阅读说明：本技术 一种fpi腔长解调方法及系统 (FPI cavity length demodulation method and system ) 是由 桂吟秋 鲁平 张津 刘德明 于 2019-09-18 设计创作，主要内容包括：本发明公开一种FPI腔长解调方法及系统,涉及一种波数光谱主频精确识别算法。该算法首先对FPI的反射光谱进行波数变换和插值,再对波数光谱进行FFT,通过Rife算法估算波数光谱频率,即通过FFT的最大谱线以及相邻次级大谱线估算出真实频率,根据该频率即可得到FPI的腔长,由腔长的变化可确定周围环境因素的变化,如温度等。该算法的计算量很小,且估算精度很高。通过该方法进行解调的FPI温度传感器测量范围只受传感器本身特性的限制,大大提高了动态范围。该方法还可用于FPI的并联复用的解调。(The invention discloses a method and a system for demodulating FPI cavity length, and relates to an algorithm for accurately identifying dominant frequency of wave number spectrum. The algorithm firstly carries out wave number transformation and interpolation on the reflection spectrum of the FPI, then carries out FFT on the wave number spectrum, estimates the frequency of the wave number spectrum through the Rife algorithm, namely estimates the real frequency through the maximum spectral line of the FFT and the adjacent secondary large spectral line, obtains the cavity length of the FPI according to the frequency, and can determine the change of the surrounding environment factors such as temperature and the like through the change of the cavity length. The algorithm has small calculation amount and high estimation precision. The measurement range of the FPI temperature sensor demodulated by the method is only limited by the characteristics of the sensor, so that the dynamic range is greatly improved. The method can also be used for demodulation of parallel multiplexing of FPIs.)

1. An FPI cavity length demodulation method is characterized by comprising the following steps:

collecting a reflection spectrum of the FPI, and performing wave number conversion on the spectrum;

interpolating the wave number data after the wave number transformation to obtain discrete wave number data with uniform intervals;

determining the wave number spectrum frequency by using a Rife algorithm based on the discrete wave number data with uniform intervals;

determining a cavity length of the FPI based on the wavenumber spectral frequencies; the change in the ambient preset environmental factor can be determined from the change in the cavity length.

2. The FPI cavity length demodulation method of claim 1, wherein said wavenumber transform of said spectrum comprises the steps of:

firstly, determining the relation between the FPI reflectivity and the spectrum wavelength;

secondly, determining the relation between the spectral wavelength and the wave number;

finally, the relationship between the FPI reflectivity and the wave number is determined.

3. The FPI cavity length demodulation method of claim 2, wherein cubic spline interpolation is performed on the wave number based on the relationship between the FPI reflectivity and the wave number, resulting in discrete wave number data with uniform spacing.

4. The FPI cavity length demodulation method according to claim 3, wherein said determining wavenumber spectral frequencies using a Rife algorithm comprises the steps of:

performing fast Fourier transform on the discrete wave number data with uniform intervals;

the wavenumber spectral frequencies are determined based on the discrete wavenumber data after the fast fourier transform.

5. The FPI cavity length demodulation method according to any of claims 1-4, wherein the relationship of wavenumber spectral frequency to FPI cavity length is expressed by the following formula:

where f represents the wavenumber spectral frequency, ω represents the spectral angular frequency, and L represents the FPI cavity length.

6. An FPI cavity length demodulation system, comprising:

the spectrum acquisition unit is used for acquiring a reflection spectrum of the FPI and performing wave number conversion on the spectrum;

the wave number interpolation unit is used for interpolating the wave number data after the wave number transformation to obtain discrete wave number data with uniform intervals;

the spectral frequency determining unit is used for determining the wave number spectral frequency by using a Rife algorithm based on the discrete wave number data with uniform intervals;

and the FPI cavity length determining unit is used for determining the cavity length of the FPI based on the wave number spectrum frequency, and the change of the surrounding preset environment factors can be determined according to the change of the cavity length.

7. The FPI cavity length demodulation system of claim 6, wherein the spectrum acquisition unit first determines the relationship between FPI reflectivity and spectral wavelength; secondly, determining the relation between the spectral wavelength and the wave number; finally, the relationship between the FPI reflectivity and the wave number is determined.

8. The FPI cavity length demodulation system of claim 7, wherein the wave number interpolation unit performs cubic spline interpolation on the wave number based on the relationship between the FPI reflectivity and the wave number, resulting in discrete wave number data with uniform spacing.

