Railway track condition monitoring using dynamic measurements on in-service vehicles

Railway track is one of the most important parts for the railway system and its condition monitoring is essential to ensure the safety of the system. Currently, the track condition is inspected by the stationary equipments or specified low-speed track inspection vehicle. The track inspection needs close the railway system. An adaptive regularization approach has been developed to identify the parameters of the track substructure from dynamic measurements on the in-service vehicles in this paper. The vehicle-track interaction system is modelled as the discrete mass-spring model on Winkler elastic foundation. The damage is defined as the stiffness reduction of the track that includes the foundation settlement, damaged rail fastener and damaged ballast. Numerical results show that the approach can identify all stiffness parameters successfully at normal moving speed and in high sampling frequency with measurement noise.