| Research topics |
Real-Time Digital Twins for Power Electronic Systems.
PhD candidates are invited to join a research programme focused on modelling, diagnostics, and control of power electronic systems using time-series data, geometry reconstruction of hidden dynamics, and lightweight, interpretable digital-twin models.
The programme focuses on developing models that are not simply large neural networks, but structured, computationally efficient models with potential for real-time operation. The central idea is to reconstruct the hidden dynamic state of a converter from current, voltage, thermal, and control-signal waveforms, and then use this information for modelling, diagnostics, control, and safety assessment.
Possible PhD directions include explainable discrete emulators for digital twins, component degradation diagnostics, adaptive predictive control, and stability and safety of digital twins operating inside control loops.
The topics are suitable for candidates interested in power electronics, control theory, dynamic modelling, signal analysis, machine learning, time-series analysis, diagnostics, HIL, DSP, FPGA, or embedded systems. Candidates are not expected to know all of these areas. More important are systematic work, scientific curiosity, and motivation to work at the intersection of power electronics, nonlinear dynamics, time-series data, and AI.
|