Observer-based fault detection for Rail Vehicle Suspension Systems

This paper is concerned with the design of an observer-based fault detection module for nonlinear rail vehicle suspension systems with the method of restructuring states and faults which can affect the system performance. An alarm signal will be sent by the designed observer if the residual evaluation is larger than a prescribed threshold when there are detectable actuator and sensor faults. The track irregularity generated by the spatial power spectral density is regarded as the system disturbances. Simulation result shows the effectiveness of the proposed method for Rail Vehicle Suspension Systems´ fault detection.