Information Estimation on Wheelset State and Track Features of Railway Vehicles

In order to get the wheelset state and track feature used for rail vehicle wheelset active control in an easy and economical way, an information estimation scheme based on state observers is proposed to simultaneously estimate the wheelset state and unknown track feature. The scheme obtains wheelset yaw velocity and lateral acceleration signal using some inertia sensors (accelerator and gyros). In addition, a single independent rotating wheelset model is set up to simulate and verify the proposed estimation scheme. Simulations show that the estimation system using a reduced order observer and a higher-order sliding mode observer presents good estimating performance, and is hardly affected by track irregularity. Errors of wheelset state estimation are less than 0.10%. In the ideal track, the error of the estimated track curvature and superelevation in curved track is no more than 0.26% and 13.40%, respectively. In the real track with irregularity, the error of the estimated track curvature is no more than 9.10%, but the filter fluctuation error should be removed when using this scheme in practical applications.