FBG demodulation system based on tunable F-P filter

Abstract

The invention provides an FBG (Fiber Bragg Grating) demodulation system based on a tunable F-P (Fabry-Perot) filter. The system comprises a sensing optical path module, a signal acquisition circuit and an upper computer, wherein the sensing light path module is used for scanning the reflection wavelength of the FBG; the signal acquisition circuit is used for converting an FBG reflected light signal into an electric signal, and performing analog-to-digital conversion and synchronous acquisition on the electric signal; and the upper computer is used for processing the acquired data, demodulating the central wavelength of the FBG, and displaying and storing a demodulation result in real time. The method has a remarkable engineering application value.

Description

Technical Field

[0001] The present invention belongs to the technical field of optical fiber sensing, and in particular relates to an FBG demodulation system based on a tunable FP filter. Background Art

[0002] Fiber Bragg Grating (FBG) sensing technology, due to its advantages such as resistance to electromagnetic interference, corrosion resistance, and ease of distributed measurement, has been widely used in structural health monitoring, earthquake early warning, and industrial vibration detection. The core operating principle of FBG sensors is to reflect changes in external physical quantities by detecting changes in their reflected wavelength. Therefore, high-precision wavelength demodulation technology is key to FBG sensing systems.

[0003] Currently, commonly used FBG demodulation methods include spectrometer detection, matched gratings, edge filtering, unbalanced Mach-Zehnder interferometry, and tunable FP filtering. Among these, demodulation methods based on tunable FP filters (TFPFs) have become a hot topic in research and application due to their high precision, reasonable cost, and wide demodulation range. However, the piezoelectric ceramic (PZT) in TFPFs inherently exhibits creep and hysteresis, resulting in a nonlinear relationship between input voltage and transmission wavelength, severely impacting the detection accuracy of the demodulation system. Furthermore, the stability and repeatability of wavelength detection in traditional demodulation systems, when exposed to complex environmental interference, need to be improved.

[0004] Therefore, how to overcome the nonlinear influence of PZT in TFPF and improve the measurement accuracy, stability and repeatability of FBG demodulation system is a technical problem that needs to be solved urgently in this field. Summary of the Invention

[0005] The present invention aims to provide an FBG demodulation system based on a tunable FP filter to solve the problems of low demodulation accuracy and poor stability caused by PZT hysteresis and creep in the prior art TFPF, and to achieve high-precision and high-stability demodulation of the FBG reflection wavelength.

[0006] A FBG demodulation system based on a tunable FP filter, the system comprising:

[0007] A sensing optical path module, a signal acquisition circuit and a host computer; wherein the sensing optical path module is used to scan the FBG reflection wavelength;

[0008] The signal acquisition circuit is used to convert the FBG reflected light signal into an electrical signal, and perform analog-to-digital conversion and synchronous acquisition on the electrical signal;

[0009] The host computer is used to process the collected data, demodulate the central wavelength of the FBG, and display and store the demodulation results in real time.

[0010] Furthermore, the sensing optical path module includes: a broadband light source, a tunable FP filter, a coupler, a circulator and an FBG sensor group. The broadband light source is used to generate a continuous optical signal covering the FBG operating wavelength range. The tunable FP filter scans the wavelength of the input light under the action of a driving voltage. The coupler and circulator send the scanned optical signal to the FBG sensor group.

[0011] Furthermore, the signal acquisition circuit includes: a photodetector and a data acquisition card, wherein the photodetector converts the FBG reflected light signal into an electrical signal, and the data acquisition card performs analog-to-digital conversion and synchronous acquisition on the electrical signal.

[0012] Furthermore, the sensing optical path module further includes an FP etalon arranged in parallel with the tunable FP filter, for providing a stable wavelength reference standard.

[0013] Furthermore, the host computer uses a multi-peak Gaussian fitting algorithm to fit the transmission peak of the FP etalon, and calculates the time difference between the FBG reflection peak and the FP etalon transmission peak, and combines the interpolation method to demodulate the central wavelength of the FBG.

