气辅注射成型论文：复杂壳体类塑件气辅注射成型工艺优化研究doc.doc

气辅注射成型论文：复杂壳体类塑件气辅注射成型工艺优化研究 【中文摘要】气辅注射成型工艺参数的确定是一个相当重要和复杂的过程,它直接对制品的质量产生影响。如何筛选成型过程中重要的工艺参数并优化组合以获得较好制品质量是目前该领域一大难题。本文将21英寸彩电前壳作为研究对象,以正交试验设计方法为基础,利用遗传算法并结合径向基神经网络建立气辅注射成型工艺参数优化系统,可用于工艺参数组合的快速确定,为气辅注射成型过程中工艺参数优化提供了一种新的求解思路。本文主要研究工作如下：(1)综合考察分析了对制品质量影响严重的一些工艺参数；归纳总结了气辅注射成型常见缺陷及解决办法。(2)将正交试验设计方法与Mold flow 2010数值模拟相结合,对21英寸彩电前壳气辅注射成型工艺参数组合进行了初步优化。(3)利用多目标最优化方法建立了气辅注射成型工艺参数优化数学模型。(4)建立了工艺参数与制品质量指标间的径向基神经网络模型。以正交试验初步优化得到的工艺参数组合为依据,再按正交设计方法获得神经网络训练数据,通过网络训练,获得工艺参数与制品质量指标的映射关系,并验证了该模型的正确性。(5)选择多目标遗传算法策略,结合建立好的径向基神经网络模型,建立气辅注射成型工艺参数优化系统。通过该系统,得到了最佳工艺参数组合。最后通过数值模拟验证表明了基于神经网络—遗传算法的多目标优化方法用于气辅注射成型工艺参数优化的可行性和有效性。 【英文摘要】How to effectively and quickly determine process parameters of gas-assisted injection molding (GAIM) is a quite important and complex process, which directly affect on quality of products. Up to now, it is still a problem how to choose the key process parameters and optimize them to obtain good product performance.21 inches color front panel is taken as study object. On the basis of orthogonal experimental design, GAIM process parameters optimization system is established by using a hybrid system that combines radical basis function (RBF) neural network with genetic algorithm (GA). It can be used to determine optimum process parameter combination rapidly, and also provide a new solution of process parameter optimization for GAIM.The main contents of this thesis are listed as follows:(1) The influence factors of several quite important process parameters for the quality of GAIM part products are comprehensive studied. The common defects of GAIM are summarized and the relevant solutions are given.(2) The preliminary optimization is done based on orthogonal experimental design and mold-flow 2010 simulation methods.(3) the GAIM process parameters optimization mathematical model is established by using multi-objective optimization method.(4) The RBF neural network is setup to show the relationship of GAIM process parameters and the quality of part products. According