Rail Wheelset Anomaly Detection via Distributed Acoustic Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting wheel flats in train wheels are time-consuming and prone to false alarms, as they rely on visual inspection or acoustic monitoring that struggles to discriminate repetitive noise from other sources, posing safety risks due to potential derailment.
Innovation Solution
The use of fibre optic distributed acoustic sensing (DAS) along rail tracks to analyze acoustic signals for characteristic frequencies proportional to train speed, allowing for the detection of anomalies like wheel flats by differentiating them from other noise sources through a distributed sensing array.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If acoustic monitoring is used to detect wheel flats, then detection coverage is improved, but false alarms increase due to inability to discriminate repetitive noise from other sources
Solution Approach 1:
The patent segments the acoustic monitoring task by analyzing acoustic signals from different spatial locations along the track separately. By dividing the track into multiple sensing zones and analyzing signals from each zone independently, the system can identify the spatial pattern characteristic of wheel flat noise (repetitive impacts at specific locations) versus other noise sources, thereby improving discrimination capability and reducing false alarms
Solution Approach 2:
The patent adds a spatial dimension to the acoustic analysis by deploying distributed sensors along the track and analyzing the positional information of acoustic signals. This transforms the problem from simple acoustic detection to spatial-temporal pattern recognition, enabling the system to distinguish wheel flat noise (which occurs at specific spatial intervals corresponding to wheel rotation) from other noise sources
2Reliability
If visual inspection is used to detect wheel flats, then false alarms are reduced, but detection time and productivity deteriorate
Solution Approach 1:
The patent replaces the mechanical visual inspection process with an acoustic sensing system. Instead of requiring inspectors to physically examine each wheel, the system uses distributed acoustic sensors to automatically detect the characteristic impact noise of wheel flats, enabling continuous monitoring without stopping trains and dramatically improving inspection productivity while maintaining high detection accuracy
Solution Approach 2:
The system enables the track infrastructure itself to perform detection by using the rail and surrounding structures as acoustic waveguides. The acoustic sensors deployed along the track passively capture noise generated by passing trains, eliminating the need for active inspection equipment on the trains or manual inspection processes
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This method provides improved detection and discrimination of wheel flats by analyzing acoustic signals over long distances and varying train speeds, enhancing the signal-to-noise ratio and reducing false alarms, thus enabling more effective preventative maintenance.
Implementation Method 1
taking data acquired by a distributed acoustic sensor having a sensing optical fibre deployed along at least part of a rail track as the train moves along that part of the track, wherein the data corresponds to acoustic signals detected by a plurality of longitudinal sensing portions of said sensing optical fibre
Data Source
Figure 1~3
Figure 4~5
AI summary
This application relates to methods and apparatus for the detection of anomalies in the wheelsets of rail vehicles, for instance for detection of defects such as wheel flats (303) of a wheel (301). The method using a distributed acoustic sensor (106) having a sensing optical fibre (104) deployed along at least part of a rail track (201) as a train (202) moves along that part of the track. The distributed acoustic sensor detects acoustic signals from a plurality of longitudinal sensing portions of the sensing optical fibre. A processor 108 analyses the acoustic signals to determine the train speed v. Having determined the train speed the processor also analyses the acoustic signals for a characteristic acoustic signal so as to detect an anomaly in a wheelset where the characteristic acoustic signal is based on the determined train speed. In particular the method may involve at least a first section of track where the train travels with a first speed v1 and a second section of track where the train travels with a second different speed v2 the characteristic acoustic signal may be a repetitive signal at a frequency that varies proportional to the train speed.