Parameter Extraction for an ANN-assisted IGBT Model in Transient Simulations

Inhalt:
The development of power IGBT devices require an accurate transient model over multiple currents and voltages. With the Artificial Neural Network (ANN) assisted diode model and the modified Angelov-based capacitance model, a new IGBT model describes both the overshoot of collector current and collectoremitter tail voltage at different bias conditions. Given a completed parameter extraction workflow, a comprehensive good fitting of the transient current and voltage waveforms is verified at both the turn-on and turn-off process for currents from 30 to 600 A and voltages from 100 to 820 V.