Datagestuurde benaderingen voor het voorspellen van hoogfrequente magnetische verliezen: van metingen tot generatieve AI

Machine Learning (ML) is nowadays a tool that allows systems to find solutions in a better and faster way after a learning process. Therefore, ML has the potential to improve the design of electric systems where many objective are required to be met. In the power converter design, there are many factors that affect a converter performance: variables, parameters, components to use, topologies, etc. These factors make design itself a complex activity as the possible solutions, resulting from the combination of each component and parameter, constitute a whole universe as such. With that amount of possible designs, computation becomes quite expensive and lasts extensive periods of time. The aim of this project is to create a machine learning environment where advance modelling is integrated for improving the design of power converters.