Estimation of primary suspension parameters from lateral dynamic response of a wheelset

Abstract

Globally, maintenance tasks for railway vehicles have generally calendar based schedules. However, apparent changes may occur in vehicle structure and environment. Suspension malfunction and substantial change in adhesion conditions can be given as an example for such situations. Since these kinds of changes may affect especially safety, necessary actions must be taken as soon as possible without waiting schedule. This is possible with condition monitoring systems, which lead vehicles to be smarter, as they can inform decision makers for actions. Dynamic response of a vehicle, which includes information about changes in vehicle's structure and environment, can be used for condition monitoring. In this study, a condition monitoring scheme is proposed to identify primary lateral suspension parameter from dynamic response of a wheelset. Identification is based on the well-known model based filtering method, namely unscented Kalman filter.