USE OF CUMULATIVE SUM (CUSUM) TEST FOR DETECTING ABRUPT CHANGES IN THE PROCESS DYNAMICS

In order to operate a successful plant or process, continuous improvement must be made in the areas of safety, quality and reliability. Central to this continuous improvement is the early or proactive detection of process faults. Even though fault detection algorithms are designed and implemented for quickly detecting incidents, most these algorithms do not have an optimal property in terms of detection delay with respect to false alarm rate. Based on this optimization property, a real-time system for detecting changes in dynamic systems is designed in this paper. This work is motivated by combining two fault detection (FD) methods; a simplified procedure of the incident detection problem is formulated by using the combination of the Artificial Neural Networks (ANN) and the cumulative sum (CUSUM) or Page-Hinkley test. It is intended to reveal any drift from the normal behavior of the process. The process behavior under its normal operating conditions is established by a reliable model. In order to obtain this reliable model for the process dynamics, the black-box identification by means of a NARX (Nonlinear Auto-Regressive with eXogenous input) model is chosen in this study. The purpose is to develop and test the fault detection method on a real incident data, to detect the change presence, and pinpoint the moment it occurred. The experimental results demonstrate the robustness of the FD method.

[1]  Mitra Fouladirad,et al.  Optimal statistical fault detection with nuisance parameters , 2005, Autom..

[2]  Thomas J. McAvoy,et al.  Nonlinear FIR Modeling via a Neural Net PLS Approach , 1996 .

[3]  Sheng Chen,et al.  Representations of non-linear systems: the NARMAX model , 1989 .

[4]  Mohamed Benbouzid,et al.  Induction motor stator faults diagnosis by a current Concordia pattern-based fuzzy decision system , 2003 .

[5]  William J. Garland,et al.  Fault detection and diagnosis using statistical control charts and artificial neural networks , 1998, Artif. Intell. Eng..

[6]  Y. Chetouani.  Fault detection by using the innovation signal: application to an exothermic reaction , 2004 .

[7]  Alain Poirel,et al.  pH modelling by neural networks. Application of control and validation data series in the Middle Loire river , 1999 .

[8]  Michel Kinnaert,et al.  Model-based fault detection and isolation for a gas–liquid separation unit , 2000 .

[9]  Konstantinos P. Ferentinos,et al.  Biological engineering applications of feedforward neural networks designed and parameterized by genetic algorithms , 2005, Neural Networks.

[10]  Paul M. Frank,et al.  Issues of Fault Diagnosis for Dynamic Systems , 2010, Springer London.

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

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

[13]  Yi Qi,et al.  Detection-delay-based freeway incident detection algorithms , 2003 .

[14]  Mark R. Warnes,et al.  On data-based modelling techniques for fermentation processes , 1996 .

[15]  G. Lorden.  PROCEDURES FOR REACTING TO A CHANGE IN DISTRIBUTION , 1971 .

[16]  Roman W. Swiniarski,et al.  Rough sets as a front end of neural-networks texture classifiers , 2001, Neurocomputing.

[17]  D. Hinkley.  Inference about the change-point from cumulative sum tests , 1971 .

[18]  D. Siegmund,et al.  A diffusion process and its applications to detecting a change in the drift of Brownian motion , 1984 .

[19]  Paul M. Frank,et al.  Fault diagnosis in dynamic systems using analytical and knowledge-based redundancy: A survey and some new results , 1990, Autom..

[20]  Fu Xiao,et al.  AHU sensor fault diagnosis using principal component analysis method , 2004 .

[21]  Yahya Chetouani,et al.  Fault Detection in a Chemical Reactor by Using the Standardized Innovation , 2006 .

[22]  Y. Chetouani.  Application of the Generalized Likelihood Ratio Test for Detecting Changes in a Chemical Reactor , 2006 .

[23]  Ranjan Ganguli,et al.  Filter design using radial basis function neural network and genetic algorithm for improved operational health monitoring , 2006, Appl. Soft Comput..

[24]  Joseph McGhee,et al.  Neural networks applied for the identification and fault diagnosis of process valves and actuators , 1997 .

[25]  Chuen-Sheng Cheng,et al.  A neural network-based procedure for the monitoring of exponential mean , 2001 .

[26]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[27]  J. Savković-Stevanović.  Neural net controller by inverse modeling for a distillation plant , 1996 .

[28]  Malcolm K. Horne,et al.  A comparison of methods used to detect changes in neuronal discharge patterns , 1997, Journal of Neuroscience Methods.

[29]  Gilles Mourot,et al.  Rainfall–runoff multi-modelling for sensor fault diagnosis , 2001 .

[30]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part III: Process history based methods , 2003, Comput. Chem. Eng..

[31]  Bin Jiang,et al.  Parameter fault detection and estimation of a class of nonlinear systems using observers , 2005, J. Frankl. Inst..

[32]  S. Joe Qin,et al.  Joint diagnosis of process and sensor faults using principal component analysis , 1998 .

[33]  Kumpati S. Narendra,et al.  Identification and control of dynamical systems using neural networks , 1990, IEEE Trans. Neural Networks.

[34]  Fabio Previdi,et al.  Identification of black-box nonlinear models for lower limb movement control using functional electrical stimulation , 2002 .

[35]  N. Aretakis,et al.  Identification of sensor faults on turbofan engines using pattern recognition techniques , 2004 .

[36]  P H Ellaway,et al.  Cumulative sum technique and its application to the analysis of peristimulus time histograms. , 1978, Electroencephalography and clinical neurophysiology.