Adaptive Fuzzy Observer-Based Fault Estimation for a Class of Nonlinear Stochastic Hybrid Systems

Shasha Fu (1) , Jianbin Qiu (2, 3) , Liheng Chen (4, 5) , Mohammed Chadli (6)

1 School of Electrical Engineering
2 State Key Laboratory of Robotics and Systems
3 Research Institute of Intelligent Control and Systems
4 College of Intelligent Systems Science and Engineering
5 Engineering Research Center of Navigation Instruments
6 IBISC - Informatique, BioInformatique, Systèmes Complexes

This article studies the fault estimation problem for a class of continuous-time nonlinear Markovian jump systems with unmeasured states, unknown bounded sensor faults, and unknown nonlinearities simultaneously. In this article, a new adaptive fuzzy observer design scheme is developed, where the completely unknown nonlinear terms are approximated by adaptive fuzzy logic systems. By means of a novel online adaptive mechanism, the asymptotic stability of the error dynamic system is guaranteed despite of sensor faults and unknown nonlinear terms. Moreover, the sliding surface switching problem in the traditional sliding mode observer techniques can be avoided for Markovian jump systems. Finally, two practical examples are given to demonstrate the effectiveness of the proposed approach.

Keywords: Adaptive fault estimation; Fuzzy logic systems (FLSs); Markovian jump systems (MJSs); Sensor faults

Domains: Automatic

Shasha Fu, Jianbin Qiu, Liheng Chen, Mohammed Chadli. Adaptive Fuzzy Observer-Based Fault Estimation for a Class of Nonlinear Stochastic Hybrid Systems. IEEE Transactions on Fuzzy Systems, 2022, 30 (1), pp.39--51. ⟨10.1109/TFUZZ.2020.3031033⟩. ⟨hal-03540339⟩