Multi-objective design optimization of insulated gate bipolar transistor structures considering ultrasonic welding packaging

This article proposes a multi-objective design optimization (MODO) framework to improve the performance (<i>i.e.</i> fatigue resistance, electrical conductivity and lightweight characteristics) of insulated gate bipolar transistor (IGBT) modules during high-frequency ultrasonic welding (USW) packaging. A high-fidelity USW coupling model is established to explore coupling mechanisms and accurately predict the welding temperature and penetration depth. To mitigate the computational burden, a novel surrogate modelling approach employing an improved sequential sampling strategy is proposed. A new non-dominated sorting genetic algorithm-II incorporating kernel density estimation is proposed to solve the MODO problem effectively. The high accuracy of the coupling model in predicting the welding temperature and penetration depth are validated by experimental results. Among all the MODO Pareto solutions, the maximum fatigue damage and conductivity loss of the IGBT structure are reduced by 87.57% and 9.65%, respectively, which demonstrates the advantages of the proposed MODO framework.