The effect of driving direction on spatially aligned track recording car measurements in turnouts

To monitor rail track quality the track is measured by parameters from a track recording car. These parameters are often hard to interpret in a turnout due to the geometric complexity. A new data processing method is presented based on the track recording car parameters are all mathematically unified to one driving direction, which is straight and facing in a left-hand turnout. The unified recording car data are then spatially aligned using the cross-correlation function (CCF) to adjust for small spatial dislocations from the GPS location system. Each turnout has assigned a specific driving direction of the passing trains and the effect of the driving directions on the recording car parameter patterns are investigated with locally weighted regression. Distinct patterns are observed for each of the two directions. Manual gauge measurements are compared to the gauge parameter from the track recording car for the trailing direction versus the facing direction.