摘要
熔融沉积成形(FDM)是快速成型(RP)最有发展前途的工艺之一,掌握提高成形件精度的控制方法是推广其应用的重要途径。在分析FDM成形件精度影响因素的基础上,提出应用误差反向传播(BP)神经网络建立预测精度模型的方法。将主要影响因素作为BP神经网络模型的输入参数,并根据最小预测误差选择输入层和中间层的维数,确定了BP模型结构。利用多组实验数据进行模型训练,建立了BP神经网络模型。模型预测与实验测量的对比结果表明,模型的预测误差在6%以内,具有很高的预测精度,可以指导实际应用。
Fused deposition modeling (FDM) is one of the most promising rapid prototype (RP) technologies. Methods for improving prototyping precision is crucial for its wide application. Based on analysis of factors of influencing the precision of prototype during FDM, a novel approach of establishing a BP neural network model to predict FDM prototyping precision was proposed in this paper. Some key factors were confirmed to be the feature pa- rameters of BP neural networks. The dimensional numbers of input layer and middle hidden layer were confirmed according to the minimum error, and then the BP model structure was fixed. The structure was trained by experimental data. The a model of BP neural network for predicting the precision of FDM prototype was finally constituted. The results show that the error can be controlled within 6% . The model can be used to guide practical application.
出处
《机械科学与技术》
CSCD
北大核心
2010年第3期399-403,共5页
Mechanical Science and Technology for Aerospace Engineering
基金
江苏省数字化制造技术重点建设实验室开放基金项目(HYDML-0806)资助