A Physics-Guided Quantitative GPR Framework for Detecting Hanging Sleepers in Ballasted Railway Tracks

Area of Science:

• Geophysics
• Civil Engineering
• Railway Engineering

Background:

• Sleeper voids, or hanging sleepers, in ballasted railway tracks pose significant risks to structural integrity and operational safety.
• Existing detection methods may lack the precision required for early identification and severity assessment of these defects.

Purpose of the Study:

• To develop and validate a physics-guided quantitative ground-penetrating radar (GPR) framework for detecting hanging sleepers.
• To enable accurate estimation of the elevation difference between sleepers and ballast interfaces for severity assessment.
• To provide a physically interpretable and automated solution for railway track inspection.

Main Methods:

• Utilizing high-frequency antennas (f≥1.5 GHz) for enhanced GPR signal acquisition.
• Implementing a framework involving signal post-processing, sleeper-region localization, and time-domain peak searching with polarity consideration.
• Employing continuous wavelet transform (CWT) for auxiliary verification and spatial continuity constraints to minimize false detections.

Main Results:

• The proposed GPR framework successfully detects hanging sleepers by quantitatively estimating elevation differences based on physical geometric relationships.
• Validation via gprMax simulations and field experiments confirmed the framework's effectiveness in detection and severity assessment.
• The method provides a physically interpretable basis for identifying hanging sleepers according to established engineering criteria.

Conclusions:

• The physics-guided GPR framework offers a robust and accurate solution for detecting hanging sleepers in ballasted railway tracks.
• This approach significantly enhances the capability for automated railway inspection, improving safety and serviceability.
• The quantitative assessment of sleeper voids provides valuable data for predictive maintenance and structural health monitoring of railway infrastructure.