An approach for the estimation of vertical wheel/rail force using dynamic signals

Abstract

The identification of Wheel/Rail Force (WRF) based on acceleration is a cheaper and convenient way to ensure safe operation. For this purpose, this paper describes an approach for the estimation of Vertical Wheel/Rail Force (VWRF) using dynamic signals from a track inspection train (TIT). Based on the vehicle dynamics theory, a VWRF estimation model is established, by which VWRF can be estimated with the measured vehicle acceleration. Rolling angle is a non-constant and hard-measured value which will affect estimation results. By deducing its relation with the difference between left and right track profiles, it can be got through the Vertical Axle Box Acceleration (VABA) and the length of the axle box. VWRF cannot be estimated only by filtering the Axle Box Acceleration (ABA). The mass participating in vibration (PVM) also has a great influence on it. With many tests and calculations, the relationship between the PVM and the wheelset mass is concluded, which made the estimation results much more accurate. The proposed method provides the basis for online high-precision estimate WRF and ensures safe operation. By comparing with a commercial WRF Inspection System, the railway application results identified that the proposed method can produce a satisfactory result.