Jacobian regularized power electronic device monitoring

A method of characterizing a parameter (e.g., threshold voltage) of a power electronic device using an artificial intelligence (AI) model includes sampling measured parameter values (e.g., voltage, current) of the power electronic device during operation and characterizing the parameter of the power electronic device using the AI model in inference mode with the measured parameter values as inputs. The AI model is trained using a joint loss function including a Jacobian regularization term. The Jacobian regularization term may depend on the norm of at least one Jacobian of a corresponding set of training inputs. A power electronics system configured to perform the method includes the power electronic device and a computing system with a processor and memory storing the AI model. The computing system may be a microcontroller. The system may also include an analog-to-digital converter (ADC) circuit, such as in the microcontroller.