Nonlinear Process Monitoring Using Supervised Locally Linear Embedding Projection

The chemical and mineral processing industries need a nonlinear process monitoring method to improve the stability and economy of their processes. Techniques that are currently available to these industries are often too computationally intensive for an industrial control system, or they are too complex to commission. In this paper, we propose using supervised locally linear embedding for projection (SLLEP) as a new nonlinear process monitoring technique to solve these issues. In addition, we suggest using a commonly available tool in modern industrial control systems, a model predictive control, to solve the quadratic program of SLLEP in real-time and with minimal effort to commission. As a case study, we demonstrate that process monitoring with SLLEP can detect and diagnose the early onset of a semiautogenous grinding (SAG) mill overload. A SAG mill overload is a highly nonlinear operating situation, and we show that principal component analysis, the best-in-class technique currently used by the industr...

[1]  L. C. Coetzee,et al.  Robust nonlinear model predictive control of a closed run-of-mine ore milling circuit , 2009 .

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

[3]  Ian K. Craig,et al.  Grinding mill circuits — A survey of control and economic concerns , 2009 .

[4]  Qi Li,et al.  Constrained model predictive control in ball mill grinding process , 2008 .

[5]  Kenneth Scott McClure.  Algorithm for nonlinear process monitoring and controller performance recovery with an application to semi-autogenous grinding , 2013 .

[6]  Lawrence K. Saul,et al.  Think Globally, Fit Locally: Unsupervised Learning of Low Dimensional Manifold , 2003, J. Mach. Learn. Res..

[7]  Young-Don Ko,et al.  SAG mill system diagnosis using multivariate process variable analysis , 2011 .

[8]  Sirkka-Liisa Jämsä-Jounela,et al.  Evaluation of PCA methods with improved fault isolation capabilities on a paper machine simulator , 2008 .

[9]  T. McAvoy,et al.  Nonlinear principal component analysis—Based on principal curves and neural networks , 1996 .

[10]  Ian K. Craig,et al.  Specification framework for robust control of a run-of-mine ore milling circuit , 1995 .

[11]  Jin Hyun Park,et al.  Fault detection and identification of nonlinear processes based on kernel PCA , 2005 .

[12]  Paul W. Cleary,et al.  Prediction of slurry transport in SAG mills using SPH fluid flow in a dynamic DEM based porous media , 2006 .

[13]  Stephen Morrell,et al.  Slurry flow in mills: grate-only discharge mechanism (Part-1) , 2003 .

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

[15]  Benwei Li,et al.  Supervised locally linear embedding projection (SLLEP) for machinery fault diagnosis , 2011 .

[16]  Bernhard Schölkopf,et al.  Nonlinear Component Analysis as a Kernel Eigenvalue Problem , 1998, Neural Computation.

[17]  Stephen Morrell,et al.  The prediction of power draw in wet tumbling mills , 1993 .

[18]  Thomas E. Marlin,et al.  Multivariate statistical monitoring of process operating performance , 1991 .

[19]  S T Roweis,et al.  Nonlinear dimensionality reduction by locally linear embedding. , 2000, Science.

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

[21]  Panagiotis D. Christofides,et al.  Data-based fault detection and isolation using feedback control: Output feedback and optimality , 2009 .

[22]  Jun Yang,et al.  Disturbance rejection of ball mill grinding circuits using DOB and MPC , 2010 .