Host
Suspendisse tristique vehicula ante tortor, ut auctor diam ornare ut. Pellentesque et ex et dolor tempor suscipit. Phasellus vel orci vel orci tincidunt interdum. Nunc quis viverra metus. In cursus tincidunt. Nam consequat nibh justo, non sollicitudin sem placerat vitae.
Client & task
Suspendisse tristique vehicula ante tortor, ut auctor diam ornare ut. Pellentesque et ex et dolor tempor suscipit. Phasellus vel orci vel orci tincidunt interdum. Nunc quis viverra metus. In cursus tincidunt. Nam consequat nibh justo, non sollicitudin sem placerat vitae.
Objectives
The rapid adoption of wide–bandgap semiconductor technologies—such as GaN and SiC devices—is driving power electronics toward increasingly higher switching frequencies, compact designs, and stringent efficiency requirements. As frequencies move into the MHz range, traditional modelling approaches for inductors and transformers fall short, especially when new magnetic materials and advanced winding structures are involved. Accurate prediction of high-frequency parasitic effects, losses, and dynamic behaviour becomes essential not only for performance optimisation but also for enabling reliable digital prototyping and system-level design.This project addresses these challenges by developing wideband, physics-informed reduced-order models (ROMs) of passive components that incorporate advanced magnetic materials, high-frequency parasitics, and efficient computational strategies. AI-based modelling and co-simulation with real converter environments play a key role in bridging the gap between material-level characterisation and full power converter design.
Objectives 1
Develop accurate, cost-effective wideband (DC to MHz) models for passive magnetic components (inductors, transformers), incorporating HF parasitics and new material characterisation.
Objectives 2
Accurate computation of eddy-current, dielectric, and iron losses in magnetic components, comparison with conventional empirical/analytical models. Improvement of the latter.
Objectives 3
Apply (AI-based) Reduced-Order Modelling (ROM) for efficient simulation.
Objectives 4
Co-simulation: test ROM in environments like MATLAB/Simulink against actual PE circuits.
Objectives 5
Validate models with real industrial applications, including GaN converters and Danfoss high-power converters.
Morbi nec accumsan sem dolor
Pellentesque et quam vel
Pellentesque et quam nisi sagittis vel ante
Pellentesque sagittis vel ante
Expected Results
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Phasellus et massa nec arcu efficitur ultrices. Proin laoreet neque eu ligula pretium, vel congue urna faucibus. Integer ut felis dui. Phasellus tincidunt lacus ac ex laoreet, ac ultricies quam condimentum. Vestibulum ante ipsum primis in est.
67%
ROM of passive components suitable for circuit simulation (co-simulation).
Sed fringilla gravida lorem, id rhoncus justo egestas sed. Nulla sagittis vel ante sit amet neque non tellus interdum tincidunt eget eu odio. Awesome!
