Fault Detection Based on GPNMF for Industrial Process

doi:10.16182/j.issn1004731x.joss.201802020

Abstract

Abstract: As a newly dimension reduction technique, non-negative matrix factorization (NMF) has been applied in varying research areas. NMF methods require the original data non-negative. However, the operating data of industrial process maybe not satisfy this restriction. To resolve the problem, a new method is presented, which can be called as generalized projection non-negative matrix factorization (GPNMF). We use GPNMF to extract the latent variables that drive a process and to combine them with process monitoring techniques for fault detection. The corresponding contribution plots are defined for fault isolation. The proposed method is applied to a 1 000 MW unit boiler process. The simulation results clearly illustrate the feasibility of the proposed method.

Key words: fault detection, fault isolation, generalized projection non-negative matrix factorization, boiler process

CLC Number:

TP277

Niu Yuguang, Wang Shilin, Lin Zhongwei, Li Xiaoming. Fault Detection Based on GPNMF for Industrial Process[J]. Journal of System Simulation, 2018, 30(2): 521-532.

References 24

[1]	Ge Z, Song Z, Gao F.Review of recent research on data-based process monitoring[J]. Industrial & Engineering Chemistry Research (S0888-5885), 2013, 52(10): 3543-3562.
[2]	Jiang Q, Yan X, Zhao W.Fault detection and diagnosis in chemical processes using sensitive principal component analysis[J]. Industrial & Engineering Chemistry Research (S0888-5885), 2013, 52(4): 1635-1644.
[3]	Mahvash A, Lakis A A.Independent component analysis as applied to vibration source separation and fault diagnosis[J]. Journal of Vibration and Control (S1077-5463), 2016, 22(6): 1682-1692.
[4]	Lee D D, Seung H S.Learning the parts of objects by non-negative matrix factorization[J]. Nature (S1476-4687), 1999, 401(6755): 788-791.
[5]	周东华, 胡艳艳. 动态系统的故障诊断技术[J]. 自动化学报, 2009, 35(6): 748-758.Zhou Donghua, Hu Yanyan. Fault diagnosis techniques for dynamic systems[J]. Acta Automatica Sinica, 2009, 35(6): 748-758.
[6]	Deng X, Tian X.Entropy principal component analysis and its application to nonlinear chemical process fault diagnosis[J]. Asia-Pacific Journal of Chemical Engineering (S1932-2143), 2014, 9(5): 696-706.
[7]	Shao R, Hu W, Wang Y, et al.The fault feature extraction and classification of gear using principal component analysis and kernel principal component analysis based on the wavelet packet transform[J]. Measurement (S0263-2241), 2014, 54(4): 118-132.
[8]	Zhao Q.The Study on Rotating Machinery Early Fault Diagnosis based on Principal Component Analysis and Fuzzy C-means Algorithm[J]. Journal of Software (S1000-9825), 2013, 8(3): 709-715.
[9]	高东, 吴重光, 张贝克, 等. 基于PCA 和 SDG 的传感器故障诊断方法研究及应用[J]. 系统仿真学报, 2011, 23(3): 567-573.Gao Dong, Wu Chongguang, Zhang Beike, et al. Method combining PCA and SDG for fault diagnosis of sensors and its application[J]. Journal of System Simulation, 2011, 23(3): 567-573.
[10]	Kano M, Tanaka S, Hasebe S, et al.Monitoring independent components for fault detection[J]. AIChE Journal (S1547-5905), 2003, 49(4): 969-976.
[11]	Lee D D, Seung H S.Algorithms for non-negative matrix factorization[C]// Advances in neural information processing systems, 2001: 556-562.
[12]	Saito N.Modern signal processing[M].California: Mathematical Sciences Research Institute, 2004: 317-340.
[13]	Ding C, Li T, Peng W, et al.Orthogonal nonnegative matrix t-factorizations for clustering[C]// Proceedings of the 12th ACM SIGKDD international conference on Knowledge discovery and data mining. ACM, 2006: 126-135.
[14]	Yang Z, Laaksonen J.Multiplicative updates for non- negative projections[J]. Neurocomputing (S0925-2312), 2007, 71(1): 363-373.
[15]	Yuan Z, Oja E.Projective nonnegative matrix factorization for image compression and feature extraction[C]// Scandinavian Conference on Image Analysis. Springer Berlin Heidelberg, 2005: 333-342.
[16]	Li Z, Wu X, Peng H.Nonnegative matrix factorization on orthogonal subspace[J]. Pattern Recognition Letters (S0167-8655), 2010, 31(9): 905-911.
[17]	Chikhi N F, Rothenburger B, Aussenac-Gilles N.A comparison of dimensionality reduction techniques for web structure mining[C]// Proceedings of the IEEE/WIC/ACM International Conference on Web Intelligence. IEEE Computer Society, 2007: 116-119.
[18]	Li X, Yang Y, Zhang W.Fault detection method for non-Gaussian processes based on non-negative matrix factorization[J]. Asia-Pacific Journal of Chemical Engineering (S1932-2143), 2013, 8(3): 362-370.
[19]	王帆, 杨雅伟, 谭帅, 等. 基于稀疏性非负矩阵分解的故障监测方法[J]. 化工学报, 2015, 66(5): 1798-1805.Wang Fan, Yang Yawei, Tan Shuai, et al. Fault detection method based on sparse non-negative matrix factorization[J]. CIESC J, 2015, 66(5): 1798-1805.
[20]	Jackson J E, Mudholkar G S.Control procedures for residuals associated with principal component analysis[J]. Technometrics (S0040-1706), 1979, 21(3): 341-349.
[21]	Tikhonov A.Solution of incorrectly formulated problems and the regularization method[C]// Soviet Math. Dokl. 1963, 5: 1035.
[22]	Albright R, Cox J, Duling D, et al.Algorithms, initializations, and convergence for the nonnegative matrix factorization[R]. Tech. rep. 919. NCSU Technical Report Math 81706. North Caroline State University. 2006.
[23]	周东华, 李钢, 李元. 数据驱动的工业过程故障诊断技术: 基于主元分析与偏最小二乘的方法[M].北京: 科学出版社, 2010: 28-30.Zhou Donghua, Ligang, Li Yuan. Data driven fault diagnosis for industrial processes: Based on principal component analysis and partial least squares[M]. Beijing: Science Press, 2010: 28-30.
[24]	Alcala C F, Qin S J.Reconstruction-based contribution for process monitoring[J]. Automatica (S0005-1098), 2009, 45(7): 1593-1600.