Modelling and forecasting of electrical consumption for demand response applications

The emergence of concepts, policies and applications like smart grid, energy communities, carbon footprint reduction, energy-efficient consumption, demand response, demand side management enables the development and implementation of algorithms and procedures dedicated to the analysis of domestic energy consumption. In this paper a method of modelling and forecasting is presented and an optimization of electric energy consumption using machine learning and a cost function based on local peaks identified in load curves is proposed. This approach supports the effort of balancing the difference between generation and consumption curves by enabling the DSO to identify peak intervals and routes towards optimization.