CN102095538B - Data demodulation method for polarization maintaining fiber stress sensing - Google Patents

Abstract

The invention discloses a data demodulation method for polarization maintaining fiber stress sensing. The method comprises the following steps of: acquiring an interference signal by using a polarization maintaining fiber stress sensing test system to acquire photovoltage data; pre-processing the photovoltage data, averaging every n points to acquire new photovoltage data, and data volume; decomposing a photovoltage data signal into an intrinsic mode function (IMF) component and an allowance by an experimental mode method; finding out a basic component and identifying a small coupling point; reconstructing coupling intensity of a signal, and calculating the stress magnitude through intensity of each coupling point so as to realize the distributed detection of a system; and calculating indexes which can show the system test accuracy, such as a signal-to-noise ratio and the like. The method can effectively improve the signal-to-noise ratio of the polarization maintaining fiber stress sensing system, and improves the identification capacity of weak coupling points and the sensitivity of a coupling test. Under the environment of relatively low signal-to-noise ratio, microstress can be detected well.

Description

The data demodulation method of Stress Profile for Polarization-Maintaining sensing

Technical field

The present invention relates to polarizer stress parameters measuring method, belong to field of optical measuring technologies, relate in particular to a kind of data demodulation method of protecting inclined to one side stresses of parts sensing based on interferometric method optics.

Background technology

During the disturbance of factor such as polarization maintaining optical fibre receives external force, magnetic field, bending, reverse the polarization coupled phenomenon can take place, through can realize the sensing of disturbance physical quantity to the measurement of polarization maintaining optical fibre polarization coupled.The distributed polarization maintaining optical fibre sensor of white light interference polarization coupled is widely used in stress, strain, temperature, gas and position sensing owing to its high spatial resolution and wide dynamic range.Its maximum characteristic is on simple optical fiber, to realize distributed sensing, and any position on the optical fiber can be as sensing unit.Because polarization maintaining optical fibre has the polarization hold facility, makes and in measuring process, need not consider the Polarization Control problem.

This system can be widely used in the structural health detection.In addition, an important use is that it can intrinsic Coupling point of detection fiber gyro ring and polarization interference.

Utilize polarization maintaining optical fibre polarization coupled principle to realize that stress sensing has polarization time domain reflectometry (POTDR), Michelson white light interference method, carrier being frequency modulated (FMCW) method, modelocked fiber cyclic laser method, detects interferometric method and synthesizing optical coherence function method based on Kerr effect detection method, wavelength.Wherein the white light michelson interferometry is simple in structure with it, and measuring accuracy is high and be widely used.

The factor that influences measuring accuracy has: the AD converter noise of the stability of mechanical scanning step-length, power supply, the noise of detector and resolution characteristic, capture card etc.In order to improve system signal noise ratio, need adopt a series of measures, cooperate corresponding data processing to improve experimental result, such as, data pre-service, Fourier smothing filtering, small echo denoising etc.When the frequency band of the useful information of signal and noise did not overlap, the Fourier smothing filtering just can be obtained effect preferably, yet, a specific cutoff frequency is set and is not suitable for nonlinear noise and the signal structure that has sudden change.In the wavelet transformation, in a single day selected when wavelet basis, in whole decomposition, restructuring procedure, just can't change, and the signal part is not had adaptivity.The essence of wavelet transformation is the Fourier transform with flexible time frequency window.The empirical modal method is suitable for handling the time dependent non-linear non-stationary signal of a large amount of frequencies based on signal local characteristics and adaptivity.

Patent 201010183969.9 is sent the in-situ stresses strain data through the timer poll, after arithmetic mean filtering, shows in real time.Patent 200920108831.5 usefulness digital signal processing chips and extensive FPGA array chip demodulate the dependent variable in the fiber grating sensing system as signal processing unit with improved binomial/trinomial fitting algorithm.Most patents all are on system architecture, to do innovation, and this patent combines the dependent variable of empirical modal method demodulation polarization maintaining optical fibre, and is simple in structure, are easy to realize.

Summary of the invention

The present invention seeks to be to improve the recognition capability of weak coupling point in the Stress Profile for Polarization-Maintaining sensing and the signal to noise ratio (S/N ratio) of raising stiffness of coupling test macro, propose a kind of data demodulation method of the Stress Profile for Polarization-Maintaining sensing based on the empirical modal method.

