Appliance-level demand identification through signature analysis

Appliance specific load monitoring is very useful in energy management solutions that are becoming a challenging task with growing energy demand. It facilitates appliance recognition and load monitoring such that optimum resource utilization can be achieved by correct appliance scheduling. In this paper we present a study of non-intrusive load recognition using steady state appliance signatures for identifying commonly used household appliances. Current harmonics, active and reactive power components acquired from data loggers are used as appliance signature in this study. This analysis enables the capability of providing detailed information on appliances in use and consumers could benefit from customized energy management recommendations. Also, suppliers could implement smart metering technology introducing appliance level information as well. We propose algorithms for non-intrusive load recognition using combination of several methods and techniques. It was seen that a higher accuracy of identification could be achieved when a combination of techniques are used rather than using a single technique