All news

We are back from Vancouver and IEEE ECCE 2026, where PELAB presented six papers on physics-informed AI for parameter and grid impedance estimation, model-free and parameter-free control, and ISOP LLC converters.

The most valuable part of the week, though, happened outside the presentation rooms. At ECCE, five minutes in the hallway can turn into a conversation with an industry team facing exactly the problem you have been studying in the lab. That happened to us more than once: we met researchers working on similar challenges from very different angles, met again with long-standing partners, and started discussions with companies interested in bringing these approaches closer to real products.

For a university lab this matters. Research that never leaves the lab, or is never tested against industrial constraints, risks staying on paper. Conferences like ECCE are where those two worlds actually meet.

Papers presented

  • Sequential Learning-Based PINN Modeling for Parameter Estimation on a Power Converter
  • Grid Impedance Estimation using Physics-Informed Neural Network Modeling
  • Hierarchical LSTM-Based Grid Impedance Estimation from Operating Transients for Grid-Following Inverters
  • Model-Free Predictive Control for High-Level Cascaded H-Bridge
  • Parameter-Free Control with Online Resonant Tank Identification for SR-DAB Converters
  • IVS and OCS Errors in Closed-Loop ISOP LLC Converters

Thank you to everyone who attended our sessions, asked hard questions or simply stopped for a coffee and a chat. If we did not manage to meet in Vancouver and you work on grid-connected converters, AI for power electronics or advanced converter control, get in touch.