A Machine Learning Approach for the Identification of Photovoltaic and Electric Vehicle Profiles in a Smart Local Energy System

Smart local energy systems (SLESs) focus on in-tegrating more renewable energy sources to the electrical distribution network. Digitalization of SLESs can be achieved through digital twins (DTs). The DT is a virtual replica of the physical energy system. The accurate recognition of the availability, capacity, and quantity of the distributed energy resources (DERs) connected to the electrical distribution network enables the development of a comprehensive DT of the local energy system. This research study proposes to use Machine Learning (ML) algorithms to identify the availability of DERs in the local area and it contributes to the field by providing a comparative analysis of the results of different classification based ML algorithms when recognising availability of solar energy generation and electric vehicles (EVs) in aggregated grid data using a comprehensive labelled dataset. The results presented in this paper show that with the application of sliding window (SW) method for photovoltaic (PV) and EV identification from aggregated data, the accuracy, precision, recall and F1 score metrics of some ML algorithms can reach an approximate value of 98 % reflecting the effectiveness of the SW method and the respective ML algorithms.