The present invention has taken into full account the coupling effect that many different power on the optical polarization device cause Coupling point, through data pre-service and improved empirical modal algorithm, make measurement result accurately, reliable.

The data demodulation method of Stress Profile for Polarization-Maintaining sensing provided by the invention comprises:

1st, use the Stress Profile for Polarization-Maintaining sensing testing system to gather interference signal, obtain photovoltage data x (t);

2nd, photovoltage data x (t) is carried out the data pre-service, every n point obtains new photovoltage data x after doing on average ₀(t), 0＜n＜q, q are that (value of n need be selected for use by the flexible in size of the data volume that collects for the number of x (t); N is big more, and then average noise reduction is good more, but if original data volume is little; Then n should correspondingly get smaller; For subsequent algorithm provides enough data), purpose is the random noise of effectively eliminating in the acquired signal, and can data volume be dwindled n doubly;

3rd, with the empirical modal method with photovoltage data-signal x ₀(t) be decomposed into M IMF component and a surplus, promptly

Wherein, x ₀(t) be photovoltage data after average, c _i(t) be the IMF component, reflected different frequencies and amplitude composition, M is the number of IMF component, 0＜M＜q ₀, q ₀Be x ₀(t) data volume size, it is after cut-off condition is satisfied in the empirical modal decomposition, the signal number that satisfies the IMF condition of being stripped from, it is with the difference of original signal, and the result can be different.r _M(t) be a surplus, the IMF component that the empirical modal algorithm can be high with frequency at first extracts;

The basic step of empirical modal is following:

3.1st, find signal x ₀(t) all Local Extremum connect maximum value and minimal value respectively, and obtain the coenvelope x of original signal with the cubic spline curve match _Up(t) and lower envelope x _Low(t);

3.2nd, calculate the average m of coenvelope and lower envelope ₁(t)=(x _Up(t)+x _Low/ 2, and ask signal x (t)) ₀(t) with the difference h of gained average ₁(t)=x ₀(t)-m ₁(t);

3.3rd, h ₁(t) regard new raw data as, (the h soon that repeats the 3.1st step and the 3.2nd step ₁(t) regard new raw data as, find h ₁(t) all Local Extremum connect maximum value and minimal value respectively, and obtain its coenvelope h with the cubic spline curve match _Up(t) and lower envelope h _Low(t) the average m of calculating coenvelope and lower envelope ₂(t)=(h _Up(t)+h _Low/ 2, and ask signal h (t)) ₁(t) with the difference h of gained average ₂(t)=h ₁(t)-m ₂(t), by that analogy), satisfying two conditions of IMF up to the signal that generates, is c with the signal definition that satisfies condition ₁(t), the definition surplus is r ₁(t), r ₁(t)=x (t)-c ₁(t);

(two conditions that each IMF need satisfy are following:

1, local maximum and local minimizing number sum will equate with the number of striding zero point or differ one at most.

2, on the time point in office, be zero by the mean value of defined coenvelope of local maximum and the defined lower envelope of local minimum.)

3.4th, surplus r ₁(t) regard new raw data as, repeat the 3.1st step to the 3.3rd step up to extracting all IMF components; When satisfying stop condition, iteration stopping (stop condition utilizes the component standard deviation (SD) of double The selection result as criterion,

Typical SD size is at 0.2 to 0.3);

4th, from the IMF component that decomposites, find out substrate component and identify little Coupling point, concrete grammar is following: each IMF component that will decomposite respectively with surplus addition, i.e. x _i(t)=c _i(t)+r _M(t), obtain separately stiffness of coupling distribution h _i(t) be used to judge substrate and little Coupling point; The stiffness of coupling computing method that generally adopt are: h=10log (I _Cf/ I _Main) ², I wherein _CfAnd I _MainBeing respectively Coupling point interferes envelope and principal maximum to interfere the amplitude of envelope.Because surplus is the trend term of signal, DC component that can representation signal; In conjunction with Stress Profile for Polarization-Maintaining sensor-based system characteristic, the substrate component is the useful signal that identifies at first, is defined as b (t), can think to identify first x of little Coupling point _i(t) component; Physical characteristics by Coupling point; Stiffness of coupling is higher than average level 10dB and promptly is judged to be little Coupling point; Therefore can from figure, identify little Coupling point and reach thereof easily, the scope that contains little Coupling point in this IMF component is superimposed to the substrate component with reconstruction signal, promptly

Wherein

It is the medium and small Coupling point part of each IMF;

5th, reconstruction signal x ' stiffness of coupling (t), stress intensity can calculate through the each point stiffness of coupling, thereby has realized the Distributed Detection of system.

