Recursive Identification for Parameter Change and Fault Detection

Recursive Identification for Parameter Change and Fault Detection. Seminar of two hours for the Nanjing University of Aeronautics and Astronautics - NUAA
 

Abstract

• This talk addresses the issue of recursive or online identification based on the Recursive Least Squares (RLS) approach. It utilises an algorithm that iteratively estimates model parameters to minimise the weighted least squares cost function using new data. It is also exploited for parameter estimation, particularly in situations where parameters might change over time, making it useful for detecting and tracking faults in various systems. By continually updating the parameter estimates with new data, RLS can adapt to changes, making it suitable for those scenarios where faults can be detected by monitoring changes in estimated parameters.

Keywords

• Recursive system identification, process monitoring, fault diagnosis, online identification, parameter estimation, change detection.

Talk's Topics

• General Readings. Introduction to Recursive Identification. Report: L. Ljung and T. Soderstrom, Theory and Practice of Recursive Identification. The MIT Press. February 6, 1987.
• Lecture's Slides. Introduction to Recursive Identification for Fault Detection and Parameter Estimation. Lecture slides in PDF Format.
• Lecture's Slides. Recursive (online) methods. Lecture Slides in PDF Format (1.7 MB). Lecture recorded movie (YouTube Channel, 55 mins.).
• Simulation Example 1. Recursive (online) Methods: Two Application Examples in Matlab. Matlab script file 1 for lecture's simulation; Matlab script file 2 for lecture's simulation. Lecture recorded movie (YouTube Channel, 24 mins).
• Simulation Example 2. Recursive Estimation for Linear Varying Parameters of a Dynamic Process: Matlab and Simulink Simulations. Implementation of the RLS algorithm with forgetting factor (example); Matlab script for result display; Simulink file for data generation with linear varying parameters; RLS Simulink block and data generation. Offline movie (YouTube Channel, 21 mins).
• Open Problems and Future Works. Some System Identification Challenges and Approaches. Slides in PDF format.

Downloads

• Identification Dataset Example for the System Identification Toolbox of Matlab. Example process data from a SISO model: (Dynamic process data file in ascii format, Matlab Dataset).
• 22. K. Sigmon, "Matlab Primer Third Edition". University of Florida, Florida, Second Edition ed., 1992. (file in formato pdf ).
• Get Started with System Identification Toolbox System Identification Graphical User Interface for Matlab. (Matlab website, English).
• System Identification Toolbox System Identification Graphical User Interface for Matlab. (pdf format file).

General References and Textbooks

• SID, "System Identification Toolbox". System Identification Graphical User Interface for Matlab (pdf format file).
• System Identification: Theory for the User, Lennart Ljung - Springer, 1999.
• Model-based Fault Diagnosis in Dynamic Systems Using Identification Techniques, S. Simani, C. Fantuzzi, R. J. Patton - Springer, 2003.
• System Identification, T. Soderstrom, P. Stoica, Prentice Hall International, Cambridge, 1989

Fundamental Publications on System Identification and Data Analysis

Books

1. Ljung, L. (1999). System Identification: Theory for the User (2nd ed.). Prentice Hall, Upper Saddle River, NJ, USA.
2. Ljung, L., Glad, T. (2021). Modeling and Identification of Dynamic Systems. Studentlitteratur AB, Lund, Sweden.
3. Soderstrom, T., Stoica, P. (1989). System Identification. Prentice Hall International, Hemel Hempstead, UK.
4. Sintelon, R., Schoukens, J. (2012). System Identification: A Frequency Domain Approach (2nd ed.). Wiley-IEEE Press, Hoboken, NJ, USA.
5. Nelles, O. (2001). Nonlinear System Identification: From Classical Approaches to Neural Networks. Springer, Berlin, Germany.
6. Goodwin, G.C., Payne, R.L. (1977). Dynamic System Identification: Experiment Design and Data Analysis. Academic Press, New York, NY, USA.

Journal Articles

7. Schoukens, J., Ljung, L. (2019). Nonlinear system identification: A user-oriented roadmap. IEEE Control Systems Magazine, 39(6), 28 - 99.
8. Pillonetto, G., Ljung, L., Chen, T. (2023). Deep networks for system identification: A survey. Annual Reviews in Control, 56, 242 - 264.
9. Wahlberg, B., Ljung, L. (2018). Algorithms and performance analysis for stochastic Wiener system identification. Automatica, 95, 277 - 289.
10. Brunton, S.L., Proctor, J.L., Kutz, J.N. (2016). Sparse Identification of Nonlinear Dynamical Systems with Control (SINDYc). IFAC-PapersOnLine, 49(18), 710 - 715.
11. Pillonetto, G., Dinuzzo, F., Chen, T., De Nicolao, G., Ljung, L. (2014). Kernel methods in system identification, machine learning and function estimation: A survey. Automatica, 50(3), 657 - 682.
12. Ljung, L. (2010). Perspectives on system identification. Annual Reviews in Control, 34(1), 1 - 12.
13. Verdult, V., Ljung, L. (2005). Nonlinear state-space system identification with application to neural networks. Automatica, 41(11), 1771 - 1784.
14. Ljung, L. (2008). Perspectives on the relation between artificial intelligence and system identification. IFAC Proceedings Volumes, 41(2), 1 - 12.
15. Gevers, M., Ljung, L. (1986). Optimal experiment designs with respect to the intended model application. Automatica, 22(5), 543 - 554.

Software Resources

16. Ljung, L. (2024). System Identification Toolbox (version 10.1). MathWorks, Natick, MA, USA. (PDF format file).

Previous Page

Simani Home Page

Dipartimento di Ingegneria