MultiBoost with ENN-based ensemble fault diagnosis method and its application in complicated chemical process

Abstract

Fault diagnosis plays an important role in complicated industrial process. It is a challenging task to detect, identify and locate faults quickly and accurately for large-scale process system. To solve the problem, a novel MultiBoost-based integrated ENN (extension neural network) fault diagnosis method is proposed. Fault data of complicated chemical process have some difficult-to-handle characteristics, such as high-dimension, non-linear and non-Gaussian distribution, so we use margin discriminant projection(MDP) algorithm to reduce dimensions and extract main features. Then, the affinity propagation (AP) clustering method is used to select core data and boundary data as training samples to reduce memory consumption and shorten learning time. Afterwards, an integrated ENN classifier based on MultiBoost strategy is constructed to identify fault types. The artificial data sets are tested to verify the effectiveness of the proposed method and make a detailed sensitivity analysis for the key parameters. Finally, a real industrial system—Tennessee Eastman (TE) process is employed to evaluate the performance of the proposed method. And the results show that the proposed method is efficient and capable to diagnose various types of faults in complicated chemical process.

Keywords

extension neural network / multi-classifier ensembles / margin discriminant projection / affinity propagation / fault diagnosis / TE process

References

[1] GaoD, WuC-g, ZhangB-k, MaXin. Signed directed graph and qualitative trend analysis based fault diagnosis in chemical industry [J]. Chinese Journal of Chemical Engineering, 2010, 18(2): 265-276

[2] MauryaM R, RengaswamyR. Application of signed digraphs-based analysis for fault diagnosis of chemical process flowsheets [J]. Engineering Application of Artificial Intelligence, 2004, 17(5): 501-518

[3] LauC K, GhoshK, HussainM A, Che HassanC R. Fault diagnosis of Tennessee Eastman process with multi-scale PCA and ANFIS [J]. Chemometrics and Intelligent Laboratory, 2013, 120: 1-14

[4] YuJie. A support vector clustering-based probabilistic method for unsupervised fault detection and classification of complex chemical processes using unlabeled data [J]. AIChE Journal, 2013, 59(2): 407-419

[5] JiangQ-c, YanX-feng. Just-in-time reorganized PCA integrated with SVDD for chemical process monitoring [J]. AIChE Journal, 2014, 60(3): 949-965

[6] WangX-f, WangS-h, DuH-z, WangPing. Fault diagnosis of chemical industry process based on FRS and SVM [J]. Control and Decision, 2015, 30(2): 353-356

[7] LuF, LiX, DuW-xia. MultiBoost with SVM-based ensemble classification method and application [J]. Control and Decision, 2015, 30(1): 81-85

[8] ChenX-y, YanX-feng. Fault diagnosis in chemical process based on self-organizing map integrated with fisher discriminant analysis [J]. Chemical Journal of Chemical Engineering, 2013, 21(4): 382-387

[9] SongB, MaY-x, FangY-f, ShiH-bo. Fault detection for chemical process based on LSNPE method [J]. Chinese Journal of Chemical Engineering Journal, 2014, 65(2): 620-627

[10] ZhouY, PedryczW, QianXu. Application of extension neural network to safety status pattern recognition of coal mines [J]. Joumal of Central South University of Technology, 2011, 18: 633-641

[11] BarakatM, DruauxF, LefebvreD, KhalilM, MustaphaO. Self adaptive growing neural network classifier for faults detection and diagnosis [J]. Neurocomputing, 2011, 74: 3865-3876

[12] ZhuZ-b, SongZ-huan. A novel fault diagnosis system using pattern classification on kernel FDA subspace [J]. Expert System with Application, 2011, 38(6): 6895-6905

[13] JiangQ-c, YanX-feng. Probabilistic weighted NPE-SVDD for chemical process monitoring [J]. Control Engineering Practice, 2014, 28: 74-89

