(375b) Designing Data-Based and Model-Based Methods for Process Monitoring and Equipment Degradation Tracking

Authors: 
Sheriff, M. Z., Texas A&M University
Karim, M. N., Texas A&M University
Kravaris, C., Texas A&M University
Nounou, H., Texas A&M University at Qatar
Nounou, M., Texas A&M University at Qatar
A wide variety of industrial processes utilize monitoring methods in order to ensure that safety and quality is maintained. Conventional model-based methods include dimensionality reduction techniques such as Principal Component Analysis (PCA) are often employed as they are computationally simple, and easy to implement in practice [1]. Unfortunately, these techniques assume that the process data is Gaussian, uncorrelated, and only contain a moderate level of noise. This work will demonstrate how wavelet-based representation of data can be utilized in order to efficiently handle data that violate these assumptions. Statistical hypothesis testing methods such as the generalized likelihood ratio chart have shown promise with respect to fault detection, and thus will be used to enhance monitoring capabilities [2], [3].

Additionally, this work will demonstrate the necessity to track deviations in the process models themselves in order to identify process drifts or equipment degradation. When a process operates under control, a controller continuously adjusts the level of manipulated variables in order to ensure that the controlled variables are being maintained within specific limits. Unfortunately, this may be at the expense of increased operating and additional costs. Therefore, this work will utilize a dynamic contour-based algorithm in order to illustrate how equipment degradation can be efficiently tracked in multiple operating regimes [4].

For both algorithms, illustrative examples using simulated synthetic data, and real data from different applications will be utilized in order to highlight the effectiveness of the developed algorithms.

References

[1] I. T. Joliffe, Principal Component Analysis, 2nd ed. New York, NY: Springer-Verlag, 2002.

[2] M. R. Reynolds and J. Y. Lou, “An Evaluation of a GLR Control Chart for Monitoring the Process Mean,” J. Qual. Technol., vol. 42, no. 3, pp. 287–310, 2010.

[3] M. Z. Sheriff, M. Mansouri, M. N. Karim, H. Nounou, and M. Nounou, “Fault detection using multiscale PCA-based moving window GLRT,” J. Process Control, vol. 54, 2017, doi: 10.1016/j.jprocont.2017.03.004.

[4] M. Z. Sheriff, H. Nounou, M. Nounou, and M. N. Karim, “Monitoring process degradation through operating regime based process monitoring,” in AIChE Spring Meeting and Global Congress on Process Safety: Process Control Monitoring and Analytics, 2019.