Optimization Design of SiC MOSFET Solid State Switch Heat Sink Based on Response Surface Methodology and Genetic Algorithm

Solid state switches with power devices as the core are widely used in power distribution systems, and thermal management is a necessary measure to ensure that solid-state switches achieve expected functions and improve reliability. As one of the important aspects of thermal management, the parameter design of the radiator directly affects the effectiveness of thermal management. To achieve high efficiency and lightweight of the heat dissipation system, heat dissipation optimization design is carried out for SiC MOSFET with the highest heat flux density in solid-state switches. Taking low heat resistance and lightweight of the radiator as optimization objectives, the number, height, and thickness of fins of the radiator are selected as independent variables. A case study is designed using response surface methodology, and results are obtained through COMSOL simulation. The quantitative relationship between optimization objectives and independent variables is obtained through regression analysis; Furthermore, the nondominated generative algorithm is used to obtain the Pareto front solution set for the regression equation, and the optimal parameter combination is selected by minimizing the thermal resistance and mass of the heat sink. In response to the heat dissipation requirements of a $1000 \mathrm{~V} / 50$ A DC solid-state device, the optimization results show that when the number of heat dissipation fins is 9, the fin height is 51.71 mm, and the fin thickness is $\mathbf{2. 5 0 1 ~ m m}$, the objective function achieves the optimal value, and the thermal resistance of the heat sink is reduced by $43 \%$ compared to expectations. This can effectively reduce the operating junction temperature of power devices, improve the reliability of solid-state switches, and form a universal optimization method for air cooling heat dissipation of solid-state switches.