An adaptive robust framework for model-based state fault detection

A goal in many applications is to combine a priori knowledge of the physical system with experimental data to detect faults in a system at an early enough stage as to conduct preventive maintenance. The information available beforehand is the mathematical model of the physical system and the key issue in the design of model-based fault detection is the effect of model uncertainties such as severe parametric uncertainties and unmodeled dynamics on their performance. This paper presents the application of a nonlinear model-based adaptive robust state fault detection that combines online parameter adaptation with robust filter structures to reduce the extent of model uncertainty to help in the improvement of the sensitivity of the fault detection scheme to faults. Simulation results are presented to demonstrate the superior performance of the proposed scheme in the early and reliable detection of incipient faults