9. The FPI cavity length demodulation system of claim 8, wherein said spectral frequency determination unit performs fast fourier transform on said uniformly spaced discrete wavenumber data; the wavenumber spectral frequencies are determined based on the discrete wavenumber data after the fast fourier transform.

10. The FPI cavity length demodulation system according to any of claims 6 to 9, wherein the relationship between wavenumber spectral frequency and FPI cavity length is expressed as the following formula:

where f represents the wavenumber spectral frequency, ω represents the spectral angular frequency, and L represents the FPI cavity length.

Technical Field

The invention relates to the technical field of optical fiber sensing, in particular to a method and a system for demodulating FPI cavity length.

Background

One significant advantage of fiber optic sensors over sensing electrical sensors is the ease of multiplexing. The multiplexing of the sensor means that multi-point or even fully distributed parameter measurement can be obtained simultaneously through a set of light source and demodulation equipment, for example, a temperature distribution map in a measurement area can be obtained through fully distributed temperature sensing. The reusability of the optical fiber sensor provides possibility for distributed sensing, the cost of a light source and demodulation equipment can be greatly reduced, and the optical fiber sensor has important significance in engineering application.

The optical fiber sensor based on Fabry-Perot interferometer (FPI) type can be flexibly prepared, so that the optical fiber sensor is widely researched by students, the preparation is generally simple, when the reflectivity of the reflecting surface of an FP cavity is low, the formed interference spectrum is similar to double-beam interference, the spectrum shape is simple, the frequency component is single, the demodulation is facilitated, at present, the phase change is demodulated through the resonant wavelength drift of the interference spectrum, and the frequency demodulation, the single longitudinal mode laser wavelength scanning demodulation, the double-wavelength demodulation, the white light interference demodulation and the like are realized through fast Fourier transform. Since the dynamic Range of a wavelength demodulation based FPI type sensor is limited by Free Spectral Range (FSR), sensitivity can be improved by replacing the more sensitive medium, but the dynamic Range is sacrificed.

In the aspect of spectrum demodulation, researchers do much work in order to obtain accurate real frequency and improve the measurement range. For frequency identification of a single-frequency function, a Maximum Likelihood Estimation algorithm (MLE) is an optimal frequency Estimation algorithm, and later learners improve the defect of the algorithm of ultra-large complexity, including a classical Kay algorithm, an L & R algorithm and an M & M algorithm. However, the frequency estimation algorithm based on the MLE is still high in complexity, slow in synchronization speed and not suitable for real-time processing.

Disclosure of Invention

Aiming at the defects of the prior art, the invention aims to solve the technical problems of small dynamic range, high accuracy, high algorithm complexity and low speed of the conventional wavelength demodulation FPI type sensor.

To achieve the above object, in a first aspect, the present invention provides an FPI cavity length demodulation method, including the following steps:

collecting a reflection spectrum of the FPI, and performing wave number conversion on the spectrum;

interpolating the wave number data after the wave number transformation to obtain discrete wave number data with uniform intervals;

determining the wave number spectrum frequency by using a Rife algorithm based on the discrete wave number data with uniform intervals;

and determining the cavity length of the FPI based on the wave number spectrum frequency, and determining the change of the surrounding preset environment factors according to the change of the cavity length.

The preset environmental factor may be temperature.

Optionally, performing wavenumber transformation on the spectrum, specifically including the following steps:

firstly, determining the relation between the FPI reflectivity and the spectrum wavelength;

secondly, determining the relation between the spectral wavelength and the wave number;

finally, the relationship between the FPI reflectivity and the wave number is determined.

Optionally, cubic spline interpolation is performed on the wave number based on the relationship between the FPI reflectivity and the wave number to obtain discrete wave number data with uniform intervals.

Optionally, the determining the wave number spectrum frequency by using the Rife algorithm specifically includes the following steps:

performing fast Fourier transform on the discrete wave number data with uniform intervals;

the wavenumber spectral frequencies are determined based on the discrete wavenumber data after the fast fourier transform.

Alternatively, the relationship of the wavenumber spectral frequency to the FPI cavity length is expressed as follows:

where f represents the wavenumber spectral frequency, ω represents the spectral angular frequency, and L represents the FPI cavity length.

