An OFDR strain demodulation method based on nearest neighbor sliding window threshold

Abstract

This invention relates to the field of fiber optic sensing technology, specifically an OFDR strain demodulation method based on a nearest-neighbor sliding window threshold. The method is implemented using the following steps: Step S1: Acquire a reference signal; Step S2: Set a threshold; Step S3: Acquire a measurement signal; Step S4: Acquire two differences; Step S5: Compare the two differences with the threshold; if the absolute values ​​of both differences are less than the threshold, proceed to Step S6; otherwise, proceed to Step S7; Step S6: Directly determine the final spectral offset; Step S7: Acquire a new reference signal; Step S8: Acquire multiple differences and determine the final spectral offset; Step S9: Calculate the strain value applied to the fiber based on the final spectral offset. This invention effectively solves the problem of accurate measurement when the strain is large using traditional OFDR demodulation technology, and is applicable to fields such as civil structure health monitoring, composite material fatigue testing, and temperature detection of new energy materials.

Description

Technical Field

[0001] This invention relates to the field of fiber optic sensing technology, specifically an OFDR strain demodulation method based on the nearest neighbor sliding window threshold. Background Technology

[0002] OFDR (Optical Frequency Domain Reflectometry) systems are interferometric fiber optic sensing systems based on narrow-linewidth laser frequency modulation. They are widely used in fields such as civil structure health monitoring, composite material fatigue testing, and temperature detection of new energy materials. During OFDR operation, OFDR demodulation (i.e., demodulating the interference signal generated by the OFDR system) yields the strain / temperature values ​​applied to the fiber, thus achieving distributed strain / temperature measurement. However, in practical applications, traditional OFDR demodulation technology, due to its inherent limitations, struggles to achieve accurate measurements when the strain is large. Specifically, when the strain applied to the fiber is large, the significant stretching of the fiber leads to a substantial decrease in the spectral similarity between the reference signal and the measured signal at corresponding positions. This can easily result in demodulation errors in the spectral offset obtained from cross-correlation calculations (e.g.,...). Figure 2 As shown in the figure, accurate measurement is difficult to achieve. Therefore, it is necessary to invent an OFDR strain demodulation method based on the nearest neighbor sliding window threshold to solve the problem that traditional OFDR demodulation techniques struggle to achieve accurate measurement when the strain is large. Summary of the Invention

[0003] To address the problem that traditional OFDR demodulation technology struggles to achieve accurate measurement when strain is large, this invention provides an OFDR strain demodulation method based on the nearest neighbor sliding window threshold.

[0004] This invention is achieved using the following technical solution:

[0005] An OFDR strain demodulation method based on a nearest-neighbor sliding window threshold is implemented using the following steps:

[0006] Step S1: Obtain the reference signal:

[0007] The OFDR system was run without strain applied to the optical fiber, and the interference signal generated during the operation was acquired twice, thus obtaining two sets of interference signals. The two sets of interference signals were used as two sets of reference signals respectively.

[0008] Step S2: Set the threshold:

[0009] Perform a sliding window cross-correlation operation on the two sets of reference signals to obtain the first set of spectral offsets; calculate the average of the absolute values ​​of each spectral offset in the first set of spectral offsets, and set the calculation result as the threshold.

[0010] Step S3: Acquire measurement signal:

[0011] The OFDR system is operated while strain is applied to the optical fiber, and the interference signal generated during operation is acquired in a single operation to obtain a set of interference signals, which are then used as the measurement signals.

[0012] Step S4: Obtain the two differences:

[0013] Select one set of reference signals from the two sets of reference signals, and perform a sliding window cross-correlation operation on the reference signals and the measured signals to obtain the second set of spectral offsets; select one spectral offset from the second set of spectral offsets, and calculate the difference between this spectral offset and the adjacent previous spectral offset and the adjacent next spectral offset respectively to obtain two difference values;

[0014] Step S5: Compare the two differences with the threshold respectively; if the absolute value of both differences is less than the threshold, proceed to step S6; otherwise, proceed to step S7.

[0015] Step S6: Directly determine the final spectral offset:

[0016] The spectral offset selected in step S4 is directly used as the final spectral offset, and then step S9 is executed.

[0017] Step S7: Obtain a new reference signal:

[0018] Keep the starting index value of the sliding window of the measurement signal in the distance domain unchanged; for the set of reference signals selected in step S4, change the starting index value of the sliding window of the set of reference signals in the distance domain, and step the sliding window of the set of reference signals in the distance domain to obtain a new set of reference signals.