[14] LiX-q, SongQ, LiuW-h, RaoH, XuS-k, LiL-cheng. Protection of nonpermanent faults on DC overhead lines in MMC-based HVDC systems [J]. IEEE Trans on Power Delivery, 2013, 28(1): 483-490

[15] GaoH-h, HeY-l, ZhuQ-xiong. Research and chemical application of data attributes decomposition based hierarchical ELM neural network [J]. Chinese Journal of Chemical Engineering, 2013, 64(12): 4348-4353

[16] HeD, ZhangJ-song. Fault strategy of time delay analysis based on Hopfield network [J]. Chinese Journal of Chemical Engineering, 2013, 64(2): 633-640

[17] PengD, HeY-l, XuY, ZhuQ-xiong. Research and chemical application of data feature extraction based AANN-ELM neural network [J]. CIESC Journal, 2012, 63(9): 2920-2925

[18] WangM-hui. Partial discharge pattern recognition of current transformers using an ENN [J]. IEEE Trans on Power Delivery, 2005, 20(3): 1984-1990

[19] WangM-h, HungC P. Extension neural network and its applications [J]. Neural Networks, 2003, 16(5): 779-784

[20] ChaoK H, WangM H, HshC C. A novel residual capacity estimation method based on extension neural network for lead-acid batteries [J]. Lecture Notes in Computer Science, 2007, 44931145-1154

[21] ChaoK H, LeeR H, WangM H. An intelligent traffic light control based on extension neural network [J]. Lecture Notes in Computer Science, 2008, 5177: 17-24

[22] TangS, ZhengY-t, WangY, ChuaT-seng. Sparse ensemble learning for concept detection [J]. IEEE Trans on Multimedia, 2012, 14(1): 43-54

[23] ShenC-h, HaoZ-huiA direct formulation for totallycorrective multi-class boosting [C]// Computer Vision and Pattern Recognition (CVPR) of 2011 IEEE Conference, 2011Washington D.C.IEEE Computer Society2585-2592

[24] DjalelB, RobertB F, NormanC, ColinF D, BalazsK. MultiBoost: A multi-purpose boosting package [J]. The Journal of Machine Learning Research, 2012, 13(1): 549-553

[25] HeJ-r, DingL-x, LiZ-k, HuQ-hui. Margin discriminant projection for dimensionality reduction [J]. Journal of Software, 2014, 25(4): 826-838

[26] ShangF H, JiaoL C, ShiJ R, WangF, GongM G. Fast affinity propagation clustering: A multilevel approach [J]. Pattern Recognition, 2012, 45(1): 474-486

[27] WangC-d, LaiJ-h, SuenC Y, ZhuJ-yong. Multi-exemplar affinity propagation [J]. Pattern Analysis and Machine, 2013, 35(9): 2223-2237

[28] LinY-q, LvF-j, ZhuS-h, YangM, CourT, YuK, CaoL-l, HuangT. Large-scale image classification: Fast feature extraction and SVM training [C]// Computer Vision and Pattern Recognition (CVPR). 2011 IEEE Conference, 20111689-1696

[29] YanS-c, XuD, ZhangB-y, ZhangH-j, ZhangQ, LinS. Graph embedding and extensions: A general framework for dimensionality reduction [J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 2007, 29(1): 40-51

[30] Shawe-TaylorJ, CristianiniNKernel methods for pattern analysis [M], 2004CambriddgeshireCambridge University Press

[31] GalarM, FernandezA, BarrenecheaE, BustinceH, HerrerF. A review on ensembles for the class inbalance problem: Bagging, Boosting, and hybrid-based approach [J]. Systems, Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on, 2012, 42(4): 463-484

[32] KrokroghA, VedelsbyJNeural network ensembles,cross validation, and active learning [M], 1995Cambridge, MAMIT Press231-238

[33] GuanR C, ShiX H, MarcheseM, YangC, LiangY. Text clustering with seeds affinity propagation [J]. IEEE Transaction on Knowledge and Data Engineering, 2011, 23(4): 627-637