Rotation Speed Estimation for Instrumented Wheelset Using Auto-Regressive Model

Abstract

This paper describes an estimation procedure of the wheel rotation speed of the instrumented wheelset by using auto-regressive model. Some of the recent signal processing techniques for the instrumented wheelset require an accurate estimation of the wheel rotation speed. An auto-regressive model is appropriate for this purpose since auto-regressive model can estimate power spectrum of signals in high accuracy compared with the discrete Fourier transform. This paper describes the result of basic verification on the use of the auto-regressive model for the signal processing of the instrumented wheelset.