CAJ | 학술논문

针对气辅成型工艺设定及控制的复杂性,使用Kriging代理模型近似拟合气辅成型工艺参数与气体穿透长度之间的非线性函数关系,建立起Kriging代理模型与APSO算法相结合的优化策略.应用所建立的优化方案,以气体的穿透深度为优化的目标,模拟了某汽车后视镜的气辅成型工艺优化过程.算例表明,基于Kriging代理模型与APSO算法耦合的优化策略可以在小样本情况下获取较高的求解精度和较快的收敛速度.
침대기보성형공예설정급공제적복잡성,사용Kriging대리모형근사의합기보성형공예삼수여기체천투장도지간적비선성함수관계,건립기Kriging대리모형여APSO산법상결합적우화책략.응용소건립적우화방안,이기체적천투심도위우화적목표,모의료모기차후시경적기보성형공예우화과정.산례표명,기우Kriging대리모형여APSO산법우합적우화책략가이재소양본정황하획취교고적구해정도화교쾌적수렴속도.
For the gas-assisted injection molding process, it is hard to find the explicit relationship between the depth of gas penetration and processing parameters.In this paper, an optimization strategy based on Kriging metamodel and adaptive particle swarm optimization (APSO) algorithm was developed.The Kriging metamodel was adopted to approximate the nonlinear relationship between processing parameters and the depth of gas penetration, and the APSO algorithm was adopted to carry out the optimization procedure.Moreover, such strategy was used for the parameter optimization of a real auto rear view mirror, in which the depth of gas penetration was set as the optimization objective.It was shown that the presented optimization strategy based on Kriging metamodel and APSO algorithm could achieve higher solving accuracy and fast convergence rate for those small sample size problems.