关键词: 适应度, 数值模拟, Kriging模型, 拉丁超立方抽样

Abstract: A wind deflector of air conditioners was taken as a typical example for injection molding in this paper. Based on the different processing parameters of air conditioner deflectors, three different kernel functions were adopted to figure out the fitness examination, and mold temperature, melt temperature, packing time, cooling time, pressure, pressure time was selected as significant factors. The Kriging model was constructed on the basis of Gaussian, generalized exponential and exponential kernel functions, and then the fitness function was established. The results suggested that the Gaussian function was the most appropriate kernel one for the simulation of forming process of air conditioner deflectors.

Key words: fitness, numerical simulation, Kriging model, Latin hypercube sampling method