Data-driven fault detection of vertical rail vehicle suspension systems

This paper concerns data driven fault detection of vertical rail vehicle suspension systems issue. The underlying vehicle system are equipped with only accelerator sensors in the four corners of the carbody, the front and trail bogie, respectively. The faults considered are the vertical damper fault and vertical spring fault. Both PCA-based and CVA-based fault detection methods are studied in this paper. When there is a detectable fault, the detector sends an alarm signal if the residual evaluation is larger than a predefined threshold. By using the professional multi-body simulation tool, SIMPACK, the effectiveness of the proposed approach is demonstrated by simulation results for several fault scenarios.