In a second aspect, the present invention provides an FPI cavity length demodulation system, comprising:

the spectrum acquisition unit is used for acquiring a reflection spectrum of the FPI and performing wave number conversion on the spectrum;

the wave number interpolation unit is used for interpolating the wave number data after the wave number transformation to obtain discrete wave number data with uniform intervals;

the spectral frequency determining unit is used for determining the wave number spectral frequency by using a Rife algorithm based on the discrete wave number data with uniform intervals;

and the FPI cavity length determining unit is used for determining the cavity length of the FPI based on the wave number spectrum frequency, and the change of the surrounding preset environment factors can be determined according to the change of the cavity length.

Optionally, the spectrum acquisition unit first determines a relationship between the FPI reflectivity and the spectral wavelength; secondly, determining the relation between the spectral wavelength and the wave number; finally, the relationship between the FPI reflectivity and the wave number is determined.

Optionally, the wave number interpolation unit performs cubic spline interpolation on the wave number based on the relationship between the FPI reflectivity and the wave number to obtain discrete wave number data with uniform intervals.

Optionally, the spectral frequency determination unit performs fast fourier transform on the uniformly spaced discrete wave number data; the wavenumber spectral frequencies are determined based on the discrete wavenumber data after the fast fourier transform.

Alternatively, the relationship of the wavenumber spectral frequency to the FPI cavity length is expressed as follows:

where f represents the wavenumber spectral frequency, ω represents the spectral angular frequency, and L represents the FPI cavity length.

Generally, compared with the prior art, the above technical solution conceived by the present invention has the following beneficial effects:

(1) the method can demodulate the change of the external environment temperature based on the change of the FPI cavity length, interpolate the data after the wave number conversion is carried out on the spectrum to obtain uniform data, and on the basis, the wave number frequency is solved based on the Rife algorithm.

(2) The invention adopts a spectrum frequency demodulation algorithm, is not limited by spectrum FSR, and can greatly improve the dynamic range of the spectrum FSR.

(3) The Rife algorithm adopted by the invention only utilizes the maximum value of FFT discrete data and the adjacent secondary large value thereof when estimating the main frequency of the FPI reflection spectrum, so that the Rife algorithm is also suitable for estimating the main frequency of the reflection spectrum of a plurality of sensing units simultaneously, and feasibility is provided for multiplexing of the FPI sensor.

Drawings

Fig. 1 is a schematic flow chart of a method for accurately identifying dominant frequencies of a wave number spectrum provided by the invention.

Fig. 2 is a main flow of the Rife algorithm provided by the present invention.

Fig. 3 is a diagram of a temperature sensing experimental apparatus provided by the present invention.

FIG. 4 is a graph of the relationship between the estimated frequency of spectral wavelength demodulation and temperature provided by the present invention.

FIG. 5 is a diagram showing the relationship between the estimated frequency and the temperature of the wave number spectrum dominant frequency accurate identification algorithm provided by the present invention.

Fig. 6 is a schematic diagram of a parallel structure of four FPI sensors according to the present invention.

FIG. 7 is a reflectance spectrum of each FPI in a simulation of four FPI cascades provided by the present invention.

FIG. 8 shows the reflection spectra after superposition in four FPI cascade simulations provided by the present invention.

FIG. 9 shows reflection spectra after wavenumber transformation and FFT processing in four FPI cascade simulations provided by the present invention.

FIG. 10 is a light path diagram of two FPI cascade structures provided by the present invention.

FIG. 11 is the relationship between the dominant frequency of two FPI cascade structures and the temperature of 20 deg.C-30 deg.C provided by the present invention.

FIG. 12 is a graph showing the relationship between the dominant frequency of two FPI cascade structures and the temperature of 20 deg.C-60 deg.C.

FIG. 13 is a block diagram of an FPI cavity length demodulation system according to the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

Aiming at the problems in the prior art, the invention aims to provide a new demodulation idea to improve the dynamic range of the FPI sensor and verify the feasibility in the aspect of FPI sensor multiplexing.

In order to achieve the purpose, the invention provides an FPI cavity length demodulation method, namely a wave number spectrum dominant frequency accurate identification algorithm.

Compared with the MLE algorithm with high complexity, the operation amount of the Rife algorithm is much simpler, the Rife algorithm is calculated only through Fast Fourier Transform (FFT), the true frequency is estimated by utilizing the maximum spectral line and the second-largest spectral line, the complexity of the algorithm is not high, the hardware implementation is easy, and the method is suitable for real-time calculation.

