Data-driven root-cause fault diagnosis for multivariate non-linear processes

Abstract In a majority of multivariate processes, propagating nature of malfunctions makes the fault diagnosis a challenging task. This paper presents a novel data-driven strategy for real-time root-cause fault diagnosis in multivariate (non-)linear processes by estimating the strength of causality using normalized transfer entropy (NTE) between measured process variables and variations of a residual signal. In this paper, a new framework for root-cause fault diagnosis applicable for multivariate nonlinear processes is proposed, which can reduce the necessary number of calculation for causality analysis among time-series. More specially, a new and fast symbolic dynamic-based normalized transfer entropy (SDNTE) technique is proposed to enable real-time application of transfer entropy, which has been considered as a burdensome approach for causality analysis. The concept of SDNTE is built upon principles of time-series symbolization, xD-Markov machine and Shannon entropy. This paper also introduces a new concept of joint xD-Markov machine to capture dynamic interactions between two time-series. The proposed root-cause fault diagnosis framework is applied on Tennessee Eastman process benchmark and its computational advantages are shown by comparing with conventional kernel PDF-based method. Moreover, the proposed strategy is applied to health monitoring of a big scale industry centrifuge to corroborates its effectiveness and feasibility in industrial applications.

[1]  Xin Jin,et al.  Optimization of symbolic feature extraction for pattern classification , 2012, Signal Process..

[2]  Sirish L. Shah,et al.  Detection and Diagnosis of Plant-wide Oscillations From Industrial Data using the Spectral Envelope Method ? , 2007 .

[3]  Jef Vanlaer,et al.  Contribution plots for Statistical Process Control: Analysis of the smearing-out effect , 2013, 2013 European Control Conference (ECC).

[4]  Tongwen Chen,et al.  Methods for root cause diagnosis of plant‐wide oscillations , 2014 .

[5]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part I: Quantitative model-based methods , 2003, Comput. Chem. Eng..

[6]  Si-Zhao Joe Qin,et al.  Reconstruction-based contribution for process monitoring , 2009, Autom..

[7]  Jianbin Qiu,et al.  Diagnostic Observer Design for T–S Fuzzy Systems: Application to Real-Time-Weighted Fault-Detection Approach , 2018, IEEE Transactions on Fuzzy Systems.

[8]  Tao Yuan,et al.  Nonstationarity and cointegration tests for fault detection of dynamic processes , 2014 .

[9]  Karim Salahshoor,et al.  Process fault isolation based on transfer entropy algorithm. , 2014, ISA transactions.

[10]  John F. MacGregor,et al.  Multivariate SPC charts for monitoring batch processes , 1995 .

[11]  Asok Ray,et al.  Symbolic dynamic analysis of complex systems for anomaly detection , 2004, Signal Process..

[12]  Heiko Hoffmann,et al.  Kernel PCA for novelty detection , 2007, Pattern Recognit..

[13]  Asok Ray,et al.  Sensor Fusion for Fault Detection and Classification in Distributed Physical Processes , 2014, Front. Robot. AI.

[14]  Asok Ray,et al.  Estimation of slowly varying parameters in nonlinear systems via symbolic dynamic filtering , 2008, Signal Process..

[15]  S. Joe Qin,et al.  Data-driven root cause diagnosis of faults in process industries , 2016, Chemometrics and Intelligent Laboratory Systems.

[16]  Sirish L. Shah,et al.  Root cause diagnosis of plant-wide oscillations using the concept of adjacency matrix , 2009 .

[17]  Sirish L. Shah,et al.  Direct Causality Detection via the Transfer Entropy Approach , 2013, IEEE Transactions on Control Systems Technology.

[18]  B. Silverman.  Density estimation for statistics and data analysis , 1986 .

[19]  Christos Georgakis,et al.  Disturbance detection and isolation by dynamic principal component analysis , 1995 .

[20]  Ping Zhang,et al.  A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process , 2012 .

[21]  Mohammed Chadli,et al.  Fuzzy Fault Detection Filter Design for T–S Fuzzy Systems in the Finite-Frequency Domain , 2017, IEEE Transactions on Fuzzy Systems.

[22]  Sirkka-Liisa Jämsä-Jounela,et al.  Hybrid approach to casual analysis on a complex industrial system based on transfer entropy in conjunction with process connectivity information , 2016 .

[23]  Asok Ray,et al.  Analytic signal space partitioning and symbolic dynamic filtering for degradation monitoring of electric motors , 2010, Signal Image Video Process..

[24]  Asok Ray,et al.  Symbolic time series analysis via wavelet-based partitioning , 2006 .

[25]  Gunnar Rätsch,et al.  Input space versus feature space in kernel-based methods , 1999, IEEE Trans. Neural Networks.

[26]  Nojun Kwak,et al.  Nonlinear Projection Trick in Kernel Methods: An Alternative to the Kernel Trick , 2013, IEEE Transactions on Neural Networks and Learning Systems.

[27]  Sirish L. Shah,et al.  Transfer Zero-Entropy and Its Application for Capturing Cause and Effect Relationship Between Variables , 2015, IEEE Transactions on Control Systems Technology.

[28]  Schreiber,et al.  Measuring information transfer , 2000, Physical review letters.

[29]  Asok Ray,et al.  Symbolic dynamic analysis of surface deformation during fatigue crack initiation , 2010 .

[30]  R. Steuer,et al.  Entropy and optimal partition for data analysis , 2001 .

[31]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part II: Qualitative models and search strategies , 2003, Comput. Chem. Eng..

[32]  T. Schreiber,et al.  Information transfer in continuous processes , 2002 .

[33]  Nina F. Thornhill,et al.  Finding the Direction of Disturbance Propagation in a Chemical Process Using Transfer Entropy , 2007, IEEE Transactions on Control Systems Technology.

[34]  Inseok Hwang,et al.  A Survey of Fault Detection, Isolation, and Reconfiguration Methods , 2010, IEEE Transactions on Control Systems Technology.

[35]  Nina F. Thornhill,et al.  Diagnosis of plant-wide oscillation through data-driven analysis and process understanding , 2003 .

[36]  Richard D. Braatz,et al.  Fault Detection and Diagnosis in Industrial Systems , 2001 .

[37]  In-Beum Lee,et al.  Fault identification for process monitoring using kernel principal component analysis , 2005 .

[38]  Rongrong Wang,et al.  Actuator and sensor faults estimation based on proportional integral observer for TS fuzzy model , 2017, J. Frankl. Inst..

[39]  Anders L. Madsen,et al.  Applications of object-oriented Bayesian networks for condition monitoring, root cause analysis and decision support on operation of complex continuous processes , 2005, Comput. Chem. Eng..

[40]  Kushal Mukherjee,et al.  Multi-sensor information fusion for fault detection in aircraft gas turbine engines , 2013 .

[41]  Kushal Mukherjee,et al.  State splitting and merging in probabilistic finite state automata for signal representation and analysis , 2014, Signal Process..

[42]  C. Yoo,et al.  Nonlinear process monitoring using kernel principal component analysis , 2004 .

[43]  A. Ray,et al.  On-line identification of language measure parameters for discrete-event supervisory control , 2005 .

[44]  S. Joe Qin,et al.  Root cause diagnosis of plant-wide oscillations using Granger causality , 2014 .

[45]  Tore Hägglund,et al.  Detection and Diagnosis of Oscillation in Control Loops , 1997 .