[0014] Furthermore, the optical path design of the sensing optical path module is as follows: the output light of the broadband light source is filtered by the tunable FP filter and then divided into three paths through the coupler. The first and second paths are connected to the FBG sensor group through the circulator, and the third path is connected to the FP etalon through the circulator. When the transmission wavelength of the tunable FP filter is scanned to match the central wavelength of the FBG sensor, the FBG reflected light returns to the coupler through the circulator and is superimposed with the transmitted light of the FP etalon at the photodetector to form a composite signal.

[0015] Furthermore, the host computer calculates the mean of the wavelength data by a sliding average filtering algorithm to reduce high-frequency noise.

[0016] Furthermore, the demodulation wavelength range of the system is 40 nm, the sampling rate is 300 Hz, the wavelength demodulation repeatability is ±4 pm, and the standard deviation is 1.08 pm.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] The FBG demodulation system based on a tunable FP filter provided by the present invention effectively overcomes the effects of hysteresis and creep of piezoelectric ceramics in the tunable FP filter by introducing an FP etalon as a wavelength reference in the sensing optical path module, combined with a multi-peak Gaussian fitting algorithm and a sliding average filtering algorithm, and significantly improves the wavelength demodulation accuracy. The optical path design adopts a three-way splitting structure, sending scanning light into the FBG sensor group and the FP etalon respectively, superimposing a composite signal through photodetectors, and combining the wavelength difference with the interpolation method to achieve high-precision demodulation of the FBG center wavelength. The system demodulation wavelength range reaches 40nm, the sampling rate is 300Hz, the repeatability is ±4pm, and the standard deviation is 1.08pm. It has the advantages of high precision, high stability and wide range. The host computer developed by LabVIEW realizes the integrated operation of data processing, real-time display and storage. It has significant engineering application value in low-frequency vibration monitoring in complex environments such as structural health monitoring and earthquake early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings illustrate various embodiments generally by way of example and not limitation, and together with the description and claims, serve to explain embodiments of the invention. Where appropriate, the same reference numerals are used throughout the drawings to refer to the same or similar parts. Such embodiments are illustrative and are not intended to be exhaustive or exclusive of the embodiments of the present apparatus or method.

[0020] Figure 1 shows a schematic diagram of the optical path of the demodulation system of the present invention;

[0021] Figure 2 Shows a schematic diagram of channel allocation of a data acquisition card of the present invention;

[0022] Figure 3 A schematic diagram comparing the three algorithm fitting algorithms of the present invention is shown;

[0023] Figure 4 Shown is a wavelength and voltage fitting schematic diagram of the present invention;

[0024] Figure 5 A schematic diagram of actual data acquisition by the FBG and FP etalons of the present invention is shown. DETAILED DESCRIPTION

[0025] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0026] This paper addresses the inherent creep and hysteresis of the PZT in the TFPF, which results in a nonlinear relationship between the input voltage and the transmission wavelength, reducing the detection accuracy of the demodulation system. This paper proposes a FBG demodulation system based on a combination of the TFPF and the F-standard. First, the principles of the demodulation system and the optical path components are introduced. Second, host computer software based on LabVIEW is developed to implement system control and data processing. Finally, the demodulation system is constructed and performance tested.

[0027] Overall design of FBG demodulation system

[0028] The basic principle of the TFPF-based FBG sensor demodulation system is to scan the wavelength of the sensor FBG through the TFPF, obtain the corresponding relationship between the filter's transmission wavelength and the scanning voltage, and then demodulate the center wavelength of the sensor FBG. By combining the TFPF with a broadband light source, high-precision scanning of the sensor FBG wavelength can be achieved, thereby realizing a wide range of spectral scanning.