Specifically, the spectral data is processed using a wavenumber transform.

Specifically, the smoothest interpolation method is used: and (3) performing cubic spline (spline) interpolation, and processing the wave number-intensity data to obtain discrete data with uniform intervals.

Specifically, a Rife algorithm is adopted to accurately estimate the true frequency of the spectrum.

In particular, the algorithm improves the dynamic range of the FPI sensor.

Specifically, the algorithm can be applied to parallel multiplexing of FPI sensing units, so that crosstalk-free temperature sensing is realized, and feasibility for realizing multiplexing of FPI sensors is provided.

Fig. 1 shows the main flow of the method for accurately identifying dominant frequencies of a wave number spectrum described in this embodiment. After the reflection spectrum of the FPI is collected, it is first necessary to perform wave number transform on the spectrum, i.e. mapping all wavelengths in the data to wave numbers one by one, and then the relationship between wave number and intensity is a uniform periodic function. Since the reflection spectrum shape of the low-finesse FPI is similar to a cosine function, the reflectance expression is as shown in formula (1), and the reflectance R is expressed as a relation with the wavelength, that is:

wherein, A is F/2, B is F/2. F represents the fineness coefficient of the spectrum and L represents the cavity length of the FPI. However, in the formula (1), the wavelength λ is in the denominator, and the spectrum is considered to have periodicity only in a window with a small wavelength range. In practice, the reflectivity is not an exact periodic function of wavelength. By the following transformations:

ω＝2nL (3)

k is defined in theoretical physics as the wavenumber (wavenumber). Then, equation (1) can be written as:

R＝A+Bcos(ω·k) (4)

the reflectance R represented by the formula (4) has an exact periodic function relationship with the wave number k, and a ═ F/2, B ═ F/2, and ω represent the direct current amount, amplitude, and angular frequency of the spectrum, respectively, and the above transformation is referred to as wave number transformation.

However, the transformed wave number data has non-uniform intervals, and subsequent FFT operation cannot be performed, so that the second step needs to perform interpolation to obtain discrete data with uniform intervals. And (3) processing the wave number-intensity data by adopting a smoothest interpolation method, namely cubic spline (spline) interpolation.

After the spectral data processing is completed, the Rife algorithm is used to estimate the wavenumber spectral frequencies, as shown in fig. 2Shown as the implementation flow of the Rife algorithm. For convenient analysis, a finite-length sine wave signal S (t) is introducedAnalytic form f (t):

wherein a represents amplitude, t represents time, f₀The frequency is represented by a frequency-dependent variable,indicating the phase.

This signal is sampled with a period Δ t, resulting in:

the number of sampling points N satisfies N · Δ T ═ T, then { f_kN-1 is a discrete sampling sequence of f (t). FFT of the sequence yields:

wherein m is 0,1,2, …, N-1,

v_mrepresenting the angular frequency, modulo the discrete sequence f (m),

combining the value range of m and the formula (8), it can be deduced that:

from the basic higher mathematical knowledge, if and only if v_mΔ t/2 → 0, | F (m) | takes the maximum value. Since m can only take a discrete series of integers, it is assumed that when m is m₀When, | f (m) | takes the maximum value. Defining the frequency estimation deviation Δ f as the difference between the corresponding frequency of the FFT maximum and the actual frequency, that is:

therefore, | δ | < 1.δ represents the distance between the frequency sampling point and the actual frequency point.

If the actual frequency f of the signal₀Greater than the estimated frequency m/N Δ t, then:

-1＜δ＜0 (12)

the maximum spectral line can be expressed as:

the adjacent large spectral lines of the stage are:

if the actual frequency f of the signal₀Less than the estimated frequency m/N Δ t, then:

0＜δ＜1 (15)

the maximum spectral line expression is unchanged, while the adjacent secondary large spectral lines are:

equations (14) and (16) may be written in combination as:

r ═ 1, which represents the difference of the equation when the estimated frequency is greater or less than the actual frequency, combining equation (13) and equation (17), the ratio of the adjacent secondary large spectral line to the maximum spectral line is:

an estimate of delta can be obtained

Thus, the final frequency estimate is:

for an FPI with a cavity length L and a cavity medium refractive index n, the wavenumber spectral frequency is as follows:

therefore, theoretically, the actual frequency of the wavenumber spectrum at an FPI with a cavity length of 80 μm is about 2.546479 × 10^-5However, since the collected spectrum frequency changes with the temperature, the cavity length derived from the above formula will be different from the actual value, and thus the change of the cavity length reflects the change of the external environment temperature.

