陶瓷涂层材料多模态数据表征学习

doi:10.12066/j.issn.1007-2861.2383

上海大学学报(自然科学版) ›› 2022, Vol. 28 ›› Issue (3): 492-503.doi: 10.12066/j.issn.1007-2861.2383

陶瓷涂层材料多模态数据表征学习

武星¹^,²^,³(), 胡明涛¹, 丁鹏³^,⁴

1.上海大学计算机工程与科学学院, 上海 200444
2.之江实验室, 浙江杭州 311100
3.上海大学材料基因组工程研究院材料信息与数据科学中心, 上海 200444
4.上海大学理学院, 上海 200444

收稿日期:2022-03-28 出版日期:2022-06-30 发布日期:2022-05-27
通讯作者: 武星 E-mail:xingwu@shu.edu.cn
作者简介:武星(1980—), 男, 教授, 博士生导师, 博士, 研究方向为多模态数据挖掘、机器学习. E-mail: xingwu@shu.edu.cn
基金资助:
国家重点研发计划资助项目(2018YFB0704400);国家自然科学基金资助项目(52073168);云南省重大科技专项资助项目(202102AB080019-3);云南省重大科技专项资助项目(202002AB080001-2);之江实验室科研攻关资助项目(2021PE0AC02);上海张江国家自主创新示范区专项发展资金重大资助项目(ZJ2021-ZD-006)

Multi-modal data representation learning for ceramic coating materials

WU Xing¹^,²^,³(), HU Mingtao¹, DING Peng³^,⁴

1. School of Computer Engineering and Science, Shanghai University, Shanghai 200444, China
2. Zhejiang Laboratory, Hangzhou 311100, Zhejiang, China
3. Center of Materials Informatics and Data Science, Materials Genome Institute, Shanghai University, Shanghai 200444, China
4. College of Sciences, Shanghai University, Shanghai 200444, China

Received:2022-03-28 Online:2022-06-30 Published:2022-05-27
Contact: WU Xing E-mail:xingwu@shu.edu.cn

摘要/Abstract

摘要：

陶瓷涂层具有耐高温、耐腐蚀、耐磨损等特性, 其热膨胀系数和热导率等参数与其性能息息相关. 为解决陶瓷涂层性能实验成本高、测试困难等问题, 提出了陶瓷涂层材料多模态数据表征学习的性能预测方法. 首先利用高斯混合模型虚拟样本生成(Gaussian mixture model virtual sample generation, GMMVSG)算法生成符合真实陶瓷涂层数据分布的样本来扩充数据集; 其次利用卷积神经网络 VGG16 对陶瓷涂层的显微结构图像数据进行特征提取, 利用 TabNet 对结构化数据进行特征提取, 将提取到的图像数据特征与结构化数据特征融合; 最终根据多模态数据表征建立基于K-最近邻(K-nearest neighbor, KNN)、支持向量机回归(support vector regression, SVR)和多层感知机(multi-layer perceptron, MLP) 3 种机器学习算法的预测模型, 对陶瓷涂层的性能指标, 即热膨胀系数和热导率进行了预测. 实验结果表明: 提出的多模态数据表征学习模型的预测结果要优于单模态数据表征学习模型, 其中基于 MLP 算法训练的多模态数据表征学习模型对陶瓷涂层性能的预测效果最好; 在测试集中, 对陶瓷涂层热膨胀系数预测的平均绝对误差(mean absolute error, MAE)和均方误差(mean square error, MSE)分别为 0.026 6 和 0.001 7, 对热导率预测的 MAE 和 MSE 分别为 0.017 9 和 0.000 7. 所提出的陶瓷涂层材料多模态数据表征学习方法有效融合了结构化数据与非结构化数据, 联合学习了各模态数据的潜在共享信息, 成功提升了对陶瓷涂料层材料性能预测的准确度.

关键词: 陶瓷涂层, 高斯混合模型, 多模态数据表征, 机器学习算法

Abstract:

Ceramic coatings have excellent temperature resistance, corrosion resistance, and wear resistance, among other advantages. Their thermal expansion coefficient and thermal conductivity are two properties directly related to their performance. To address the issues of high experimental costs and challenging test conditions, we propose a method to predict the performance of ceramic coating materials based on multimodal data representation learning. To enlarge the data set, this method uses the Gaussian mixture model virtual sample generation (GMMVSG) algorithm to generate samples that match the real ceramic-coating data distribution. The method extracts micro-structural image data's features using the very deep convolutional neural network VGG16, extracts structured data's features using TabNet, and fuses the features of the extracted image data with those of the structured data. the final prediction models based on three machine learning algorithms-K-nearest neighbor (KNN), support-vector-machine regression (SVR), and multi-layer perceptron（MLP）—are established by using multimodal data representation to predict the thermal expansion coefficient and thermal conductivity of the performance index of ceramic coatings. The experimental results show that the proposed multimodal-data representation-learning model has a better prediction performance than that of the single-modal-data machine-learning model, and that the former model based on the MLP can most accurately predict ceramic coating performance. In the test set, the mean absolute and mean square errors for the prediction of the thermal expansion coefficient are 0.026 6 and 0.001 7, respectively, and the mean absolute and mean square errors for the prediction of thermal conductivity are 0.017 9 and 0.000 7, respectively. Our proposed learning method for multimodal data representation of ceramic coating materials effectively combines structured and unstructured data to learn both types of modal data with potentially shared information and successfully improves the pred.

