Design of fault diagnosis observer for a class of nonlinear systems

Rong SUN,Sheng LIU,Yu-fang ZHANG
DOI: https://doi.org/10.7641/CTA.2013.21297
2013-01-01
Abstract:Most of fault diagnosis algorithms deal with linear systems and nonlinear systems with states depending linearly on faults. According to essential properties of a nonlinear system, we propose a fault diagnosis algorithm for a class of nonlinear systems based on parameter estimation. The systems model is decomposed by the (B,K,φ) realization into models in which the states are affected nonlinearly by faults. By using the decoupling technology for disturbances, we make the resultant residuals to be completely robust to the unknown input disturbances. Stability of the algorithm is verified by using the Lyapunov function. Simulation results show that the proposed algorithm converges rapidly and provides perfect diagnosis for a class of nonlinear systems.
What problem does this paper attempt to address?