On the Use of Wavelet Analysis for the Detection and Position Estimation of Short Wavelength Rail Head Faults

In railway condition monitoring, the use of contact forces for infrastructure diagnosis is not yet widely adopted. However, wheel-rail interaction data collected via instrumented wheelsets can provide strategic insights into short-wavelength irregularities and faults that are often overlooked by conventional optical equipment commonly used in onboard track geometry monitoring. Unlike rail geometry measurements, force-based measurements are particularly sensitive to impulsive irregularities and defects, including rail head issues such as squats, voids, dipped joints, corrugation, and weld irregularities. Consequently, data from instrumented wheelsets can offer valuable information on both the magnitude and location of such irregularities along the track, for example, through wavelet analysis. This paper proposes a method based on continuous wavelet transform for identifying and localizing faults along the track using unconventional force recordings.