基于径向基网络预测PLA 3D打印零件的磨损性能  

Predicting the Wear Performance of 3D Printed Parts of PLA by Radial Basis Function

作　　者：张艳英 常亮[2] ZHANG Yanying;CHANG Liang(Puyang Medical College,Puyang,Henan 457000,China;Zhengzhou University,Zhengzhou,Henan 450001,China)

机构地区：[1]濮阳医学高等专科学校,河南濮阳457000 [2]郑州大学,河南郑州450001

出　　处：《塑料》2025年第2期135-141,147,共8页Plastics

基　　金：河南省科技厅科技攻关项目(212102210565)。

摘　　要：聚乳酸(PLA)零件磨损性能在生物医学领域的应用十分重要。采用响应面模型(RSM)研究了熔融沉积工艺(FDM)中挤出温度、打印方向及层高对PLA零件比磨损率的影响,利用径向基网络(RBF)预测了比磨损率。结果表明,挤出温度和层高是影响比磨损率的显著主效应,打印方向×层高为显著的交互效应,挤出温度^(2)、打印方向^(2)及层高^(2)为显著的二阶效应。最优RBF网络结构为3-18-1,预测比磨损率与实验值回归系数R为0.9836,均方误差(MSE)为1.02×10^(-3),与响应面模型(1.31)相比,具有更好的预测性能。The wear performance of polylactic acid(PLA)parts was crucial for biomedical application.The response surface model(RSM)was used to study the effects of extrusion temperature,print orientation,and layer height on the specific wear rate of PLA parts in the melt deposition modeling(FDM),and a radial basis function network(RBF)was used to predict the specific wear rate of parts.The results indicated that extrusion temperature and layer height were significant main effects on specific wear rate,and print direction×layer height was significant interaction effect,while extrusion temperature^(2),print direction^(2),and layer height^(2) were significant second-order effects.The optimal RBF network was 3-18-1,with a regression coefficient R of 0.9836 between predicted specific wear rate and experimental values,and mean square error(MSE)of 1.02×10^(-3),which was better than the MSE of 1.31 of the response surface model,demonstrating excellent predictive performance.

关 键 词：响应面 径向基网络 聚乳酸 熔融沉积 磨损性能 

分 类 号：TQ323.4[化学工程—合成树脂塑料工业] TQ391.7