基于极差分析法与GA-ELM的电器连接器壳体注射成型工艺优化

以某电器连接壳体为例,借助Moldflow软件对正交试验方案组合进行模拟,对正交试验模拟结果进行极差分析,得到各工艺参数对塑件翘曲变形量的影响程度为:保压时间>模具温度>注射时间>熔体温度>保压压力。极差分析得到的最优工艺参数组合对应的翘曲变形量与正交试验方案中最小翘曲变形量相比降低了6.7%。关键点采用遗传算法优化后的预测模型(GA-ELM)对塑件翘曲变形量进行预测。由于传统极限学习算法(ELM)的权值和阈值随机产生,网络系统预测稳定性及精度较差,故通过GA全局寻优能力寻找最佳的权值和阈值,得到GA-ELM。选择正交试验前80%样本作为训练集训练ELM与GA-ELM模型,通过样本后20%作为测试集验证ELM与GA-ELM模型预测精度。对比分析可看到:使用GA-ELM预测模型比直接使用ELM预测模型预测结果有更高预测精度及稳定性。此GA-ELM模型可用来预测该塑件翘曲变形量。对同类模具设计优化提供一定的思路及理论参考。

Optimization of Injection Molding Process of Electrical Connector Shell Based on Range Analysis and GA-ELM

Taking an electrical appliance connecting shell as an example,Moldflow software was used to simulate the combination of orthogonal experiment schemes,carry out a range analysis of the orthogonal experiment simulation results,and obtain the influence degree of each process parameter on the warpage deformation of the plastic part as follows:holding time>mold temperature>injection time>melt temperature>holding pressure.The warpage deformation corresponding to the optimal process parameter combination obtained by the range analysis was reduced 6.7%comparing with the minimum warpage deformation in the orthogonal experiment scheme.The key point was to use genetic algorithm-extreme learning algorithm(GA-ELM)to predict the warpage deformation of plastic parts.Because the weights and thresholds of the traditional ELM(extreme learning algorithm)were randomly generated,the network system prediction stability and accuracy were poor,so the GA global optimization capability was used to find the best weights and thresholds to obtain the prediction model optimized by the GA-ELM.The 80%samples before the orthogonal experiment as the training set was chosen to train the ELM and GA-ELM models,and the 20%test set after the samples was selected to verify the prediction accuracy of the ELM and GA-ELM models.The comparative analysis shows that using the GA-ELM prediction model has higher prediction accuracy than directly using the ELM prediction model,and the GA-ELM prediction model has higher stability.This GA-ELM model can be used to predict the warpage of the plastic part.It can provide certain ideas and reference for similar mold design optimization.