This paper presents a fault detection and isolation(FDI) scheme for a nonlinear
systems with a actuator and sensor faults. A residual generator based on the
observer model generate the information for a fault detection. The proposed fault
estimators are activated for a fault isolation and applied to estimate the
time-varying lumped faults(model uncertainty + fault). but a fault estimator
error dose not converge to zero since the derivative of lumped fault is not zero.
Then the fuzzy neural network(FNN) is used to estimate the fault estimator error.
Simulation results are presented to illustrate the effectiveness and the
applicability of the approaches proposed.