基于关联度的注塑工艺参数稳健优化设计研究

Research on Robust Design of Craft Parameters in Plastic Injection Based on Relation Degree

在注塑成形过程中,考虑塑料物性参数等随机因素波动的影响,选取最大翘曲变形量为成形质量指标,提出了基于关联度的变量筛选、神经网络响应面及蚁群优化的注塑成形工艺参数稳健设计方法。基于左侧距和位值的定义,提出了关联度的计算公式;建立了基于关联度的稳健优化模型;提出了混合交叉变异的蚁群算法。将基于关联度的稳健设计方法应用于某遥控器外壳的注塑成形工艺优化,数值模拟及实际注塑试验表明,该方法减小了最大翘曲量的标准差,并有效地提高了塑件的尺寸公差。

In plastic injection process, the fluctuating effect of random factors, such as the property parameters of plastic materials, were considered. The maximum warpage quantity was selected as a measurement of molding quality, and robust design method was put forward based on variable filter using relation degree, neural network and ant colonies algorithm. Firstly, the calculation formula of relation degree was put forward based on the definitions of left side--distance and location value, the robust optimum model of craft parameters in plastic injection process was put forward based on relation degree, the ant colonies algorithm with crossover and mutation was also put forward. The proposed robust design method based on relation degree was applied to the craft optimization of a remote controller in plastic injection process. It shows that the standard deviation of maximum warpage is decreased and the dimensional tolerance of plastic product is improved by this method, according to CAE and actual injection experiments.