[0019] Step S8: Obtain multiple differences and determine the final spectral offset:

[0020] Perform a sliding window cross-correlation operation on the new set of reference signals and the measurement signals to obtain the third set of spectral offsets; subtract each spectral offset in the third set of spectral offsets from the adjacent previous spectral offset in step S4 to obtain multiple difference values; select the minimum difference value from all the difference values ​​and take the spectral offset value corresponding to the minimum difference value as the final spectral offset value.

[0021] Step S9: Calculate the strain applied to the optical fiber based on the final spectral offset.

[0022] Further, in steps S2, S4, and S8, the specific steps of the sliding window cross-correlation operation are as follows: First, perform a fast Fourier transform on the two sets of signals to convert them from the frequency domain to the distance domain; then, use a sliding window to window and truncate the two sets of signals, and interpolate and pad them with zeros; then, perform an inverse fast Fourier transform on the two sets of signals to convert them from the distance domain to the frequency domain; finally, perform a cross-correlation operation on the two sets of signals to obtain a set of cross-correlation peak offsets, which are a set of spectral offsets.

[0023] Further, the specific steps of step S9 are as follows: multiply the horizontal and vertical coordinates of the final spectral offset by the spatial resolution and frequency resolution of the OFDR system, respectively, thereby demodulating the strain value of the optical fiber over distance.

[0024] Compared to traditional OFDR demodulation techniques, the OFDR strain demodulation method based on a nearest-neighbor sliding window threshold described in this invention employs a novel demodulation principle, ensuring accurate measurement even with large strain. Specifically, even with significant strain applied to the optical fiber, this invention guarantees high similarity between the reference signal and the measured signal's spectra at corresponding positions, thereby ensuring correct demodulation of the spectral offset obtained from cross-correlation calculations (e.g., ...). Figure 3 As shown in the figure, this enables accurate measurement.

[0025] This invention effectively solves the problem that traditional OFDR demodulation technology is difficult to achieve accurate measurement when the strain is large, and is applicable to fields such as health monitoring of civil structures, fatigue testing of composite materials, and temperature detection of new energy materials. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of step S7 in this invention.

[0027] Figure 2 This is a schematic diagram of the demodulation results of traditional OFDR demodulation technology.

[0028] Figure 3 This is a schematic diagram of the demodulation results of the present invention. Detailed Implementation

[0029] An OFDR strain demodulation method based on a nearest-neighbor sliding window threshold is implemented using the following steps:

[0030] Step S1: Obtain the reference signal:

[0031] The OFDR system was run without strain applied to the optical fiber, and the interference signal generated during the operation was acquired twice, thus obtaining two sets of interference signals. The two sets of interference signals were used as two sets of reference signals respectively.

[0032] Step S2: Set the threshold:

[0033] Perform a sliding window cross-correlation operation on the two sets of reference signals to obtain the first set of spectral offsets; calculate the average of the absolute values ​​of each spectral offset in the first set of spectral offsets, and set the calculation result as the threshold.

[0034] Step S3: Acquire measurement signal:

[0035] The OFDR system is operated while strain is applied to the optical fiber, and the interference signal generated during operation is acquired in a single operation to obtain a set of interference signals, which are then used as the measurement signals.

[0036] Step S4: Obtain the two differences:

[0037] Select one set of reference signals from the two sets of reference signals, and perform a sliding window cross-correlation operation on the reference signals and the measured signals to obtain the second set of spectral offsets; select one spectral offset from the second set of spectral offsets, and calculate the difference between this spectral offset and the adjacent previous spectral offset and the adjacent next spectral offset respectively to obtain two difference values;

[0038] Step S5: Compare the two differences with the threshold respectively; if the absolute value of both differences is less than the threshold, proceed to step S6; otherwise, proceed to step S7.

[0039] Step S6: Directly determine the final spectral offset:

[0040] The spectral offset selected in step S4 is directly used as the final spectral offset, and then step S9 is executed.

[0041] Step S7: Obtain a new reference signal:

[0042] Keep the starting index value of the sliding window of the measurement signal in the distance domain unchanged; for the set of reference signals selected in step S4, change the starting index value of the sliding window of the set of reference signals in the distance domain, and step the sliding window of the set of reference signals in the distance domain to obtain a new set of reference signals.

[0043] Step S8: Obtain multiple differences and determine the final spectral offset:

[0044] Perform a sliding window cross-correlation operation on the new set of reference signals and the measurement signals to obtain the third set of spectral offsets; subtract each spectral offset in the third set of spectral offsets from the adjacent previous spectral offset in step S4 to obtain multiple difference values; select the minimum difference value from all the difference values ​​and take the spectral offset value corresponding to the minimum difference value as the final spectral offset value.

[0045] Step S9: Calculate the strain applied to the optical fiber based on the final spectral offset.

