Surrogate-based evaluation of power device TSOA in power electronic converters

Power electronic converters are the core equipment in renewable energy development and utilization. With the escalating demands for higher power density and reliability, fully unlocking their intrinsic performance potential has become imperative. Accurate characterization of the Thermal Safe Operating Area (TSOA) of power devices in converters is a prerequisite for unlocking their performance potential. However, existing TSOA evaluation methods entail extremely high computational costs, thereby constraining converter performance enhancement. In this work, an electrothermal coupling model of the converter is employed to construct a dataset linking operating parameters to the maximum junction temperature of power device chips. The Kriging surrogate model is then utilized to uncover the intrinsic mapping relationships within the dataset, thereby enabling rapid and high-resolution evaluation of the TSOA of power devices in converters.