Non-intrusive Load Monitoring Using Water Consumption Patterns

In this paper, we tackle the problem of non-intrusive load monitoring (NILM). The purpose of algorithm NILM, is to disaggregate the total power consumption of a house-hold into individual consumption of appliances by analyzing changes in the power signal using analytical methods. One of the main challenges in this field is the existence of appliances consuming nearly-equal power. Different studies tried to extract and define specific features for these appliances to overcome this challenge. In this research, we incorporate the water consumption patterns of appliances into our analysis to separate otherwise-indistinguishable appliance. More precisely, we perform NILM via an event-based multi-label classification method in which water consumption patterns are employed to improve accuracy. To demonstrate the efficiency of the proposed method, numerical results are provided for four appliances of AMDP dataset.