Can set up stiffness of coupling and the optic fibre force size and Orientation has following relation according to the theoretical model of three sections polarization maintaining optical fibres of polarization coupled serial connection strain gauge:

$h={\mathrm{<figure-callout\; id="2"\; label="F"\; filenames="110225081828.png,110225081848.png"\; state="\{\{state\}\}">F</figure-callout>}}^2{\sin }^2\left(2\mathrm{\&alpha;}\right)\&CenterDot;{\left\{\frac{\sin \left[\mathrm{\&pi;}\sqrt{1+{\mathrm{<figure-callout\; id="2"\; label="F"\; filenames="110225081828.png,110225081848.png"\; state="\{\{state\}\}">F</figure-callout>}}^2+2F\cos \left(2\mathrm{\&alpha;}\right)}(l/L_{b0})\right]}{\sqrt{1+{\mathrm{<figure-callout\; id="2"\; label="F"\; filenames="110225081828.png,110225081848.png"\; state="\{\{state\}\}">F</figure-callout>}}^2+2F\cos \left(2\mathrm{\&alpha;}\right)}}\right\}}^2---\left(1\right)$

L wherein _BoBe beat length of polarization maintaining optical fiber, F is a stress intensity, and α is a stress direction, and l is the action length of power, can obtain the optic fibre force size by the stool and urine of polarization coupled point stiffness of coupling.

Advantage of the present invention and beneficial effect:

The present invention can effectively improve the signal to noise ratio (S/N ratio) of Stress Profile for Polarization-Maintaining sensor-based system, increases the recognition capability of weak coupling point, improves the coupling measurement sensitivity.Under the lower environment of signal to noise ratio (S/N ratio), also can quite good detecting go out microstress.

Description of drawings

Fig. 1 is the process flow diagram of the data demodulates algorithm of Stress Profile for Polarization-Maintaining sensing;

Fig. 2 is a system construction drawing of making the Stress Profile for Polarization-Maintaining sensing of the white light interference method;

Output interferogram when Fig. 3 is a Coupling point;

Fig. 4 is the original signal that computer acquisition is arrived;

Fig. 5 is the corresponding stiffness of coupling figure of Fig. 4;

Fig. 6 carries out empirical modal to Fig. 4 to decompose the first seven IMF component and surplus that obtains;

Fig. 7 be with the first six IMF respectively with the surplus addition after the stiffness of coupling figure that tries to achieve;

Fig. 8 is with the later signal of demodulating algorithm reconstruct;

Fig. 9 is (promptly use a diameter 125 μ m, the optical fiber of clapping long 3.1mm experimentizes, and the action direction of power is 15 °, and action length is the polarization maintaining optical fibre of 12.5mm) under embodiment 1 situation, the size of stiffness of coupling and the relation between the stress intensity;

Figure 10 is the stress envelope that this stress sensing system demodulation is come out under embodiment 1 situation.

Among the figure, the 1st, SLD light source, the 2nd, the polarizer, the 3rd, polarization maintaining optical fibre, the 4th, collimation lens, the 5th, turntable, the 6th, Glan prism, the 7th, beam splitter, the 8th, quiet arm, the 9th, swing arm, the 10th, convergent lens, the 11st, stepper motor, the 12nd, capture card, the 13rd, computing machine; I _CfInterfere the amplitude of envelope, I for exciting mould and coupled modes _MainFor exciting mould to interfere the amplitude of envelope, d is Michelson interferometer two arm optical path differences.

