Condition monitoring of railway vehicle suspension using adaptive multiple model approach

This paper demonstrates the possibility to detect suspension failures of railway vehicles using a multiple-model approach from on-board measurement data. The railway vehicle model used includes the lateral and yaw motions of the wheelsets and bogie, and the lateral motion of the vehicle body, with sensors measuring the lateral acceleration and yaw rate of the bogie, and lateral acceleration of the body. The detection algorithm is formulated based on the Interacting Multiple-Model (IMM) algorithm adding a method updating estimation model. The IMM method has been applied for detecting faults in vehicle suspension systems in a simulation study. The mode probabilities and states of vehicle suspension systems are estimated based on a Kalman filter (KF). This algorithm is evaluated in simulation examples. Simulation results indicate that the algorithm effectively detects on-board faults of railway vehicle suspension systems in realistic situation.