Author(s)

YASUKUNI NAGANUMA, TAKAYUKI UEMATSU

Abstract

This paper describes a method for restoring actual track geometry using data measured with a newly developed inertial trolley and the application of this method. First, we present the result of an estimation of actual geometry obtained through the application of an inverse filter, which is a classical digital signal processing technique. Recently, the authors developed a method of restoring actual track geometry using a Bayesian filter. In this paper, we show that the restoration by the Kalman filter, which is the most popular Bayesian filter, is superior to the inverse filter. In addition, as an application of this technique, we proposed a compensation method for gauge and cant irregularity data measured by means of a three-wheel trolley, and we demonstrated that gauge irregularity as defined may be obtained without any additional sensors.

Keywords

actual track geometry, track geometry measuring trolley, inverse filter, Kalman filter