Interval Sliding Mode Observer Based Incipient Fault Detection with Application to a High-Speed Railway Traction Device

In this paper, a novel interval sliding mode observer is designed to detect incipient faults for a class of non-Lipschitz nonlinear systems with mismatched uncertainties. The interval estimation concept is introduced to design interval estimator for the nonlinear subsystem with uncertainties bounded by known intervals. Then novel injection functions are designed to ensure that the sliding motion takes place and maintains thereafter. At last, new residual generators and adaptive threshold generators are designed, and the corresponding fault detectability is studied. Case study on a traction device in CRH (China Railway High-Speed) is presented to demonstrate the effectiveness of proposed incipient fault detection scheme.