D29.8 - A Computationally Efficient Method of Power Module Junction Temperature Estimate Considering Ripple Elements for Yearlong Reliability Simulation of Power Electronics Converters
Power module plays a crucial role in the reliability of power electronics converters, where junction temperature is the key stressor. Electro-thermal modeling method is widely applied for junction temperature estimation. Prior studies focus on developing various models to improve the accuracy, while few research emphasize the computation efficiency, which is equally important considering the reliability assessment with a yearlong mission profile. This digest proposes a computationally efficient algorithm to compute power module junction temperature with transient ripple components. First, a decoupled framework is introduced where the loss model, dependent only on junction temperature, is separated from the thermal model, eliminating the need for co-simulation. Second, an algorithm to decouple the average temperature through a 1D thermal-resistance model and ripple component through a thermal-impedance model, along with an adaptive simulation-time-resolution setting, is proposed. Finally, a case study on a T-type power module for solar applications demonstrates the effectiveness of the proposed method, achieving an nRMSE of 1.51% with a simulation time speedup of 700×.
