Multi-Objective Parameter Optimization of Pulse Tube Refrigerator Based on Kriging Metamodel and Non-Dominated Ranking Genetic Algorithms

Abstract

Structure parameters have an important influence on the refrigeration performance of pulse tube refrigerators. In this paper, a method combining the Kriging metamodel and Non-Dominated Sorting Genetic Algorithm II (NSGA II) is proposed to optimize the structure of regenerators and pulse tubes to obtain better cooling capacity. Firstly, the Kriging metamodel of the original pulse tube refrigerator CFD model is established to improve the iterative solution efficiency. On this basis, NSGA II was applied to the optimization iteration process to obtain the optimal and worst Pareto front solutions for cooling performance, the heat and mass transfer characteristics of which were further analyzed comparatively to reveal the influence mechanism of the structural parameters. The results show that the Kriging metamodel presents a prediction error of about 2.5%. A 31.24% drop in the minimum cooling temperature and a 31.7% increase in cooling capacity at 120 K are achieved after optimization, and the pressure drop loss at the regenerator and the vortex in the pulse tube caused by the structure parameter changes are the main factors influencing the whole cooling performance of the pulse tube refrigerators. The current study provides a scientific and efficient design method for miniature cryogenic refrigerators.