[0029] The demodulation system consists of a sensing optical path, a signal acquisition circuit, and a host computer. The sensing optical path comprises a broadband light source, a TFPF, a coupler, a circulator, and an FBG sensor group. The signal acquisition circuit comprises a photodetector and a data acquisition card. The host computer software, developed using LabVIEW, is used for signal processing, analysis, and result display.

[0030] The working principle of the demodulation system is that the power supply drives the broadband light source to generate continuous signal light that enters the FP filter. At the same time, the sawtooth signal generator generates a periodic voltage signal to drive the PZT inside the TFPF, so that the transmission wavelength of the filter is periodically scanned within a certain spectral range. The filtered signal light enters the 1×3 coupler and is divided into three paths, one of which enters the FBG sensor 1 through the circulator 1, the other enters the FBG sensor 2 through the circulator 2, and the third light enters the FP standard through the circulator. When the wavelength scanned by the filter matches the central wavelength of the FBG, the light of this wavelength is reflected back to the circulator by the FBG and enters the photodetector. The FP standard transmits signal light of a specific wavelength, enters the photodetector and couples with the light signal reflected by the FBG. The two are superimposed by the coupler to form a composite signal. After the photodetector performs photoelectric conversion on the composite light signal, it is sent to the data acquisition card for digital signal processing. Figure 1 As shown,

[0031] In the data acquisition card, after processing by algorithms such as filtering, Gaussian fitting, and peak finding, calculations will be performed in LabVIEW. By comparing the corresponding time difference between the FBG reflection peak and the etalon transmission peak during the scanning process, the current center wavelength of the FBG sensor is determined, thereby achieving accurate demodulation of the measured physical quantity.

[0032] Demodulation system optical path design

[0033] Broadband light source

[0034] The broadband light source is a core component in the sensing optical path. Its key performance indicators are spectral range and output optical power. Because the FBG sensor operates in the C-band, the selected broadband light source must cover the wavelength range of 1528-1568nm. Considering the significant light intensity attenuation in the optical path, a high output optical power was selected. A broadband light source module from DenseLight was selected, with a maximum output power of 25mW.

[0035] Signal Generator

[0036] In this demodulation system, the signal generator is primarily used to modulate the TFPF, generating a stable sweep drive signal. Signal generators typically output sawtooth or similar waveforms. Their operating principle is to use voltage control to create a periodic, linear rise and fall pattern in the output signal. Because triangular wave signals offer linear and uniform variation, stable frequency, and precisely controllable amplitude, they are widely used as a standard sweep control signal in demodulation systems. This ensures the accuracy and repeatability of the FP filter's sweep wavelength, effectively improving the measurement accuracy and stability of the demodulation system.

[0037] Tunable FP filter

[0038] The demodulation system uses a SiCART Tunable Filter. The filter primarily consists of a tunable FP cavity, with fiber optic cables extending from each end of the FP filter to connect to the optical system. Pins are located on the underside of the cavity for inputting the modulation voltage. The filter's operating range is 1500-1600 nm, meeting the operating range of the FBG sensor; the Full Width at Half Maximum (FWHM) is 0.169 nm. According to sensor demodulation theory, the closer the FWHM is to the 3dB bandwidth of the FBG, the higher the wavelength detection sensitivity and the better the demodulation performance.

[0039] FP etalon

[0040] PZT is susceptible to hysteresis, creep effects, and environmental factors, resulting in a nonlinear relationship between the modulation voltage and cavity length. To minimize this impact on demodulation accuracy, a FP etalon was introduced as the wavelength calibration reference for the demodulation system, thereby improving the system's accuracy and stability. The FP etalon, manufactured by Primanex, was selected as the wavelength calibration reference for the demodulation system. Its performance specifications include a wavelength range of 1510–1590 nm, an FSR of 100 GHz, and a temperature-compensated wavelength drift of less than 10 pm between -20°C and 70°C. Based on the demodulation theory in Chapter 2, the selected etalon has a narrow 3dB bandwidth, which effectively improves the system's wavelength detection accuracy.