Specifically, after the wave number spectrum frequency is solved, the corresponding cavity length L can be solved and deduced, the cavity length L is compared with the initial cavity length to obtain a cavity length change value, and the corresponding actual temperature is determined according to a pre-tested cavity length change and temperature change relation.

The data interval after the wave number transformation is not uniform, the subsequent FFT operation cannot be carried out, the uniform data can be obtained only by interpolation, the smoothest interpolation method, namely cubic spline (spline) interpolation, is adopted to process the wave number-intensity data to obtain a wave number spectrum with the length of 2048, the operation of the cubic spline interpolation method is simple and can be realized by a computer, and the convergence is goodSmooth and stable, and the frequency finally estimated by the Rife algorithm is about 2.546438 × 10^-5The error between the two is only-0.00162%.

Table (1) lists FPI wavenumber spectra for different cavity lengths (80 μm to 90 μm) and estimates the percent error of frequency by the 2048 data points and 4096 data points Rife algorithm. Specifically, the data points are data points obtained by interpolating the wave number data. As can be seen from the table, the 2048-point Rife algorithm is applied to the frequency estimation of the FPI, and the percentage error does not exceed 0.09%; the percentage error of the 4096-point Rife algorithm is not more than 0.007%, and the precision is improved by one order of magnitude, so that the optimization of the cubic spline interpolation method can improve the precision of the Rife algorithm.

TABLE 1 estimated error Table for the Rife Algorithm

In order to better show the advantages of the embodiment, the wave number spectrum dominant frequency precise identification algorithm is applied to temperature sensing, namely FPI is applied to the technical field of temperature sensors, and after the dominant frequency of the acquired spectrum is solved, the corresponding cavity length change is obtained to solve the environment temperature.

Fig. 3 shows a diagram of a temperature sensing experimental apparatus, which includes an Amplified Spontaneous Emission (ASE), an Optical Spectrum Analyzer (OSA), FOC, an Optical circulator, and a temperature control module (TEC).

The relationship between the estimated frequency and temperature by the spectral resonance wavelength shift demodulation method is shown in fig. 4. The measurement range of the FPI temperature sensor is only about 5 ℃, but the spectrum frequency demodulation algorithm provided by the invention is not limited by spectrum FSR, and the relation graph between the estimated frequency and the temperature is shown in figure 5, so that the dynamic range is obviously improved greatly.

Since the Rife algorithm only uses the maximum value of the FFT discrete data and the adjacent secondary large value when estimating the dominant frequency of the FPI reflection spectrum, the Rife algorithm is also suitable for estimating the dominant frequency of the reflection spectrum of a plurality of sensing units at the same time. And (3) simulating the multi-FPI parallel reflection spectrum by utilizing Matlab, and estimating corresponding main frequency.

As shown in fig. 6, which is a schematic diagram of a parallel structure of four FPI temperature sensors, input light is uniformly incident to the four FPIs through a 2 × 4 coupler, and reflected light is output through another arm of the coupler. The reflection spectra of the four FPIs are shown in fig. 7, wherein (a), (b), (c), and (d) in fig. 7 correspond to the reflection spectra of the four FPIs, respectively. After the parallel structure, the spectra will be linearly superimposed, and the spectrum generated after the superimposition is shown in fig. 8.

The spectra formed after the superimposition are still periodic (non-strict) images, but the information of the respective spectra cannot be read at all. The spectrum is subjected to wave number conversion and FFT to obtain the result shown in fig. 9. It can be seen that the four FPIs are easily distinguishable in the frequency domain due to the difference in initial cavity length. The larger the lumen length, the larger the corresponding dominant frequency. The four main frequencies on the frequency spectrum can be accurately estimated by the Rife algorithm, the estimation results are shown in Table 2, and the results show that if the cavity length difference of the four FPIs is 10 μm, the maximum estimation error is as high as 0.227%, and the corresponding temperature estimation error is +/-1.25 ℃.

TABLE 2 estimated error Table at 10 μm cavity length intervals

If the cavity length spacing between the FPIs is increased to 20 μm, e.g., 60 μm, 80 μm, 100 μm, and 120 μm, respectively, the corresponding estimation errors are shown in Table 3, it can be seen that the maximum error is reduced to 0.128%, and thus the temperature error is about + -0.7 deg.C.

