On the thermal management of a power electronics system: Optimization of the cooling system using genetic algorithm and response surface method

Abstract

The present study has designed a new type of cooling system consisted of jet impinging and combined with the pin fins and vortex generators to reduce the operating temperature of a power electronics module exposed to high heat flux. At the same time, its thermal/fluid performance was numerically evaluated. Then, the response surface method and genetic algorithm were employed to optimize the performance of the designed cooling systems. The design points for numerical simulations were determined by the design of experiment methods. The minimum average operating temperature of the power modules, as well as the minimum pumping power, were considered as the objective functions. Results indicated that the temperature of the power electronic module exhibits the highest and lowest sensitivity to the coolant flow rate and the vertical position of the vortex generator, respectively. At the optimal point proposed by the genetic algorithm, the average temperature of the insulated-gate bipolar transistor and Diode respectively were 75 C and 59 C, and the pumping power was 3.9 W. The comparison between the numerical results and the genetic algorithm's predictions indicated that the applied algorithm can predict the operating temperature with an error of less than 0.3% and the pumping power with an error of 10%. The homogeneous temperature distribution was created at the optimum point, its maximum value was 85, and was located on the insulated-gate bipolar transistor's center.