Classification of ON-OFF states of appliance consumption signatures

Nonintrusive load monitoring (NILM) is a procedure for the analysis of the changes in the power (current and voltage) that goes into households and classifying the appliances used in the house according to their individual energy consumption. Utility companies use smart electric meters accompanied with NILM to examine the particular uses of electric power in households. Focus of this paper is on the analysis of the “ACS-F2 Database of Appliance Consumption Signatures”. The challenge lies in predicting the states of the electrical devices based on the measuring data which had been previously stored. Machine learning techniques have demonstrated to be effective in classification and pattern recognition tasks. In this paper, different algorithms implemented in the WEKA software are going to be used for the classification.