Digitalization is supporting the electricity-energy system in the necessary effort to decarbonize Italy and other European Union states. The Covid-19 pandemic has shown how digitalization is fundamental in many fields, from healthcare to education, from work to electricity and energy. One of the pillars of digitalization is certainly the use of artificial intelligence and big data management/analysis techniques. This work presents some studies intended to promote the digitalization of the electrical system.

As part of the time series analysis, some clustering methodologies were explored with the aim of grouping consumption curves, both in terms of user categories and in terms of individual user behavior.

As far as the topological analysis of electrical networks is concerned, the field of transforming networks from the IEC CIM format to the GraphML format was explored, which is useful for applying graph analysis and graph machine learning techniques. The possibility of solving the powerflow calculation by means of techniques and platforms specific to the graphs world, such as GraphFrames and Tigergraph, was also studied. As part of the big data analysis, the results of the analysis of the data of the MV electricity distribution network for the areas of Brescia and Milan in the weeks of the national lockdown during the first wave of the Covid-19 pandemic are presented.

In the context of data streaming analysis, the potential of applying machine learning algorithms in real time is demonstrated. In particular, a prototype classification algorithm is described, capable of predicting the occupancy status of electric car charging stations.

To conclude, in the context of deep learning applied to the electrical system, the potential of training an artificial neural network to recognize flashovers on insulator strings filmed in the videos collected by the LANPRIS monitoring system is shown.