[0046] In steps S2, S4, and S8, the specific steps of the sliding window cross-correlation operation are as follows: First, perform a fast Fourier transform on the two sets of signals to convert them from the frequency domain to the distance domain; then, use a sliding window to window and truncate the two sets of signals, and interpolate and pad them with zeros; then, perform an inverse fast Fourier transform on the two sets of signals to convert them from the distance domain to the frequency domain; finally, perform a cross-correlation operation on the two sets of signals to obtain a set of cross-correlation peak offsets, which are a set of spectral offsets.

[0047] The specific steps of step S9 are as follows: multiply the horizontal and vertical coordinates of the final spectral offset by the spatial resolution and frequency resolution of the OFDR system, respectively, and thereby demodulate to obtain the strain value of the optical fiber over distance.

[0048] While specific embodiments of the present invention have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of the present invention is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, but all such changes and modifications fall within the scope of protection of the present invention.

Claims

1. A nearest neighbor sliding window threshold based OFDR strain demodulation method, characterized in that: The method is realized by the following steps: Step S1: obtaining reference signals: The OFDR system is operated when no strain is applied on the optical fiber, and the interference signals generated by the operation are collected twice, thereby obtaining two groups of interference signals, and the two groups of interference signals are taken as two groups of reference signals respectively; Step S2: setting a threshold value: The two groups of reference signals are subjected to sliding window cross-correlation operation, thereby obtaining a first group of frequency spectrum shifts; the average value of the absolute values of each frequency spectrum shift in the first group of frequency spectrum shifts is calculated, and the calculation result is set as the threshold value; Step S3: obtaining a measurement signal: The OFDR system is operated when strain is applied on the optical fiber, and the interference signals generated by the operation are collected once, thereby obtaining a group of interference signals, and the group of interference signals is taken as the measurement signal; Step S4: obtaining two difference values: A group of reference signals is selected from the two groups of reference signals, and the group of reference signals is subjected to sliding window cross-correlation operation with the measurement signal, thereby obtaining a second group of frequency spectrum shifts; a frequency spectrum shift is selected from the second group of frequency spectrum shifts, and the frequency spectrum shift is subtracted from the adjacent previous frequency spectrum shift and the adjacent next frequency spectrum shift respectively, thereby obtaining two difference values; Step S5: comparing the two difference values with the threshold value respectively; if the absolute values of the two difference values are both less than the threshold value, step S6 is executed; otherwise, step S7 is executed; Step S6: directly determining a final frequency spectrum shift: The selected frequency spectrum shift in step S4 is directly taken as the final frequency spectrum shift, and then step S9 is executed; Step S7: obtaining a new reference signal: The starting index value of the sliding window of the measurement signal in the distance domain remains unchanged; for the selected group of reference signals in step S4, the starting index value of the sliding window of the group of reference signals in the distance domain is changed, and the sliding window of the group of reference signals in the distance domain is subjected to step index, thereby obtaining a group of new reference signals; Step S8: obtaining multiple difference values and determining a final frequency spectrum shift: The group of new reference signals is subjected to sliding window cross-correlation operation with the measurement signal, thereby obtaining a third group of frequency spectrum shifts; each frequency spectrum shift in the third group of frequency spectrum shifts is subtracted from the adjacent previous frequency spectrum shift in step S4, thereby obtaining multiple difference values; the minimum difference value is selected from the multiple difference values, and a frequency spectrum shift corresponding to the minimum difference value is taken as the final frequency spectrum shift; Step S9: according to the final frequency spectrum shift, the strain value applied on the optical fiber is calculated.

2. The OFDR strain demodulation method based on nearest neighbor sliding window threshold according to claim 1, characterized in that: In the steps S2, S4 and S8, the specific steps of the sliding window cross-correlation operation are as follows: firstly, the fast Fourier transform is performed on the two groups of signals, so that the two groups of signals are converted from the frequency domain to the distance domain; then, the sliding window is used to cut off the two groups of signals, and the two groups of signals are subjected to interpolation zero padding; then, the inverse fast Fourier transform is performed on the two groups of signals, so that the two groups of signals are converted from the distance domain to the frequency domain; finally, the cross-correlation operation is performed on the two groups of signals, thereby obtaining a group of cross-correlation peak shift values, and the group of cross-correlation peak shift values is a group of frequency spectrum shifts.

3. The OFDR strain demodulation method based on nearest neighbor sliding window thresholding of claim 1, wherein: The specific step of the step S9 is: multiplying the horizontal and vertical coordinates of the final spectrum offset by the spatial resolution and the frequency resolution of the OFDR system respectively, so as to demodulate the strain value of the optical fiber in the distance.

Images