Detectability performance properties of learning-based nonlinear fault diagnosis

The learning approach to fault diagnosis provides a methodology for designing monitoring architectures which can be used for detection, identification and accommodation of failures in dynamical systems. This paper considers the issues of detectability conditions and detection time in a nonlinear fault diagnosis scheme based on the learning approach. First, conditions are derived to characterize the range of detectable faults. Then, non-conservative upper bounds are computed for the detection time of incipient and abrupt faults. Finally, it is shown that the detection time decreases monotonically as the values of certain design parameters increase.