Optimal Design of High-temperature SiC Power Module Based on Gene Algorithm

Silicon carbide (SiC) devices possess outstanding high-temperature performance, allowing them to operate for extended periods in high-temperature environments while maintaining high reliability. However, most existing SiC modules directly adopt packaging technologies designed for silicon chips, which limits their ability to fully utilize the superior high-temperature performance of SiC devices. To reduce thermomechanical stress in high-temperature applications and enhance the reliability of SiC modules, an optimization method combining artificial neural networks (ANN) and Genetic Algorithm (GA) is proposed. The genetic algorithm is used to automatically solve the multi-objective model of maximum principal stress and junction temperature, thereby obtaining the optimal layout structure. Finally, a multi-chip parallel half-bridge power module with a rated voltage of 1200V and a rated current exceeding 300A is fabricated. The module's high-temperature performance is evaluated through power cycling tests.