[0041] Photodetectors

[0042] Photodetectors convert optical signals into electrical signals for subsequent data acquisition and analysis by the data acquisition card. A multi-channel PIN optical detection module, manufactured by Beijing Keyang Optoelectronics Technology Co., Ltd., was selected. This module offers high stability and reliability, and can detect wavelengths from 800 to 1700 nm, fully covering the operating band of this system and meeting the test requirements of the designed FBG sensor demodulation system. Its specific performance indicators are shown in Table 1.

[0043] Table 1 Photodetector performance indicators

[0044] Wavelength range Peak responsivity Conversion gain@1550nm 800~1700nm 0.9A / w@1550 4×104V / W

[0045] Data acquisition card

[0046] After the spectral signal is converted into an electrical signal by a photodetector, it is converted to digital by a data acquisition card. The designed FBG sensor demodulation system uses a PCI-9215 synchronous data acquisition card manufactured by National Instruments (NI). This card features 8 channels and 16-bit sampling, enabling simultaneous acquisition of signals from multiple channels and ensuring consistency and accuracy between the reference and measurement channel data.

[0047] Connect the reference signal of the FP etalon to the AI0 channel, the reflectance spectrum signal of the FBG sensor to the AI1 channel, the output signal of the signal generator to the AI2 channel, and the synchronous square wave signal to the AI7 channel. This signal serves as the synchronous trigger source for data acquisition, thereby improving the overall synchronization and stability of the demodulation system.

[0048] The demodulation system host computer software designed in the present invention is jointly developed and implemented based on LabVIEW 2024 and Matlab. Among them, LabVIEW is a graphical programming environment widely used in the development of data measurement, analysis and monitoring systems.

[0049] Fitting peak finding algorithm

[0050] Centroid method

[0051] The centroid algorithm determines the feature location through the signal energy distribution. In applications, the horizontal coordinate and power value of each point are usually used to calculate the position of the signal center. This method uses the power value of each point as a weight and combines it with the sampling point position to finally obtain a more accurate signal center position. The centroid method is simple to calculate and is widely used in the field of signal processing. The formula (1) of the centroid method is as follows

[0052]

[0053] Cubic spline interpolation method

[0054] To avoid the computational complexity and instability associated with high-order interpolation, researchers have proposed a piecewise low-order interpolation method. This method first divides the data interval into several smaller regions, and then performs low-order interpolation within each region. This method reduces the number of interpolation operations and computational complexity. Currently, the cubic spline interpolation function is more commonly used, but it relies on a key condition: it is not suitable for fitting large deflection curves, especially at locations with significant changes, such as peak tops. This limitation may affect the algorithm's peak-finding accuracy.

[0055] Gaussian fitting method

[0056] The optical power function obtained by convolving the FBG reflection function and the TFPF transmission function still has a Gaussian distribution. Therefore, a Gaussian fitting algorithm is used for curve fitting. In the fitting process, a cubic spline interpolation function is used, but this method has difficulty fitting the large deflection curve at the top of the peak, which affects the accuracy of peak finding. To solve this problem, a Gaussian fitting algorithm is introduced, which uses the symmetry of the Gaussian function to accurately fit the peak area of ​​the curve. The specific expression of the Gaussian fitting algorithm is as follows

[0057] y=Aexp(-(xu) 2 / 2σ 2 ) (2)

[0058] The three algorithms are fitted and compared on the collected original data. Figure 3 As shown in the figure, the cubic spline interpolation method's fitting results are basically consistent with the original data, but when there is a lot of noise, there are still fitting errors in the peak area. The sliding average method's fitting results can effectively smooth the data and reduce noise, but due to the high noise content in the engineering data, the peak position is offset, affecting peak detection accuracy. The Gaussian fitting results can accurately capture the peak, and the peak position is consistent with the original data, significantly improving demodulation accuracy, especially in the case of strong noise.

