Robust Parameter Identification for Railway Suspension Systems

Detecting faults at the right time is crucial both for adequate maintenance and optimal operation of railway vehicles. The increasing complexity of the modern railway system pose a challenge for appropriate fault diagnosis and isolation (FDI). In this paper, a sliding-mode-based parameter identification algorithm is proposed as a solution for the identification of parameters in a simplified railway system. It is assessed on different faulty cases where the accuracy of the estimation and sensitivity to faults are shown, as well as robustness to model uncertainties. The data used for simulations and algorithm evaluation is generated using a validated Multibody Simulation (MBS) model implemented in Simpack, while the numerical simulations are carried out through MATLAB/Simulink.