面向制造过程数据的两阶段无监督特征选择方法

doi:10.3901/JME.2019.17.133

机械工程学报 ›› 2019, Vol. 55 ›› Issue (17): 133-144.doi: 10.3901/JME.2019.17.133

• 数字化设计与制造 • 上一篇

面向制造过程数据的两阶段无监督特征选择方法

张洁^1,2, 盛夏¹, 张朋¹, 秦威¹, 赵新明¹

1. 上海交通大学机械与动力工程学院上海 200240;
2. 东华大学机械工程学院上海 201620

收稿日期:2018-09-19 修回日期:2019-02-03 发布日期:2020-01-07
通讯作者: 张洁(通信作者),女,1963年出生,教授。主要研究方向为制造系统的建模、优化调度与控制、制造信息工程和智能制造系统。E-mail:mezhangjie@dhu.edu.cn
作者简介:盛夏,男,1993年出生。主要研究方向为数字化制造与智能制造。E-mail:shengxia0111@sjtu.edu.cn;张朋,男,1988年出生,博士研究生。主要研究方向为制造系统优化调度与控制E-mail:pgnahz@qq.com;秦威,男,1985年出生,副教授。主要研究方向为复杂(制造)系统的建模、控制与优化,制造大数据分析,制造智能、人工智能在制造业的应用。E-mail:wqin@sjtu.edu.cn;赵新明,女,1960年出生,副教授。主要研究方向为机械设计及图形处理技术。E-mail:zhaoxm@sjtu.edu.cn
基金资助:
国家自然科学基金资助项目（U1537110，51435009）。

Two-stage Unsupervised Feature Selection Method Oriented to Manufacturing Procedural Data

ZHANG Jie^1,2, SHENG Xia¹, ZHANG Peng¹, QIN Wei¹, ZHAO Xinming¹

1. Institute of Intelligent Manufacturing and Information Engineering, Shanghai Jiao Tong University, Shanghai 200240;
2. College of Mechanical Engineering, Donghua University, Shanghai 201620

Received:2018-09-19 Revised:2019-02-03 Published:2020-01-07

摘要/Abstract

摘要： 现代化制造车间无时无刻不在产生大量数据，其中绝大部分以无标签结构化原始数据的形式存储在现代化制造企业的工业大数据平台中。这些制造数据一方面具有很大的潜在价值，另一方面因为其具有高噪声、高冗余性的特点，难以直接分析与利用。因此，针对制造过程原始数据的特点，以去除制造数据冗余性、挖掘原始数据局部结构为目的，提出一种两阶段无监督特征选择方法。该方法的第一阶段采用遗传算法产生的原始数据的低维子集作为径向基神经网络（Radial basis fuction neural network，RBFNN）的输入，利用RBFNN复现原始数据的全部维度，并以降维率及复现精度作为遗传算法（Genetic algorithm，GA）的适应度函数，通过GA多次迭代学习高维特征的低维表示，删除原始数据集中的冗余特征与噪声特征。第二阶段采用拉普拉斯特征得分（Laplacian score，LS）逐维评价剩余特征对于反映数据局部几何结构的作用，挖掘对改善分类性能更有效的特征。通过与LS等无监督特征选择算法对比，验证了提出的两阶段无监督特征选择方法能够有效降低制造数据的冗余性，并提高数据的分类性能。

关键词: 无监督特征选择, 遗传算法, 径向基神经网络, 拉普拉斯得分, 制造过程数据

Abstract: In a modernized manufacturing workshop, myriads of data are incessantly produced and a large part of those are stored in the industrial big data platform of the modern manufacturing enterprise in the form of structuralized unlabeled raw data. Those manufacturing data are of great latent exploitative value, because of their characteristics of high-noise and high-redundancy, however, direct analysis and utilization of them are impossible. Aiming at reducing the redundancy of manufacturing procedural data and excavating their local structure, a two-stage unsupervised feature selection method is proposed. In the first stage of the method, subset of the original feature set generated by genetic algorithm(GA) is utilized as the input features of radius basis function neural network(RBFNN), to reconstruct the unabridged original feature set. The ratio of dimensionality reduction and reconstructional accuracy are calculated jointly as the fitness function of GA, which is optimized by iteration to learn a low-dimensional representation of high-dimensional features, removing redundant and noisy features of the origin feature set. In the second stage, Laplacian score(LS) is employed to evaluate the locality preserving power of the remaining features, unearthing features which are prone to improving the performance of classification. By comparing with other unsupervised feature selection method, the method proposed here is proven more effective in reducing the redundancy of manufacturing data and simultaneously enhancing the performance of classification.

Key words: unsupervised feature selection, genetic algorithm, radius basis function neural network, Laplacian score, manufacturing procedural data

中图分类号:

TG156

张洁, 盛夏, 张朋, 秦威, 赵新明. 面向制造过程数据的两阶段无监督特征选择方法[J]. 机械工程学报, 2019, 55(17): 133-144.

