MBPLS-based rail vehicle suspension system fault detection

The suspension system plays a crucial role of rail vehicles. The fault detection of the suspension system is an effective way to ensure the security, stable operation of rail vehicles. This paper concerns the fault detection issue of rail vehicle suspension systems with the extended form of Partial Least Squares(PLS), which is Multi-block Partial Least Squares (MBPLS). The signal information used in the fault detection is obtained from the SIMPACK and MATLAB co-simulation environment. In this paper, the typical primary spring and damper faults and secondary spring and damper faults are detected successfully using MBPLS. MBPLS is applied to the block data, and the statistical index SPE and T² are used to monitor the performance of the suspension system. Compared with DPCA, the effectiveness of the proposed approach is demonstrated by the simulation results for several fault scenarios of primary and secondary faults.