Key words: ceramic coatings, Gaussian mixture models, multimodal data representation, machine learning algorithm

中图分类号:

TB 35

武星, 胡明涛, 丁鹏. 陶瓷涂层材料多模态数据表征学习[J]. 上海大学学报(自然科学版), 2022, 28(3): 492-503.

WU Xing, HU Mingtao, DING Peng. Multi-modal data representation learning for ceramic coating materials[J]. Journal of Shanghai University（Natural Science Edition）, 2022, 28(3): 492-503.

图/表 10

图1

图2

图3

图4

图5

表1

表2

表3

图6

图7

参考文献 22

[1]	倪嘉, 史昆, 薛松海, 等. 航空发动机用热障涂层陶瓷材料的发展现状及展望[J]. 材料导报, 2021, 35(S1): 163-168.
[2]	刘丹丹, 樊自拴. 超高温陶瓷涂层的研究进展[J]. 材料保护, 2020, 53(5): 105-110.
[3]	魏晨光. 陶瓷涂层物理性能评价的相对法模型及验证[D]. 北京: 中国建筑材料科学研究总院, 2015.
[4]	Lu Z, Jiang T, Min Z, et al. Review of thermal properties of graphite: coefficient of thermal expansion and thermal conductivity[J]. New Carbon Materials, 2022, 37(3): 1-12. doi: 10.1016/S1872-5805(22)60573-0
[5]	Liu B, Vu-Bac N, Zhuang X, et al. Stochastic full-range multiscale modeling of thermal conductivity of polymeric carbon nanotubes composites: a machine learning approach[J]. Composite Structures, 2022, 289: 115393. doi: 10.1016/j.compstruct.2022.115393
[6]	Baltrušaitis T, Ahuja C, Morency L P. Multimodal machine learning: a survey and taxonomy[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 41(2): 423-443. doi: 10.1109/TPAMI.2018.2798607
[7]	银正强. 基于多模态表征学习的源代码搜索研究[D]. 西安: 电子科技大学, 2021.
[8]	宋云峰, 任鸽, 杨勇, 等. 基于注意力的多层次混合融合的多任务多模态情感分析[J]. 计算机应用研究, 2022, 39(3): 716-720.
[9]	田彦涛, 黄兴, 卢辉遒, 等. 基于注意力与深度交互的周车多模态行为轨迹预测[J]. 吉林大学学报(工学版), 2022(1): 1-9.
[10]	薛景瑜. 基于深度学习的多模态阿尔兹海默症预测方法研究[D]. 济南: 齐鲁工业大学, 2021.
[11]	Maimaitijiang M, Sagan V, Sidike P, et al. Soybean yield prediction from UAV using multimodal data fusion and deep learning[J]. Remote Sensing of Environment, 2020, 237: 111599. doi: 10.1016/j.rse.2019.111599
[12]	Pakdamanian E, Sheng S, Baee S, et al. Deeptake: prediction of driver takeover behavior using multimodal data[C]// Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems. 2021: 1-14.
[13]	Liu T, Huang J, Liao T, et al. A hybrid deep learning model for predicting molecular subtypes of human breast cancer using multimodal data[J]. IRBM, 2022, 43(1): 62-74. doi: 10.1016/j.irbm.2020.12.002
[14]	McClenny L, Haile M, Attari V, et al. Deep multimodal transfer-learned regression in data-poor domains[DB/OL]. [2022-02-15]. https://doi.org/10.48550/arXiv.2006.09310.
[15]	Li L, Damarla S K, Wang Y, et al. A Gaussian mixture model based virtual sample generation approach for small datasets in industrial processes[J]. Information Sciences, 2021, 581: 262-277. doi: 10.1016/j.ins.2021.09.014
[16]	Sha'abani M, Fuad N, Jamal N, et al. KNN and SVM classification for EEG: a review[M]. New York: Springer, 2020: 555-565.
[17]	Awad M, Khanna R. Support vector regression[M]. Berkeley: Efficient Learning Machines, 2015: 67-80.
[18]	Taud H, Mas J F. Multilayer perceptron (MLP)[M]. New York: Springer, 2018: 451-455.
[19]	Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[DB/OL]. [2022-02-15]. https://web.cs.ucdavis.edu/-yjlee/teaching/ecs289g-winter2018/alexnet.pdf.
[20]	Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition[DB/OL]. [2022-02-15]. https://ui.adsabs.harvard.edu/abs/2014arXiv1409.1556S/abstract.
[21]	Arik S O, Pfister T. Tabnet: attentive interpretable tabular learning[C]// AAAI. 2021: 6679-6687.
[22]	Sharma K, Giannakos M. Multimodal data capabilities for learning: what can multimodal data tell us about learning?[J]. British Journal of Educational Technology, 2020, 51(5): 1450-1484. doi: 10.1111/bjet.12993

陶瓷涂层材料多模态数据表征学习

Multi-modal data representation learning for ceramic coating materials

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 22

相关文章 3

编辑推荐

Metrics

本文评价

[1]	武星, 高进, 丁鹏. 聚丙烯复合材料老化数据集成学习[J]. 上海大学学报(自然科学版), 2022, 28(3): 440-450.
[2]	侯丽敏, 张伟涛, 施丹, 刘焕成. 基于鼾声的声学特征估计睡眠呼吸暂停指数[J]. 上海大学学报(自然科学版), 2019, 25(4): 435-444.
[3]	侯丽敏缪炜. 基于韵律信息的语种辨识系统[J]. 上海大学学报(自然科学版), 2009, 15(5): 441-444.