摘要
为了提高气溶胶喷射3D打印质量的稳定性和准确性,建立基于Back Propagation(BP)神经网络的打印质量预测模型.该模型以鞘气流量、载气流量和打印速度为主要参数,并预测气溶胶喷印中的两个重要指标:线条宽度和线条粗糙度.同时,采用了K折交叉验证方法对神经网络模型进行训练,并对网络结构进行了评估.测试结果表明,该模型具有较高的预测精度和稳定性,能够准确地预测线条宽度和线条粗糙度.
To improve the stability and accuracy of aerosol jet 3D printing quality,a quality prediction model was developed based on Back Propagation(BP) neural networks.In this research,the main process parameters such as sheath gas flow rate,carrier gas flow rate,and printing speed were utilized as model inputs to predict two important indicators in aerosol jet printing:line width and line roughness.Additionally,K-fold cross-validation was employed to train the neural network model and evaluated its network structure.The test results demonstrate that the developed model exhibits high predictive accuracy and stability,enabling precise predictions of line width and line roughness.
作者
杨宝生
葛建军
张海宁
YANG Bao-sheng;GE Jian-jun;ZHANG Hai-ning(School of Computer Science and Information engineering,Hefei University of Technology,Hefei 230009,China;Suzhou Industrial Investment Holding Group Co.,Ltd.,Suzhou 234000,Anhui,China;Institute of Intelligent Manufacturing,School of Information Engineering,Suzhou University,Suzhou 234000,Anhui,China)
出处
《兰州文理学院学报（自然科学版）》
2024年第3期48-53,共6页
Journal of Lanzhou University of Arts and Science(Natural Sciences)
基金
安徽省教育厅安徽高校自然科学研究项目(KJ2021A1111)
宿州学院博士科研启动基金项目(2021BSK023)
宿州学院校级科研平台项目(2021XJPT51)。
关键词
气溶胶喷射打印
BP神经网络
线条形态
质量预测
aerosol jet printing
BP neural network
printed line morphology
printing quality prediction
