Temperature Evaluation of Power Devices in DC–DC Converters Based on Multitimescale Electrothermal Coupling Model

Junction temperature is a critical parameter for ensuring the safety and field reliability of power devices in operational dc–dc converters. Traditional electrothermal methods, which rely on iterative calculations coupling electrical and thermal models, face challenges due to the disparity in timescales, where power devices manifest dynamic operation on fast timescales, while thermal effects evolve over much longer timescales. To tackle this challenge, the article introduces a novel multitimescale electrothermal coupling model, which balances computational demand and accuracy in calculating the junction temperature of power devices under various operating conditions such as switching frequency, load profiles, and ambient temperature. The proposed model integrates a circuit model encompassing a power loss model and a state model for the power device, with a thermal model constructed using the finite element method. For reducing computational complexity, the proper orthogonal decomposition (POD) method is applied to the thermal model. The junction temperature is iteratively calculated using the multitimescale electrothermal model of the power device, which is validated in an LLC resonant dc–dc converter, demonstrating that the proposed method achieves high computational speed and accuracy. The results confirm that the multitimescale electrothermal model exhibits excellent performance in managing the complex thermal dynamics of power devices under real-world operating conditions.