A Robust Takagi-Sugeno Fault Diagnostic Scheme for Remaining Useful Life Estimation

The paper deals with a problem of remaining useful life estimation for a class of fuzzy systems described with a well-known Takagi-Sugeno modeling paradigm. Such an approach is especially worth of investigation due to the fact that it allows predicting a time to the failure of each actuator. As a result, it makes it possible to planning their maintenance aiming at expanding their useful lifetime. In order to do so, the proposed methodology employs a fault estimation strategy. Such a fault estimation is accompanied with a state estimation. The proposed methodology enables suitable robustness while estimating the state as well as both actuator and sensor faults. Subsequently, the obtained estimates are used for predicting the actuator fault evolution, and hence, assessing actuators remaining useful time. The final part of the paper illustrates an example with an application of the proposed approach to the laboratory multi-tank system.