Distributed fibre optic sensing for monitoring rail networks

• US 10,850,754 B2
• Filed: 12/01/2016
• Issued: 12/01/2020
• Est. Priority Date: 12/01/2015
• Status: Active Grant

• Current Assignee: Optasense, Inc. (Luna Innovations, Inc.)
• Original Assignee: Optasense, Inc. (Luna Innovations, Inc.)
• Inventors: Cole, Stephen Patrick
• Primary Examiner(s): Smith, Jason C
• Application Number:
  US15/778,867
• Publication Number:
  US 20180354534A1
• Time in Patent Office: 1,461 Days
• CPC Class Codes: B61L 23/041 Obstacle detection
• CPC Class Codes: B61L 23/30 using automatic section blocking
• CPC Class Codes: B61L 25/021 Measuring and recording of train speed
• CPC Class Codes: B61L 25/025 Absolute localisation, e.g. providing geodetic coordinates
• CPC Class Codes: B61L 3/06 controlling by electromagnetic or particle radiation, e.g. by light beam
• CPC Class Codes: G01H 9/004 using fibre optic sensors light guides per se G02B6/00, acousto-optical devices specially adapted for gating or modulating in optical wave guides G02F1/125

Abstract

This application describes methods and apparatus for monitoring of rail networks using fibre optic distributed acoustic sensing (DAS), especially for condition monitoring. One method involves taking (902) a first data set corresponding to measurement signals from a plurality of channels of at least one fibre optic distributed acoustic sensor (100) having a sensing fibre (101) deployed to monitor at least part of the path of the rail network (201). The first data set corresponds to measurement signals acquired as a train (202) passed along a first monitored section of the rail network. The method involves identifying (903) a speed of the train through the first monitored section and dividing (904) the first data set into a plurality of time windows. Each time window contains a different subset of the first data set, with the measurement signal for each successive channel in a time window being delayed with respect to the previous channel by a time related to the identified train speed. For each time window, any appropriate time shift is identified (905) and applied (906) to the measurement signals for a channel so as to substantially align the measurement signals of the channels within the time window. The data from the time windows is then combined (907) after any time shifts have been applied to form an aligned first data set; and a characteristic train signal is derived (908) from the aligned first data set. The characteristic signal may be removed from the aligned first data set (1007) to leave remainder data. The characteristic trains signal and/or remainder data may be analysed for condition monitoring.

Claims

1. A method of monitoring a rail network comprising:

taking a first data set corresponding to measurement signals from a plurality of channels of at least one fibre optic distributed acoustic sensor having a sensing fibre deployed to monitor at least part of the path of the rail network, wherein the first data set corresponds to measurement signals acquired as a train passed along a first monitored section of the rail network;

identifying a speed of the train through the first monitored section;

dividing the first data set into a plurality of time windows, each time window comprising a different subset of the first data set, wherein each time window comprises measurement signals from the plurality of channels, the measurement signal for each successive channel in a time window being delayed with respect to the previous channel by a time related to the identified train speed;

for each time window, identifying and applying any time shift to the measurement signals for a channel so as to substantially align the measurement signals of the channels within the time window;

combining the time windows after any time shifts have been applied to form an aligned first data set; and

deriving a characteristic train signal from the aligned first data set.

2. A method as claimed in claim 1 wherein said time windows overlap with one another.

3. A method as claimed in claim 1 wherein the characteristic train signal is derived by summing the measurement signals of a plurality of the channels of the aligned first data set.

4. A method as claimed in claim 1 wherein identifying and applying any time shift to the measurement signals for a channel so as to substantially align the measurement signals of the channels within the time window comprises summing the measurement signals of at least some of the channels to form a pilot signal and iteratively for each channel identifying a time shift to align that measurement signal for that channel to the pilot signal and updating the pilot signal.

5. A method as claimed in claim 1 wherein identifying and applying any time shift to the measurement signals for a channel so as to substantially align the measurement signals of the channels within the time window comprises applying at least first and second alignment passes.

6. A method as claimed in claim 5 wherein the first alignment pass is applied to measurement signals in the time windows that have been at least one:

low pass filtered and processed to a lower resolution.