[0059] Demodulation system test analysis

[0060] During performance testing, a spectrometer was used to calibrate the relationship between the TFPF's transmission wavelength and the driving voltage. During calibration, the broadband light source power was set to 25mW, and the signal generator output a 10V DC voltage in increments of 2V. The spectrometer recorded the corresponding TFPF transmission wavelength position. Calculation revealed that the relationship between wavelength and voltage is y = -4.64x + 1566.25.

[0061] Table 2 shows that the central wavelength fluctuates between 1549.994nm and 1550.272nm. The FBG central wavelength fluctuation error is greater than 300pm, and the fluctuation range is 0.212nm. Using the standard deviation formula, the standard deviation of the data is 93.5pm.

[0062] Table 2 Actual test data of wavelength and voltage fitting demodulation

[0063] Demodulation times Wavelength (nm) Error (pm) 1 1550.029 52 2 1550.272 106 3 1550.189 212 4 1549.994 17 5 1550.022 45 6 1550.190 212 7 1550.083 45 8 1550.166 17 9 1550.174 19 10 1549.996 52

[0064] The FBG and FP etalons actually collect data, such as Figure 5 As shown in the figure, t1 is the time point corresponding to the first peak on the right side of the etalon calibration peak, Δt is the time point difference between two adjacent peaks of the etalon, and t G is the time point corresponding to the FBG central wavelength.

[0065] The FBG center wavelength λ can be calculated by formula (3): b .

[0066]

[0067] Where λ b is the center wavelength of the demodulated FBG, λ1 is the center wavelength of the first peak of the etalon, and Δλ is the center wavelength difference between two adjacent peaks.

[0068] Table 3 Experimental results data

[0069]

[0070]

[0071] The data was processed by the host computer software and the demodulated FBG center wavelength was 1549.977 nm. The system was used to collect data 10 times and the data was processed by the host computer software. The demodulation results and their errors are shown in Table 3.

[0072] The table shows that the maximum deviation of the FBG center wavelength measured by the demodulation system reached 50 pm. The center wavelength fluctuated between 1549.927 nm and 1550.027 nm, with a fluctuation range of 0.1 nm. Using the standard deviation formula, the standard deviation of this data set was calculated to be 15.56 pm.

[0073] In the actual demodulation process, the peak position of the FP etalon will drift slightly due to the influence of external environmental factors. If only the fixed peak of the etalon is used as a reference, measurement errors may be introduced, thereby reducing the demodulation accuracy of the FBG central wavelength. In order to reduce this error and improve measurement accuracy, this study uses a multi-peak Gaussian fitting method to fit and analyze the peaks of the FP etalon. As shown in Table 4, the peak t of the etalon and subsequent peaks are obtained by the Gaussian fitting peak finding algorithm. Define the parameters is the average time interval between two adjacent peaks of the etalon, then formula (3) can be further modified to

[0074]

[0075] The 10 experimental data were calculated using formula (4), and the results and corresponding errors are shown in Table 4.

[0076] Table 4. Experimental results data

[0077]

[0078]

[0079] As shown in Table 4, the central wavelength fluctuates between 1549.957nm and 1549.997nm, which is relatively close to the actual wavelength of 1549.977nm. The maximum deviation is 20pm, the fluctuation range is reduced to 0.04nm, and the standard deviation is 6.37pm. Compared with the data calculated by taking only a specific peak, the standard deviation is reduced by nearly 59.07%.

[0080] To further improve demodulation accuracy, a sliding average is used to calculate the mean of consecutive sampling points, reducing high-frequency noise and short-term fluctuations, thereby improving signal smoothness and further improving demodulation accuracy. The results and corresponding errors are shown in Table 5.

[0081] Table 5 Experimental results data

[0082] Demodulation times Wavelength (nm) Error (pm) 1 1549.976 1 2 1549.979 2 3 1549.981 4 4 1549.979 2 5 1549.980 3 6 1549.976 1 7 1549.973 4 8 1549.975 2 9 1549.978 1 10 1549.979 2

[0083] As shown in Table 5, the central wavelength fluctuates between 1549.970nm and 1549.985nm, which is relatively close to the actual wavelength of 1549.977nm. The maximum deviation is 4pm, the fluctuation range is reduced to 8mm, and the standard deviation is 1.08pm, which further improves the demodulation accuracy.

