为解决控制图质量异常模式识别中, 实时质量数据呈现出高维非线性等复杂特征导致模型过拟合以及失真等问题, 提出一种基于融合特征 分布随机近邻嵌入(-distributed stochastic neighbor embedding, -SNE) 降维的控制图质量异常模式识别方法. 首先, 从生产过程动态数据流中提取其统计特征、形状特征并与原始特征进行融合, 形成动态数据流的高维融合特征; 然后利用 -SNE 算法对融合特征进行降维, -SNE 算法能够有效地处理线性和非线性数据, 并产生更有意义的聚类; 进而利用一维卷积神经网络(one-dimensional convolutional neural networks, 1DCNN)作为分类器实现复杂产品制造过程的质量异常模式识别; 最后, 通过仿真实验将本文所提方法与单一类型特征方法、融合特征方法以及融合特征主成分分析法(principal component analysis, PCA)、核主成分分析(kernel PCA, KPCA) 和局部线性嵌入算法(locally linear embedding, LLE) 降维方法的识别模型进行比较, 并利用锂离子电池极片制造过程为例进一步说明本文模型的有效性与实用性. 仿真与实例结果表明, 本文所提算法具有更高的识别效率和精度, 特别适用于处理在复杂产品制造过程背景下的高维非线性数据.
Abstract
To address the issues of overfitting and distortion caused by the complex characteristics of high-dimensional nonlinear real-time quality data in control chart quality abnormal pattern recognition, a control chart quality abnormal pattern recognition method based on fused feature -distributed stochastic neighbor embedding (-SNE) dimensionality reduction is proposed. Firstly, statistical features and shape features are extracted from the dynamic data stream of the production process, and fused with the original features to form high-dimensional fused features of the dynamic data stream. Then, the fused features are reduced using the -SNE algorithm, which effectively handles both linear and nonlinear data and produces more meaningful clustering. Furthermore, a one-dimensional convolutional neural networks (1DCNN) is used as a classifier to achieve quality abnormal pattern recognition in complex product manufacturing processes. Finally, through simulation experiments, the proposed method is compared with single-type feature methods, fused feature methods, and fused feature principal component analysis (PCA), kernel PCA(KPCA) and locally linear embedding (LLE) dimensionality reduction methods for the recognition models. The effectiveness and practicality of the proposed model are further demonstrated using the example of the lithium-ion battery electrode manufacturing process. Simulation and case study results show that the proposed algorithm has higher recognition efficiency and accuracy, particularly suitable for handling high-dimensional nonlinear data in the context of complex product manufacturing processes.
关键词
控制图 /
模式识别 /
复杂产品制造过程 /
分布随机近邻嵌入(-SNE) /
卷积神经网络(CNN)
{{custom_keyword}} /
Key words
control chart /
pattern recognition /
complex product manufacturing process /
-distributed stochastic neighbor embedding (-SNE) /
convolutional neural networks (CNN)
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 郭凯明, 陈昊, 颜色. 贸易成本与中国制造[J]. 金融研究, 2022(3): 1-19. Guo K M, Chen H, Yan S. Trade cost and “Made in China”[J]. Journal of Financial Research, 2022(3): 1-19.
[2] 冯烽, 熊昭. 装备制造业发展的国际经验及对中国的启示[J]. 经济学家, 2023, 4(4): 56-66. Feng F, Xiong Z. International experience in the development of equipment manufacturing industry and its enlightenment to China[J]. Economist, 2023, 4(4): 56-66.
[3] 王海宇. 考虑多异常波动的时变参数EWMA图多目标优化设计[J]. 系统工程理论与实践, 2021, 41(11): 3031- 3042. Wang H Y. Multi-objective optimization design of EWMA chart with time-varying parameters for multiple assignable shifts[J]. Systems Engineering—Theory & Practice, 2021, 41(11): 3031-3042.
[4] 徐旭东, 马立乾. 基于迁移学习和卷积神经网络的控制图识别[J]. 计算机应用, 2018, 38(2): 290-295. Xu X D, Ma L Q. Control chart recognition based on transfer learning and convolution neural network[J]. Journal of Computer Applications, 2018, 38(2): 290-295.
[5] 潘婧, 柴洪峰, 孙权, 等. 超高维删失数据的联合特征筛选方法研究[J]. 系统工程理论与实践, 2023, 43(1): 169-190. Pan J, Chai H F, Sun Q, et al. Joint feature screening method for ultrahigh dimensional censored data[J]. Systems Engineering—Theory & Practice, 2023, 43(1): 169-190.
[6] Wang H. A new method for feature selection in pattern recognition of statistical process control chart[C]// 2022 China Automation Congress (CAC). IEEE, 2022: 190-193.
[7] Pham D T, Wani M A. Feature-based control chart pattern recognition[J]. International Journal of Production Research, 1997, 35(7): 1875-1890.
[8] Hassan A, Baksh M S N, Shaharoun A M, et al. Improved SPC chart pattern recognition using statistical features[J]. International Journal of Production Research, 2003, 41(7): 1587-1603.
[9] Ranaee V, Ebrahimzadeh A, Ghaderi R. Application of the PSO-SVM model for recognition of control chart patterns[J]. ISA Transactions, 2010, 49(4): 577-586.
[10] 刘玉敏, 赵哲耘. 基于特征选择与SVM的质量异常模式识别[J]. 统计与决策, 2018, 34(10): 47-51. Liu Y M, Zhao Z Y. Quality anomaly pattern recognition based on feature selection and SVM[J]. Statistics & Decision, 2018, 34(10): 47-51.
[11] Ebrahimzadeh A, Ranaee V. Control chart pattern recognition using an optimized neural network and effcient features[J]. ISA Transactions, 2010, 49(3): 387-393.
