NILM Scheme Enabled Wide Range of Electrical Load Identification Using SST and SVM Techniques

The accurate electrical load identifications in residential sectors may help to implement type of use (TOU) tariff for better energy management in today's era of everincreasing demands of energy. For the different residential electrical loads, the sample values of the voltage and current signals are first acquired to generate various features using Sample Shifting Technique (SST). On these features Support Vector Machine (SVM) is applied to train and identify the loads. For different electrical home appliances having a wide range of power the proposed technique is applied, and loads are successfully identified. Various performance metric parameters determined to determine the accuracy of the proposed technique. A hardware based acquired data is used to verify the effectiveness of the proposed system.