An impulsive noise filter for rail vibration measurements using a laser Doppler vibrometer on a moving platform

This paper proposes a new impulsive noise (i.e., IN) filter for non-contact rail vibration measurements performed by a laser Doppler vibrometer (i.e., LDV) placed on a moving rail car. The development of the filter is motivated by the fact that such measurements can be contaminated by IN, which is caused by a high level of speckle noise due to the movement of the LDV. Consequently, some portions of the LDV signals can be distorted, leading to missing measurements from some rail segments. The proposed filter consists of two main steps: (i) detecting the signal segments containing the IN (i.e., detection step), and (ii) replacing the detected segments to estimate their IN-free values (i.e., estimation step). The performance of the filter was evaluated on a synthetic signal, which was created by combining a rail vibration signal in the vertical direction (i.e., vibration component) with a speckle noise signal of an LDV placed on a moving platform (i.e., noise component). The reason to use a synthetic signal was to know the vibration and noise components separately. This allowed for the examination of IN's behavior, which showed that IN manifests itself as artificial peaks as well as high-frequency noise in the proximity of artificial peaks (i.e., artificial fluctuations). Accordingly, the proposed filter was designed to detect and replace both artificial peaks and fluctuations. Qualitative and quantitative evaluations of the filtered synthetic signal showed that the proposed filter exhibited good performance in mitigating the IN. © 2024 Elsevier Ltd