ZHANG Jie, SHENG Xia, ZHANG Peng, QIN Wei, ZHAO Xinming. Two-stage Unsupervised Feature Selection Method Oriented to Manufacturing Procedural Data[J]. Journal of Mechanical Engineering, 2019, 55(17): 133-144.

参考文献

[1] 张洁,秦威,鲍劲松. 制造业大数据[M]. 上海:上海科学技术出版社,2016. ZHANG Jie,QIN Wei,BAO Jingsong. Big data in manufacturing industry[M]. Shanghai:Shanghai Scientific and Technical Publishers,2016.
[2] 杨文献,姜节胜. 基于数据挖掘的柴油机气门故障诊断技术研究[J]. 机械工程学报,2004,40(10):25-29. YANG Wenxian,JIANG Jiesheng. Data mining based technique for diagnosing the faults presenting on engine valves[J]. Chinese Journal of Mechanical Engineering,2004,40(10):25-29.
[3] 闫伟. 基于数据挖掘的复杂产品关键质量特性识别的方法研究[D]. 天津:天津大学,2012. YAN Wei. On Critical-to-quality characteristics identification for complex products using data mining[D]. Tianjin:Tianjin University,2012.
[4] KEOGH E,MUEEN A. Curse of dimensionality[J]. IND. ENG. Chem.,2017,29(1):48-53.
[5] 徐峻岭,周毓明,陈林,等. 基于互信息的无监督特征选择[J]. 计算机研究与发展,2012,49(2):372-382. XU Junling,ZHOU Yuming,CHEN Lin,et al. An unsupervised feature selection approach based on mutual information[J]. Journal of Computer Research and Development,2012,49(2):372-382.
[6] WANG Junliang,ZHANG Jie. Big data analytics for forecasting cycle time in semiconductor wafer fabrication system[J]. International Journal of Production Research,2016,54(23):7231-7244.
[7] WANG Junliang,ZHANG Jie,WANG Xiaoxi. A data driven cycle time prediction with feature selection in a semiconductor wafer fabrication system[J]. IEEE Transactions on Semiconductor Manufacturing,2018,31(1):173-182,5.
[8] ISLAM M R,UDDIN M S,KHAN S,et al. Multi-core accelerated discriminant feature selection for real-time bearing fault diagnosis[C]//International Conference on Industrial,Engineering and Other Applications of Applied Intelligent Systems. Springer International Publishing,2016:645-656.
[9] ZHU P,ZUO W,ZHANG L,et al. Unsupervised feature selection by regularized self-representation[J]. Pattern Recognition,2015,48(2):438-446.
[10] ZHAO Z,WANG L,LIU H. Efficient spectral feature selection with minimum redundancy[C]//Twenty-Fourth AAAI Conference on Artificial Intelligence,AAAI 2010,Atlanta,Georgia,USA,July. DBLP,2010.
[11] JOHN G H,KOHAVI R,PFLEGER K. Irrelevant features and the subset selection problem[J]. Precision Engineering,2004,2(4):314-318.
[12] ZHAO Z,LIU H. Spectral feature selection for supervised and unsupervised learning[C]//Machine Learning,Proceedings of the Twenty-Fourth International Conference. DBLP,2007:1151-1157.
[13] ZHAO Z,WANG L,LIU H. Efficient spectral feature selection with minimum redundancy[C]//Twenty-Fourth AAAI Conference on Artificial Intelligence,AAAI 2010,Atlanta,Georgia,USA,July. DBLP,2010.
[14] CAI D,ZHANG C,HE X. Unsupervised feature selection for multi-cluster data[C]//ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM,2010:333-342.
[15] 姚旭,王晓丹,张玉玺,等. 特征选择方法综述[J]. 控制与决策,2012,27(2):161-166. YAO Xu,WANG Xiaodan,ZHANG Yuxi,et al. Summary of feature selection algorithms[J]. Control and Decision,2012,27(2):161-166.
[16] ZHANG X L,CHEN W,WANG B J,et al. Intelligent fault diagnosis of rotating machinery using support vector machine with ant colony algorithm for synchronous feature selection and parameter optimization[J]. Neurocomputing,2015,167:260-279.
[17] CERRADA M,SÁNCHEZ R V,CABRERA D,et al. Multi-stage feature selection by using genetic algorithms for fault diagnosis in gearboxes based on vibration signal[J]. Sensors,2015,15(9):23903-23926.
[18] TABAKHI S,MORADI P,AKHLAGHIAN F. An unsupervised feature selection algorithm based on ant colony optimization[J]. Engineering Applications of Artificial Intelligence,2014,32:112-123.
[19] GUYON I,ELISSEEFF A. An introduction to variable and feature selection[J]. Journal of Machine Learning Research,2003,3:1157-1182.
[20] HOQUE N,BHATTACHARYYA D K,KALITA J K. MIFS-ND:A mutual information-based feature selection method[J]. Expert Systems with Applications,2014,41(14):6371-6385.
[21] PARK J,SANDBERG I W. Universal approximation using radial-basis-function networks[J]. Neural Computation,1991,3(2):246-257.
[22] WHITLEY D. A genetic algorithm tutorial[J]. Statistics and Computing,1994,4(2):65-85.
[23] LEONTARITIS I J,BILLINGS S A. Model selection and validation methods for non-linear systems[J]. International Journal of Control,1987,45(1):311-341.
[24] LECUN Y,BENGIO Y,HINTON G. Deep learning[J]. Nature,2015,521(7553):436.
[25] BELKIN M,NIYOGI P. Laplacian eigenmaps and spectral techniques for embedding and clustering[C]//International Conference on Neural Information Processing Systems:Natural and Synthetic. MIT Press,2002:585-591.
[26] HE X,NIYOGI P. Locality preserving projections[J]. Advances in Neural Information Processing Systems,2003,16(1):186-197.
[27] HE Xiaofei,DENG Cai,PARTHA N. Laplacian score for feature selection[J]. Advances in neural Information Processing Systems,2006:507-514.

