An iterative learning observer-based approach to fault detection and accommodation in nonlinear systems

This paper presents a general framework for fault detection and accommodation using iterative learning strategy. An iterative learning observer (ILO) updated online by past system output errors as well as input is constructed for the purpose of fault detection. This observer is different from conventional Luenberger observer where the observer's state is only a function of the most recent input, output and the estimation error. Furthermore, using iterative learning (IL) approach, an automatic control reconfiguration scheme for accommodation of faults is also described. One of the main features of the proposed scheme is that the control reconfiguration is achieved automatically based only on the response of the overall systems, and the IL controller does not require a fault detection and isolation subsystems. An example is employed to verify the effectiveness of the IL observer and IL fault accommodation scheme.