A Multi-Objective Optimization Method Based on NSGA-II Algorithm for Electro-Thermal-Stress Collaborative Design of Intelligent Power Modules

Intelligent power module (IPM) has always been important for achieving efficient energy conversion due to its high integration level and high reliability. The IPM based on the insulated metal baseplate (IMB) can reduce parasitic inductance and improve its thermal cycle life. However, the low thermal conductivity of the insulating resin in the IMB makes the thermal performance of the module extremely sensitive to temperature changes. Strongly coupled with electrical-thermal-stress multi-physical characteristics, the impact of temperature sensitivity on performance is further exacerbated, limiting the heat dissipation efficiency and reliability of the module, especially in high power density applications. For that, this paper proposes a multi-objective optimization method based on the non-dominated sorting genetic algorithm II (NSGA-II), where the packaging structure and chip layout are ingeniously integrated into the Pareto frontier construction process. Simulations and experiments show that trade-offs are necessary since the three objectives cannot be optimized at once, with maximum reductions of 22.59% in loop inductance, 21.56% in chip junction temperature, and 33.43% in solder layer stress.