面向制造过程数据的两阶段无监督特征选择方法

Two-stage Unsupervised Feature Selection Method Oriented to Manufacturing Procedural Data

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	石小秋, 李炎炎, 邓丁山, 龙伟. 基于自适应变级遗传杂草算法的FJSP研究[J]. 机械工程学报, 2019, 55(6): 223-232.
[2]	刘秀全, 姬景奇, 郑健, 王向磊, 陈国明, 张浩. 基于涡激抑制的深水钻井隔水管系统浮力块配置智能优化[J]. 机械工程学报, 2019, 55(4): 164-171.
[3]	朱茂桃, 吴新佳, 郑国峰, 李利平, 韩鹏飞, 上官文斌. 基于短时傅里叶变换的汽车零部件耐久性载荷信号编辑方法[J]. 机械工程学报, 2019, 55(4): 126-134.
[4]	谢延敏, 张飞, 王子豪, 黄仁勇, 杨俊峰, 胡云川. 基于渐变凹模圆角半径的高强钢扭曲回弹补偿[J]. 机械工程学报, 2019, 55(2): 91-97.
[5]	杨志军, 陈超然, 黄观新. 面向机器人优化设计的GA-非均匀Kriging-梯度投影混合全局优化算法[J]. 机械工程学报, 2019, 55(11): 61-68.
[6]	李明龙, 杨文婧, 易晓东, 王彦臻, 王戟. 面向灾难搜索救援场景的空地协同无人群体任务规划研究[J]. 机械工程学报, 2019, 55(11): 1-9.
[7]	姜兴宇, 王蔚, 张皓垠, 张凯, 李丽. 考虑质量、成本与资源利用率的再制造机床优化选配方法[J]. 机械工程学报, 2019, 55(1): 180-188.
[8]	魏鑫, 张泽群, 唐敦兵, 杨长祺, 金永乔, 秦威. 面向节能的导弹结构件混线生产作业车间多目标调度研究[J]. 机械工程学报, 2018, 54(9): 45-54.
[9]	赵宏晨,王刚,刘晓明. 多种群遗传算法与截断奇异值分解法在开关电弧反演中的应用[J]. 电气工程学报, 2018, 13(6): 14-19.
[10]	周生海,尹航,顾颖,王翀,柏峰. 基于POP NSGA-Ⅱ的配电网故障恢复重构[J]. 电气工程学报, 2018, 13(6): 28-35.
[11]	崔健,王久和. 基于最优阻尼注入的Buck变换器无源控制研究[J]. 电气工程学报, 2018, 13(6): 7-13.
[12]	常予, 焦敬品, 李光海, 何存富, 吴斌. 漏磁传感器励磁结构影响因素分析及优化设计[J]. 机械工程学报, 2018, 54(24): 7-17.
[13]	夏毅敏, 林赉贶, 贺飞, 吴遁, 李正光, 暨智勇. 盘形滚刀刀圈结构地质适应性设计方法[J]. 机械工程学报, 2018, 54(1): 10-17.
[14]	陈波, 高殿荣, 杨超, 毋少峰, 张光通. 大功率远程射雾器结构参数多目标智能协同优化^*[J]. 机械工程学报, 2017, 53(6): 166-175.
[15]	胡瑞飞, 殷鸣, 刘雁, 苏真伟, 殷国富. 切削稳定性约束下的铣削参数优化技术研究^*[J]. 机械工程学报, 2017, 53(5): 190-198.