TABLE 3 estimated error Table for 20 μm cavity Length spacing

Therefore, through analog simulation, the reflection superposition spectrum of the FPI probes can also estimate the respective corresponding main frequency through a Rife algorithm when the FPI probes are subjected to parallel multiplexing; if the initial cavity length spacing of the FPI is large (e.g., 20 μm), the effect of cross-talk between sensors during the estimation process is small. Thus demonstrating the feasibility of the cavity length demodulation algorithm for FPI multiplexing.

After the simulation was completed, two FPI temperature sensing units FPI1 and FPI2 having initial cavity lengths of about 99.90 μm and 120.12 μm, respectively, were prepared and connected in parallel, the optical paths are as shown in fig. 10, and the ASE light source, the spectrometer and the FPI sensor were connected to both ends of the 3dB coupler, respectively. The FPI1 was placed in the card slot of the TEC temperature control module, the FPI2 was placed in air, and the TEC temperature was adjusted to rise from 20 ℃ to 30 ℃ and reflectance spectra were recorded at 1 ℃ intervals. After the reflection spectrum is subjected to wave number conversion, two main frequencies of the wave number spectrum are estimated by a Rife algorithm, and the relationship between the main frequencies and the temperature is shown in fig. 11. As can be seen from the figure, the relationship between the main frequency of the FPI1 and the outside temperature is linear, while the main frequency of the FPI2 is basically not influenced by the temperature and is jittered only around one level, and the standard deviation of jitter is 3.086 multiplied by 10^-8The average jitter was about. + -. 0.4 ℃.

If the temperature of the FPI1 is raised from 20 ℃ to 60 ℃, its test results are shown in fig. 12. As with the single FPI measurement, the relationship between the dominant frequency of the FPI1 and temperature is a quadratic function when measuring a wide range of temperatures. While the FPI2 has its dominant frequency substantially unchanged due to its placement in air.

It can be seen that the FPI1 dominant frequency changes over the temperature range of 20 ℃ to 60 ℃. Ambient temperatures in the range of 20 ℃ to 60 ℃ can be detected by the major frequency change of the FPI 1. Wherein, the change of the FPI1 main frequency reflects the change of the FPI1 cavity length, and the corresponding relation exists between the cavity length change and the temperature change, so that the external environment temperature can be determined through the change of the FPI1 cavity length.

Since the FPI2 is placed in the air, its temperature does not change, so its dominant frequency remains substantially unchanged, further confirming that the change in ambient temperature can be detected by the dominant frequency of the FPI, i.e., the cavity length, and the effect of crosstalk between the sensors FPI1 and FPI2 is small.

Fig. 13 is a structural diagram of an FPI cavity length demodulation system provided in the present invention, including: a spectrum acquisition unit 1310, a wavenumber interpolation unit 1320, a spectral frequency determination unit 1330, and an FPI cavity length determination unit 1340.

A spectrum collection unit 1310 for collecting a reflection spectrum of the FPI and performing a wave number transform on the spectrum;

a wave number interpolation unit 1320, configured to interpolate the wave number data after the wave number transformation, to obtain discrete wave number data with uniform intervals;

a spectral frequency determining unit 1330 configured to determine a wavenumber spectral frequency by using a Rife algorithm based on the uniformly spaced discrete wavenumber data;

and an FPI cavity length determining unit 1340 for determining the cavity length of the FPI based on the wavenumber spectrum frequency, wherein the change of the cavity length can determine the change of the ambient factors, such as temperature and the like.

Optionally, the spectrum collection unit 1310 first determines a relationship between FPI reflectivity and spectral wavelength; secondly, determining the relation between the spectral wavelength and the wave number; finally, the relationship between the FPI reflectivity and the wave number is determined.

Alternatively, the wave number interpolation unit 1320 performs cubic spline interpolation on the wave number based on the relationship between the FPI reflectivity and the wave number, to obtain discrete wave number data at uniform intervals.

Optionally, the spectral frequency determination unit 1330 performs fast fourier transform on the uniformly spaced discrete wave number data; the wavenumber spectral frequencies are determined based on the discrete wavenumber data after the fast fourier transform.

Alternatively, the relationship of the wavenumber spectral frequency to the FPI cavity length is expressed as follows:

where f represents the wavenumber spectral frequency, ω represents the spectral angular frequency, and L represents the FPI cavity length.

It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.