[0084] This paper primarily designs and studies a TFPF-based FBG sensor demodulation system. First, the key optical path components and their performance in the demodulation system are introduced. Second, a host computer platform for the demodulation system is constructed using LabVIEW. The effects of the centroid algorithm, continuous fitting, and Gaussian fitting on the demodulation system's performance are analyzed. Finally, a complete demodulation test system is constructed to evaluate the impact of different demodulation algorithms on system performance. Experimental results demonstrate a detection range of 40 nm, a demodulator repeatability of ±4 pm, a sampling rate of 300 Hz, and a standard deviation of 1.08 pm, fully demonstrating the superior measurement accuracy of the designed sensor demodulation system.

[0085] The above is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, can make equivalent replacements or changes based on the technical solutions and inventive concepts of the present invention, which should be covered by the scope of protection of the present invention.

Claims

1. A FBG demodulation system based on a tunable FP filter, characterized in that: The system comprises: A sensing optical path module, a signal acquisition circuit and a host computer; wherein the sensing optical path module is used to scan the FBG reflection wavelength; The signal acquisition circuit is used to convert the FBG reflected light signal into an electrical signal, and perform analog-to-digital conversion and synchronous acquisition on the electrical signal; The host computer is used to process the collected data, demodulate the central wavelength of the FBG, and display and store the demodulation results in real time.

2. The FBG demodulation system based on the tunable FP filter according to claim 1, characterized in that: The sensing optical path module includes: a broadband light source, a tunable FP filter, a coupler, a circulator and an FBG sensor group. The broadband light source is used to generate a continuous optical signal covering the FBG operating wavelength range. The tunable FP filter scans the wavelength of the input light under the action of a driving voltage. The coupler and circulator send the scanned optical signal to the FBG sensor group.

3. The FBG demodulation system based on a tunable FP filter according to claim 1, characterized in that: The signal acquisition circuit includes: a photoelectric detector and a data acquisition card, wherein the photoelectric detector converts the FBG reflected light signal into an electrical signal, and the data acquisition card performs analog-to-digital conversion and synchronous acquisition on the electrical signal.

4. The FBG demodulation system based on a tunable FP filter according to claim 1, characterized in that: The sensing optical path module further includes an FP etalon arranged in parallel with the tunable FP filter, for providing a stable wavelength reference standard.

5. The FBG demodulation system based on a tunable FP filter according to claim 1, characterized in that: The host computer uses a multi-peak Gaussian fitting algorithm to fit the transmission peak of the FP etalon, and calculates the time difference between the FBG reflection peak and the FP etalon transmission peak, and combines the interpolation method to demodulate the central wavelength of the FBG.

6. The FBG demodulation system based on a tunable FP filter according to claim 1, characterized in that: The optical path design of the sensing optical path module is as follows: the output light of the broadband light source is filtered by a tunable FP filter and then divided into three paths through a coupler. The first and second paths are connected to the FBG sensor group through a circulator, and the third path is connected to the FP etalon through a circulator. When the transmission wavelength of the tunable FP filter is scanned to match the center wavelength of the FBG sensor, the FBG reflected light returns to the coupler through the circulator and is superimposed with the transmitted light of the FP etalon at the photodetector to form a composite signal.

7. The FBG demodulation system based on a tunable FP filter according to claim 1, characterized in that: The host computer calculates the mean of the wavelength data by a sliding average filtering algorithm to reduce high-frequency noise.

8. The FBG demodulation system based on a tunable FP filter according to claim 1, characterized in that: The demodulation wavelength range of the system is 40 nm, the sampling rate is 300 Hz, the wavelength demodulation repeatability is ±4 pm, and the standard deviation is 1.08 pm.

Images