Surrogate-Assisted Multi-Objective Optimization of GaN HEMTs Using a TCAD DoE Dataset
An end-to-end workflow that couples a space-filling TCAD design of experiments dataset with hybrid surrogate models and surrogate-assisted NSGA-II optimization to rapidly map the breakdown voltage and on-resistance trade-off under realistic electrical constraints is presented, demonstrating a reliable and efficient fra...