[12] Zhang M, Yuan Y, Wang R, et al. Recognition of mixture control chart patterns based on fusion feature reduction and fireworks algorithm-optimized MSVM[J]. Pattern Analysis and Applications, 2020, 23: 15-26.
[13] 郑轶松, 王谦, 何曙光. 基于主成分分析的半导体封装质量控制系统[J]. 系统工程, 2009, 27(1): 108-112. Zheng Y S, Wang Q, He S G. Quality controlling of semiconductor assembly based on principal components analysis[J]. Systems Engineering, 2009, 27(1): 108-112.
[14] 刘玉敏, 张帅. 基于MPCA-LSSVM的生产制造过程异常监控模型[J]. 统计与决策, 2018, 34(13): 81-84. Liu Y M, Zhang S. Abnormal monitoring model of manufacturing process based on MPCA-LSSVM[J]. Statistics & Decision, 2018, 34(13): 81-84.
[15] Guan F, Cheng L. Abnormal quality pattern recognition of industrial process based on multi-support vector machine[J]. Journal of Software, 2018, 13(9): 506-519.
[16] Zhang M, Zhang X, Wang H, et al. Features fusion exaction and KELM with modified grey wolf optimizer for mixture control chart patterns recognition[J]. IEEE Access, 2020, 8: 42469-42480.
[17] 常志朋, 陈闻鹤, 王治莹. 核主成分马田系统及其应用[J]. 系统工程理论实践, 2021, 41(9): 2447-2456. Chang Z P, Chen W H, Wang Z Y. Kernel principal component Mahalanobis-Taguchi system and its application[J]. Systems Engineering—Theory & Practice, 2021, 41(9): 2447-2456.
[18] 赵春华, 汪成康, 华露, 等. 基于融合特征约减和支持向量机的控制图模式识别[J]. 中国机械工程, 2017, 28(8): 930- 935. Zhao C H, Wang C K, Hua L, et al. Control chart pattern recognition based on fusion feature reduction and SVM[J]. China Mechanical Engineering, 2017, 28(8): 930-935.
[19] 殷秀丽, 谢丽蓉, 杨欢, 等. 特征选择与t-SNE结合的滚动轴承故障诊断[J]. 机械科学与技术, 2023, 42(11): 1784- 1793. Yin X L, Xie L R, Yang H, et al. Rolling bearing fault diagnosis combined feature selection with t-distributed stochastic neighbor embedding[J]. Mechanical Science and Technology for Aerospace Engineering, 2023, 42(11): 1784-1793.
[20] Electric W. Statistical quality control handbook[M]. Indianapolis: Western Electric Company, 1958.
[21] 刘玉敏, 刘莉. 基于统计特征的动态过程质量异常模式识别[J]. 统计与决策, 2017(19): 32-36. Liu Y M, Liu L. Dynamic process of quality abnormal pattern recognition based on statistical features[J]. Statistics & Decision, 2017(19): 32-36.
[22] 王海燕, 侯琳娜. 基于随机森林的统计控制图模式识别研究[J]. 工业工程, 2019, 22(5): 118-125. Wang H Y, Hou L N. A study of pattern recognition of statistical control chart based on random forest[J]. Industrial Engineering Journal, 2019, 22(5): 118-125.
[23] 高淑芝, 王拳, 张义民. EEMD熵特征和t-SNE相结合的滚动轴承故障诊断[J]. 机械设计与制造, 2023(6): 229-233. Gao S Z, Wang Q, Zhang Y M. Rolling bearing fault diagnosis combining EEMD entropy feature and t-SNE[J]. Machinery Design & Manufacture, 2023(6): 229-233.
[24] Van der Maaten L, Hinton G. Visualizing data using t-SNE[J]. Journal of Machine Learning Research, 2008, 9(86): 2579-2605.
[25] 王宁, 李盼盼, 赵哲耘, 等. 基于卷积神经网络的智能制造过程质量异常诊断[J]. 运筹与管理, 2022, 31(6): 220-225. Wang N, Li P P, Zhao Z Y, et al. Quality abnormal recognition model based on convolutional neural network[J]. Operations Research and Management Science, 2022, 31(6): 220-225.
[26] 赵哲耘, 刘玉敏, 王宁. 基于改进卷积神经网络的动态过程质量异常模式识别[J]. 工业工程与管理, 2021, 26(4): 69-76. Zhao Z Y, Liu Y M, Wang N. Dynamic process quality abnormal patterns recognition method based on improved convolutional neural network[J]. Industrial Engineering and Management, 2021, 26(4): 69-76.
[27] 张珂, 宋文丽, 石怀涛, 等. 基于改进核主元分析的故障检测方法研究[J]. 控制工程, 2017, 24(2): 418-424. Zhang K, Song W L, Shi H T, et al. Fault detection based on improved kernel principal component analysis[J]. Control Engineering of China, 2017, 24(2): 418-424.
[28] 李正欣, 郭建胜, 毛红保, 等. 多元时间序列相似性度量方法[J]. 控制与决策, 2017, 32(2): 368-372. Li Z X, Guo J S, Mao H B, et al. Similarity measure for multivariate time series[J]. Control and Desicion, 2017, 32(2): 368-372.
[29] Fan M, Zhang X, Hu J, et al. Adaptive data structure regularized multiclass discriminative feature selection[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 33(10): 5859-5872.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金重点资助项目(U1904211);河南省高等学校青年骨干教师培养项目(2021GGJS006);河南省高校哲学社会科学创新人才支持计划(2023-CXRC-19);郑州大学人文社会科学优秀青年科研团队(2023-QNTD-01)
{{custom_fund}}
PDF(1022 KB)
