基于Kriging代理模型及改进PSO的变模温注塑成型翘曲变形优化  

Warping Deformation Optimization of Dynamic Mold Temperature Injection Molding Based on Kriging Surrogate Model and Improved PSO Algorithm

作　　者：陈川 吕永锋[1] CHEN Chuan;LYU Yongfeng(Zhejiang Polytechnic University of Mechanical&Electrical Engineering,Hangzhou,Zhejiang 310053,China;Zhejiang University of Technology,Hangzhou,Zhejiang 310014,China)

机构地区：[1]浙江机电职业技术大学,浙江杭州310053 [2]浙江工业大学,浙江杭州310014

出　　处：《塑料》2024年第4期166-172,共7页Plastics

基　　金：浙江省自然科学基金(LGG19E050019);浙江省教育厅一般科研项目(Y202249583);校级科教融合课题(A-0271-22-016)。

摘　　要：为提升变模温注塑成型的产品质量,提高优化效率,将翘曲变形作为优化目标,模具温度、保压压力、熔体温度、保压时间、冷却时间等工艺参数作为变量因素,采用Moldflow模拟变量因素对优化目标的影响,通过最优拉丁超立方试验设计选出试验样本,建立Kriging代理模型,并且,检测代理模型拟合精度。采用改进PSO算法,得到最优翘曲模型预测值及最佳工艺参数组合。对比改进后的PSO与标准PSO,平均迭代次数、迭代时间约降低了45%,最优适应度、平均适应度、最差适应度、局部最优解及未实现收敛等参数均得到提升。通过实验验证可知,与优化前翘曲值(1.293 mm)相比,优化后翘曲值(0.7512 mm)降低了41.9%,误差为4.84%。结果表明,基于Kriging代理模型及改进PSO能有效地优化变模温成型工艺,降低了翘曲变形量,对于生产应用有指导意义。To improve the quality of Dynamic Mold Temperature injection molding products and the optimization efficiency,warpage deformation was chosen as the optimization target,the process parameters such as mold temperature,holding pressure,melt temperature,holding time and cooling time were chosen as variable factors.The influence of variable factors on the optimization goal was simulated by Moldflow.The test samples were selected by the optimal Latin hypercube experimental design.Then the Kriging surrogate model was established and the fitting accuracy of the surrogate model was tested.The improved PSO algorithm was used to obtain the optimal warpage model prediction value and the optimal process parameter combination.Improved PSO compared with standard PSO,average number of iterations and iteration time were reduced about 45%.Optimal fitness,average fitness,worst fitness,local optimal solution and unrealized convergence of the improved PSO were improved.Compared with the warpage value before optimization(1.293 mm),the warpage value of the improved PSO(0.7512 mm)was reduced by 41.9%,and the error was 4.84%,through experimental verification.The results showed that the dynamic mold temperature molding process could be effectively optimized based on Kriging surrogate model and improved PSO algorithm,and the warpage deformation could be reduced significantly.The optimization method also had guiding value for production application.

关 键 词：变模温注塑成型 改进PSO Kriging代理模型 翘曲 优化 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术] TQ320.66[自动化与计算机技术—计算机科学与技术]