Modelling variable loading through machine learning prediction model for an impact assessment concerning distribution networks

In this thesis we will deal with the application of new machine learning technologies to an Irish power grid model. The project involves the collection, preparation, cleaning and employement of historical load data of 4000 Irish customers to train the SVR and RFR algorithms. The impact of forecasting associated with the application of two models to and example of Irish distribuition Network and the implementation on system losses, is analysed.