IEEE Energy Conversion Congress and Exposition; Atlanta, GA, USA; 2010. pp. 4527–4533.
[38] Zeifman M, Roth K. Nonintrusive appliance load monitoring: review and outlook. IEEE Transactions on Consumer Electronics 2011; 57 (1): 76-84.
[39] Carrie Armel K, Gupta A, Shrimali G, Albert A. Is disaggregation the holy grail of energy efficiency? The case of electricity. Energy Policy 2013; 52: 213-234.
[40] Najmeddine H, El Khamlichi Drissi K, Pasquier C, Faure C, Kerroum K et al. States of art on load monitoring methods. In: 2nd IEEE International Conference on Power and Energy; Johor Baharu, Malaysia; 2008. pp. 1256- 1258.
[41] De Paiva Penha D, Castro ARG. Home appliance identification for NILM systems based on deep neural networks. International Journal of Artificial Intelligence and Applications 2018; 9 (2): 69-80.
[42] Liu Q, Wu H, Liu X, Linge N. Single appliance recognition using statistical features based k-nn classification. In: International Conference on Cloud Computing and Security; Nanjing, China; 2017. pp. 631-640.
[43] Mallat S. A Wavelet Tour of Signal Processing. 2nd ed. San Diego, CA, USA: Academic Press, 1999.
[44] van Erp M, Vuurpijl L, Schomaker L. An overview and comparison of voting methods for pattern recognition. In: Proceedings of Eighth International Workshop on Frontiers in Handwriting Recognition; Niagara on the Lake, Canada; 2002. pp. 195-200.
[45] Bilski P, Winiecki W. Non-intrusive appliance load identification with the ensemble of classifiers. In: Proceedings of the International Workshop on Non-Intrusive Load Monitoring; Vancouver, Canada; 2016. pp. 1-5.
[46] Pascoal C, Oliveira MR, Pacheco A, Valadas R. Theoretical evaluation of feature selection methods based on mutual information. Neurocomputing 2016; 226: 168-181.
[47] Cover T, Thomas J. Elements of Information Theory. 2nd ed. New York, NY, USA: Wiley and Sons, 2006.
[48] Yang H, Moody J. Data visualization and feature selection: new algorithms for non-Gaussian data. Advances in Neural Information Processing Systems 1999; 12: 687-693.
[49] Gao J, Giri S, Kara EC, Berges M. Plaid: A public dataset of high-resolution electrical appliance measurements for load identification research: demo abstract. In: Proceedings of the 1st ACM Conference on Embedded Systems for Energy-Efficient Buildings; New York, NY, USA; 2014. pp. 198–199.
[50] Kelly J, Knottenbelt W. Neural NILM: Deep neural networks applied to energy disaggregation. In: 2nd ACM International Conference on Embedded Systems for Energy-Efficient Built Environments; Seoul, South Korea; 2015. pp. 55-64.
[51] Batra N, Parson O, Berges M, Singh A, Rogers A. A comparison of nonintrusive load monitoring methods for commercial and residential buildings 2014. arXiv preprint, arXiv: 1408.6595.
[52] Djordjević S, Simić M. Nonintrusive identification of residential appliances using harmonic analysis. Turkish Journal of Electrical Engineering & Computer Sciences 2018; 26 (2): 780-791.