7. A method as claimed in claim 6 wherein the first alignment pass is used to determine a first set of time shifts which are applied to the channels each time window before the second alignment pass.

8. A method as claimed in claim 1 further comprising removing the characteristic train signal from a data set of interest to leave a remainder data set.

9. A method as claimed in claim 8 wherein the data set of interest comprises the aligned first data set.

10. A method as claimed in claim 8 wherein the data set of interest comprises an aligned second data set, the aligned second data set corresponding to a second data set of measurement signals from a plurality of channels of at least one fibre optic distributed acoustic sensor acquired as the train passed along a second monitored section of the rail network that has been processed to align the measurement signals from different channels.

11. A method as claimed in claim 10 further comprising forming the aligned second data set by:

taking the second data set;

identifying a speed of the train through the second monitored section;

dividing the second data set into a plurality of time windows, each time window comprising a different subset of the second data set, wherein each time window comprises measurement signals from the plurality of channels, the measurement signal for each successive channel in a time window being delayed with respect to the previous channel by a time related to the identified train speed through the second section;

for each time window, identifying and applying any time shift to the measurement signals for a channel so as to substantially align the measurement signals of the channels within the time window;

combining the time windows after any time shifts have been applied to form the aligned second data set.

12. A method as claimed in claim 8 comprising processing the remainder data set to reverse any time shifts applied to align the measurement data from different channels.

13. A method as claimed in claim 8 wherein a deconvolution is applied to the data set of interest based on the characteristic train signal to leave the remainder data set.

14. A method as claimed in claim 8 comprising applying interferometric imaging techniques to the data set of interest based on the characteristic train signal as a source signal.

15. A method as claimed in claim 8 comprising analysing the remainder data set with respect to one or more previously acquired remainder data sets for the same section of track.

16. A method as claimed in claim 15 comprising comparing the remainder data set with at least one reference track profile.

17. A method as claimed in claim 15 comprising performing trend analysis based on said remainder data set and said one or more previously acquired remainder data sets.

18. A method as claimed in claim 15 comprising identifying any train related components of signals in the remainder data.

19. A method as claimed in claim 1 comprising analysing the characteristic train signal by comparing the characteristic train signal to one or more reference train signals for that train.

20. A rail network monitoring apparatus comprising;

a memory for storing a first data set corresponding to measurement signals from a plurality of channels of at least one fibre optic distributed acoustic sensor having a sensing fibre deployed to monitor at least part of the path of the rail network, wherein the first data set corresponds to measurement signals acquired as a train passed along a first monitored section of the rail network; and

a processor configured to;

identify a speed of the train through the first monitored section;

divide the first data set into a plurality of time windows, each time window comprising a different subset of the first data set, wherein each time window comprises measurement signals from the plurality of channels, the measurement signal for each successive channel in a time window being delayed with respect to the previous channel by a time related to the identified train speed;

for each time window, identify and apply any time shift to the measurement signals for a channel so as to substantially align the measurement signals of the channels within the time window;

combine the time windows after any time shifts have been applied to form an aligned first data set; and

derive a characteristic train signal from the aligned first data set.

21. A rail network monitoring apparatus comprising;

a memory for storing a first data set corresponding to measurement signals from a plurality of channels of at least one fibre optic distributed acoustic sensor having a sensing fibre deployed to monitor at least part of the path of the rail network, wherein the first data set corresponds to measurement signals acquired as a train passed along a first monitored section of the rail network; and

a processor configured to;

identify a speed of the train through the first monitored section;

divide the first data set into a plurality of time windows, each time window comprising a different subset of the first data set, wherein each time window comprises measurement signals from the plurality of channels, the measurement signal for each successive channel in a time window being delayed with respect to the previous channel by a time related to the identified train speed;

for each time window, identify and apply any time shift to the measurement signals for a channel so as to substantially align the measurement signals of the channels within the time window;

combine the time windows after any time shifts have been applied to form an aligned first data set; and

remove a characteristic train signal from the aligned first data set to leave a remainder data set.