Monitoring Rail Condition Based on Sound and Vibration Sensors Installed on an Operational Train
The paper presents a measurement setup capable of collecting wheel/rail contact noise and vibration signals from a passenger train. A data analysis method based on machine learning is developed for detecting events from the acquired data and classifying them according to relevant railway track components and noise phenomena. A classification rate higher than 84 % is achieved.
KeywordsFalse Positive Rate Gaussian Mixture Model Sound Pressure Level True Positive Rate Validation Dataset
Unable to display preview. Download preview PDF.
- Lozano-Angulo, J.A.: Detection and one class classification of transient events in train track noise. Master’s Thesis, Technical University of Denmark, Denmark (2012)Google Scholar
- Masri, P.: Computed modelling of sound for transformation and synthesis of musical signal. Ph.D. dissertation, University of Bristol, UK (1996)Google Scholar
- Mediante, E.C.: Sound recognition techniques: application to city noise. Bachelor thesis, Technical University of Denmark, Denmark (2012)Google Scholar