Isolating the root cause of propagated oscillations in process plants

Oscillations are a common type of propagated disturbance, whose sources might be attributable to a number of different phenomena such as poor controller tuning or actuator nonlinearity. A number of data-driven methods have already been proposed to isolate the source loop of nonlinearity induced plant-wide oscillations. Amongst these the bi-amplitude ratio index, correlation dimension, maximal Lyapunov exponent, nonlinearity index and spectral ICA show promise. The propagation of oscillations is first examined in order to gain an understanding of how this might affect the performances of the various techniques. The various methods are then described and their performance on a set of simulation generated data and two industrial case studies are compared. Copyright © 2004 John Wiley & Sons, Ltd.

[1]  John Howell,et al.  Correlation Dimension and Lyapunov Exponent Based Isolation of Plant-Wide Oscillations , 2004 .

[2]  A. Wallen.  Valve diagnostics and automatic tuning , 1997, Proceedings of the 1997 American Control Conference (Cat. No.97CH36041).

[3]  Chunming Xia,et al.  Control loop measurement based isolation of faults and disturbances in process plants , 2003 .

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

[5]  Alf Isaksson,et al.  A Method for Detection of Stiction in Control Valves , 1998 .

[6]  P. Grassberger,et al.  Characterization of Strange Attractors , 1983 .

[7]  Nina F. Thornhill,et al.  DETECTION OF DISTRIBUTED OSCILLATIONS AND ROOT-CAUSE DIAGNOSIS , 2001 .

[8]  Peter A. Cook,et al.  Nonlinear dynamical systems , 1986 .

[9]  John W. Cox,et al.  A practical approach for large-scale controller performance assessment, diagnosis, and improvement , 2003 .

[10]  Karl Johan Åström,et al.  Dominant pole placement for multi-loop control systems , 2002, Autom..

[11]  Nina F. Thornhill,et al.  Diagnosis of poor control-loop performance using higher-order statistics , 2004, Autom..

[12]  P. Grassberger,et al.  Measuring the Strangeness of Strange Attractors , 1983 .

[13]  S. Joe Qin,et al.  Control performance monitoring — a review and assessment , 1998 .

[14]  Holger Kantz,et al.  Practical implementation of nonlinear time series methods: The TISEAN package. , 1998, Chaos.

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

[16]  H. Kantz,et al.  Nonlinear time series analysis , 1997 .

[17]  Alexander Horch.  A simple method for detection of stiction in control valves , 1999 .

[18]  Nina F. Thornhill,et al.  Detection of multiple oscillations in control loops , 2003 .

[19]  Xiaoyun Zang,et al.  Discrimination between bad tuning and non-linearity induced oscillations through bispectral analysis , 2003, SICE 2003 Annual Conference (IEEE Cat. No.03TH8734).

[20]  Nina F. Thornhill,et al.  ANALYSIS OF PLANT-WIDE DISTURBANCES THROUGH DATA-DRIVEN TECHNIQUES AND PROCESS UNDERSTANDING , 2002 .

[21]  Karl Johan Åström,et al.  Assessment of Achievable Performance of Simple Feedback Loops , 1988 .

[22]  A qualitative shape analysis formalism for monitoring control loop performance , 2001 .

[23]  Chunming Xia,et al.  Isolating multiple sources of plant-wide oscillations via independent component analysis , 2003 .

[24]  Tore Hägglund,et al.  A Control-Loop Performance Monitor , 1995 .