Optimization Model for Demand Flexibility Based on a Non-Intrusive Load Disaggregation Tool

In the last years there has been a growing spread of smart meters that measure and communicate residential electricity consumption, allowing the development of new energy efficiency services. An interesting application involves the disaggregation of the main home appliances from the aggregated consumption signal. This is essential, to make predictions and optimizations for the implementation of Demand Side Management (DSM) strategies, also applicable to Energy Communities perspective. This paper presents an infrastructure and a set of algorithms to collect data from Italian second-generation smart meters and break down the total power measured by them into those used by main individual appliances. By using Non-Intrusive Load Monitoring (NILM) techniques, the proposed methodology can identify when a specific appliance is operating and create an appliance's properties database through unsupervised clustering algorithms applied to the detected devices. The system is tested using data collected from three households in Italy and results are reported in the paper. As a further development of this work, a demonstration of the application of the proposed NILM algorithm outputs as inputs for the optimization of houses power demand is performed, in order to maximize the shared energy in the context of Renewable Energy Communities.