Heavy-haul railways play a crucial role in bulk freight transportation, particularly for coal and mineral resources. However, under long-term high-axle-load operation, rails in small-radius curve sections are subjected to intense wheel–rail interaction forces. These conditions accelerate defect formation such as head checks and bolt-hole cracks, which can ultimately lead to rail fracture if not detected in time.
Ultrasonic inspection is the primary non-destructive testing method used in heavy-haul railway maintenance. In practice, inspection vehicles collect ultrasonic A-scan signals to identify internal rail defects. However, in small-radius curves, strong flange-related structural vibrations, low-frequency disturbances, and high-frequency electrical noise are coupled together. These non-Gaussian noise components overlap spectrally and temporally with defect echoes, severely masking critical defect information.
The team published their study in Communications in Transportation Research (https://doi.org/10.26599/COMMTR.2026.9640021).
"Traditional denoising methods assume relatively simple noise characteristics," the researchers explain. "But in real heavy-haul environments, noise is strongly coupled and non-Gaussian. Simply applying low-pass filtering or single-domain feature extraction is often insufficient."
To address this challenge, the research team first constructed a physics-based ultrasonic A-scan signal model under noise-coupled conditions. The model explicitly characterizes defect echoes together with structural vibration noise, low-frequency irrelevant components, and high-frequency electrical noise. This modeling step provides a theoretical foundation for understanding why conventional signal processing methods fail under complex field conditions.
Based on this signal model, the team developed a multi-feature fusion filtering framework within an ideal binary mask (IBM) paradigm. The framework integrates three complementary feature extraction mechanisms: Variational Mode Decomposition (VMD)–based inter-layer correlation analysis to capture synchronized multi-mode defect responses; Continuous Wavelet Transform (CWT) time–frequency clustering to identify triangular energy ridges associated with defect echoes; Sliding-window waveform morphology analysis using kurtosis and peak-width features to discriminate impulsive defect pulses from background noise.
Rather than treating feature fusion as simple vector concatenation, the proposed method performs decision-level fusion based on physically interpretable indicators from time, frequency, and modal domains. The fused features are used to construct an indexed binary mask, which selectively preserves defect-related samples while suppressing unsupported noise components.
Field-based experiments were conducted using 285 simulated noise-contaminated defect signals derived from real inspection data. Results show that the proposed method achieves accurate defect localization in more than 90% of test cases. Compared with single-feature methods, the multi-feature fusion framework significantly reduces missed detections and large localization deviations.
From an engineering perspective, this improvement directly enhances the accuracy of B-scan image reconstruction and defect positioning. By operating at the A-scan level and preserving peak amplitude integrity, the method provides practical support for intelligent condition-based maintenance of heavy-haul railways.
The researchers note that future work will extend the framework to multi-channel ultrasonic inspection systems, where inter-channel coupled noise presents additional challenges.
DOI Link:
https://doi.org/10.26599/COMMTR.2026.9640021
About Communications in Transportation Research
Communications in Transportation Research was launched in 2021, with academic support provided by Tsinghua University and China Intelligent Transportation Systems Association. The Editors-in-Chief are Professor Xiaobo Qu, a member of the Academia Europaea from Tsinghua University, and Professor Xiaopeng (Shaw) Li from University of Wisconsin–Madison. The journal mainly publishes high-quality, original research and review articles that are of significant importance to emerging transportation systems, aiming to serve as an international platform for showcasing and exchanging innovative achievements in transportation and related fields, fostering academic exchange and development between China and the global community.
It has been indexed in SCIE, SSCI, Ei Compendex, Scopus, CSTPCD, CSCD, OAJ, DOAJ, TRID and other databases. It was selected as Q1 Top Journal in the Engineering and Technology category of the Chinese Academy of Sciences (CAS) Journal Ranking List. In 2022, it was selected as a High-Starting-Point new journal project of the "China Science and Technology Journal Excellence Action Plan". In 2024, it was selected as the Support the Development Project of "High-Level International Scientific and Technological Journals". The same year, it was also chosen as an English Journal Tier Project of the "China Science and Technology Journal Excellence Action Plan Phase Ⅱ". In 2024, it received the first impact factor (2023 IF) of 12.5, ranking Top1 (1/58, Q1) among all journals in "TRANSPORTATION" category. In 2026, its 2025 IF was announced as 12.7, maintaining the Top1 position (1/66, Q1) in the same category.
From Volume 6 (2026), Communications in Transportation Research will be published by Tsinghua University Press on the SciOpen platform with the official journal website at https://www.sciopen.com/journal/2097-5023 . We kindly request that all new manuscript submissions be made through the journal's submission system at https://mc03.manuscriptcentral.com/commtr
