Method of Fault Identification in Technical Systems Based on High Order Sliding Mode Observers

The problem of fault identification in technical systems described by linear equations under the external disturbances is considered. The problem is solved based on sliding mode observers of two types: on the basis of high-order observers and on the basis of observers with weakened existing conditions. The peculiarity of such observers is that they do not rely on the matching condition, which is necessary for designing conventional sliding mode observers, but require the fulfillment of minimum phase condition for the original system. In contrast to the standard approach, both types of observers are designed not on the basis of the original system, but on the basis of its reduced-order model, which has selective sensitivity to faults and disturbances. The model is constructed in an identification canonical form and allows reducing the complexity of estimation procedure and removing the minimum phase condition. To implement the first type, an ordinary reduced-order Luenberger observer and a high-order sliding mode observer are constructed using information from the first observer; the identification expression uses information from the second observer and data on the coefficients of the characteristic equation of the first observer. To implement the second type, a reduced-order sliding mode observer is designed, which directly generates the required estimate. The disadvantage of high-order observers is that they require the linearity of the original system. In some cases, the disadvantage can be overcome by using virtual sensors, which are nonlinear observers that estimate the unmeasured components of the system state vector. An example is given that demonstrates the capabilities of virtual sensors. Based on the Matlab package, the original system and the constructed observers were simulated both in the absence of measurement noise and in their presence. The simulation confirmed the correctness of the assumptions and theoretical constructions and showed that high-order observers are less sensitive to such noises.

1. Zhirabok A., Zuev A., Sergiyenko O., Shumsky A. Identification of faults in nonlinear dynamical systems and their sensors based on sliding mode observers, Autom. Remote Control, 2022, vol. 83, pp. 214—236, DOI: 10.31857/S0005231022020052.

2. Edwards C., Spurgeon S., Patton R. Sliding mode observers for fault detection and isolation, Automatica, 2000, vol. 36, pp. 541—553, DOI: 10.1016/S0005-1098(99)00177-6.

3. Yan X., Edwards C. Nonlinear robust fault reconstruction and estimation using sliding mode observers, Automatica, 2007, vol. 43, pp. 1605—1614, DOI: 10.1016/j.automatica.2007.02.008.

4. Kalsi K., Hui S., Zak S. Unknown input and sensor fault estimation using sliding-mode observers, Proc. 2011 ACC, San Francisco, 2011, pp. 1364—1369.

5. Keijzer T., Ferrari R. Threshold design for fault detection with first order sliding mode observers, Automatica, 2022. vol. 146, art. 110600, DOI: 10.1016/j.automatica.2022.110600.

6. Hamdi H., Rodrigues M., Rabaout B., Braiek N. A fault estimation and fault tolerant control based sliding mode observer for LPV descriptor systems with time delay, Int. J. Appl. Math. Comput. Sci., 2021, vol. 31, pp. 247—258, DOI: 10.34768/amcs-2021-0017.

7. Sergiyenko O., Tyrsa V., Zhirabok A., Zuev A. Sliding mode observer based fault identification in automatic vision system of robot, Control Engineering Practice, 2023, vol. 139, art. 105614, DOI: 105614.10.1016/j.conengprac.2023.105614.

8. Floquet T., Barbot J.-P. A сanonical Form for The design of unknown input sliding mode observer, Advanced in Variable Structure, Eds. C. Edwards et al. Springer Verlag, 2006, pp. 217—292.

9. Fridman L., Levant A., Davila J. High-order sliding-mode observation and identification for linear systems with unknown inputs, Proc. 45th IEEE Conf, Decision & Control, 2006, pp. 5567—5572.

10. Fridman L., Levant A., Davila J. Observation of linear systems with unknown inputs via high-order sliding-modes, Int. J. Systems Science, 2007, vol. 38, pp. 773—791, DOI: 10.1080/00207720701409538.

11. Floquet T., Edwards C., Spurgeon S. On sliding mode observers for systems with unknown inputs, Int. J. Adaptive Control and Signal Processing, 2007, vol. 21, pp. 638—656, DOI: 10.1002/acs.958.

12. Shtessel Yu., Chava B., Edwards C. Second order sliding mode control using nonlinear dynamic sliding manifold: Lyapunov approach, 2015 IEEE 54th Annual Conf. Decision and Control (CDC), December 15—18, 2015, Osaka, Japan, pp. 4103—4108.

13. Levant A. Higher-order sliding modes, differentiation and output feedback control, Int. J. Control, 2003, vol. 76, pp. 924—941, DOI: 10.1080/0020717031000099029.

14. Wang X., Tan C., Zhou D. A novel sliding mode observer for state and fault estimation in systems not satisfying matching and minimum phase conditions, Automatica, 2017, vol. 79, pp. 290—295, DOI: 10.1016/j.automatica.2017.01.027.

15. Zhirabok A., Zuev A., Protsenko А., Gornostaev I., Serdukova A. Method of sensor fault identification based on high order sliding mode observers, Tomsk State University. J. of Control and Computer Science, 2025, no 73, pp. 81—89 (in Russian), DOI: 10.17223/19988605/73/10.

16. Zhirabok A., Zuev A., Filaretov V., Shumsky A. Fault identification in nonlinear systems based on sliding mode observers with weakened existence conditions, J. Computer Systems Sciences Int., 2022, vol. 61, pp. 313—321, DOI: 10.1134/S1064230722030169.

17. Witczak M. Fault diagnosis and fault tolerant control strategies for nonlinear systems, Berlin, Springer, 2014.

18. Zhirabok A. N., Zuev A. V., Bobko E. Yu. Method of virtual sensors design for replacement of faulty physical sensors, Mekhatronika, Avtomatizatsiya, Upravlenie, 2023, vol. 24, pp. 526—532 (in Russian), DOI: 10.17587/mau.24.526-532.