Stable learning scheme for failure detection and accommodation

This paper presents a methodology for constructing automated fault diagnosis and accommodation architectures using online approximators and adaptation/learning schemes. In this framework, neural network models constitute an important class of online approximators. Changes in the system dynamics are monitored by an online approximation model, which is used not only for detecting but also for accommodating system failures. A systematic procedure for constructing nonlinear estimation algorithms and stable learning schemes is developed, and simulation studies are used to illustrate the results.<
>