An inverse dynamics method for railway vehicle systems

Abstract

The wheel–rail action will obviously be increased during the vehicles in high-speed operation state. However, in many practical cases, direct measurement of the wheel–rail contact forces cannot be performed with traditional procedures and transducers. An inverse mathematical dynamic model for the estimation of wheel–rail contact forces from measured accelerations was developed. The inverse model is a non-iteration recurrence method to identify the time history of input excitation based on the dynamic programming equation. Furthermore, the method overcomes the weakness of large fluctuations which exist in current inverse techniques. Based on the inverse dynamic model, a high-speed vehicle multibody model with twenty-seven Degree of Freedoms (DOFs) is established. With the measured responses as input, the inverse vehicle model can not only identify the responses in other parts of vehicle, but also identify the vertical and lateral wheel–rail forces respectively. Results from the inverse model were compared with experiment data. In a more complex operating condition, the inverse model was also compared with results from simulations calculated by SIMPACK.

First published online: 22 May 2013