Embodiment

Embodiment 1:

Fig. 1 is the process flow diagram of the data demodulates algorithm of Stress Profile for Polarization-Maintaining sensing; Fig. 2 is a hardware configuration of the present invention, wherein, the 1st, SLD light source, the 2nd, the polarizer, the 3rd, polarization maintaining optical fibre; The 4th, collimation lens, the 5th, turntable, the 6th, Glan prism, the 7th, beam splitter, the 8th, quiet arm; The 9th, swing arm, the 10th, convergent lens, the 11st, stepper motor, the 12nd, capture card, the 13rd, computing machine.

The principle and the course of work of the inventive method are following:

The distributed Stress Profile for Polarization-Maintaining sensing testing system of white light interference polarization coupled is based on polarization maintaining optical fibre polarization coupled principle.It utilizes external disturbance to cause polarization maintaining optical fibre generation polarization coupled, obtains the position and the size of disturbance point through the Michelson interferometer demodulation, and its structure is as shown in Figure 2.

The light that SLD light source 1 sends gets into polarization maintaining optical fibre 3 behind the optical fiber polarizer 2, the optical fiber polarizer docks with polarization maintaining optical fibre through adapter, causes that one excites mould in optical fiber, to transmit.Do the time spent when certain position of polarization maintaining optical fibre receives external force, can make this point that polarization coupled takes place and inspire one with the coupled modes of importing light wave polarization direction quadrature.The light wave that carries disturbance information incides on the turntable 4, gets into Michelson interferometer 7,8,9,10 through prism 6 backs.Because the existence of optic fiber polarization mould dispersion makes two patterns have delay inequality Δ τ at the optical fiber exit end, obtain interference signal through index glass 9 these delay inequalities of compensation that move in the Michelson interferometer, it can be expressed as following form:

$I_{\mathrm{out}}=I_0\{1+\exp [-{\left(\frac{d}{L_c}\right)}^2]\cos \left(k_{0}d\right)+\sqrt{h-{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="110225081828.png,110225081848.png"\; state="\{\{state\}\}">h</figure-callout>}}^2}\exp (-\frac{L_c^2{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="110225081828.png,110225081848.png"\; state="\{\{state\}\}">d</figure-callout>}}^2}{2})\cos (\mathrm{\&Delta;\&beta;l}-\frac{{\mathrm{\&omega;}}_0}{c}d)\}---\left(2\right)$

I in the formula ₀For interfering DC component, Δ β is that two feature axis propagation constants of polarization maintaining optical fibre are poor, L _cBe the coherence length of laser, d is Michelson interferometer two arm optical path differences, and c is the light velocity, and h is the Coupling point stiffness of coupling, and l is the length of Coupling point apart from the optical fiber exit end.Fig. 3 representes to have in the optical fiber interferogram of a Coupling point.

The amplitude of stiffness of coupling and interference signal envelope has following relation:

h＝10log(I _cf/I _main) ² (3)

I in the formula _CfInterfere the amplitude of envelope, I for exciting mould and coupled modes _MainInterfere the amplitude of envelope for exciting mould.Distributed sensing system not only is embodied in the position that can locate each disturbance point, goes back its size of energy measurement.Native system is according to the location positioning of interferometer two arm optical path difference d realization stress point, through measuring the size that the Coupling point stiffness of coupling comes deterministic force.Stress point stiffness of coupling and optic fibre force size and Orientation have following relation:

$h={\mathrm{<figure-callout\; id="2"\; label="F"\; filenames="110225081828.png,110225081848.png"\; state="\{\{state\}\}">F</figure-callout>}}^2{\sin }^2\left(2\mathrm{\&alpha;}\right)\&CenterDot;{\left\{\frac{\sin \left[\mathrm{\&pi;}\sqrt{1+{\mathrm{<figure-callout\; id="2"\; label="F"\; filenames="110225081828.png,110225081848.png"\; state="\{\{state\}\}">F</figure-callout>}}^2+2F\cos \left(2\mathrm{\&alpha;}\right)}(l/L_{b0})\right]}{\sqrt{1+{\mathrm{<figure-callout\; id="2"\; label="F"\; filenames="110225081828.png,110225081848.png"\; state="\{\{state\}\}">F</figure-callout>}}^2+2F\cos \left(2\mathrm{\&alpha;}\right)}}\right\}}^2---\left(2\right)$

L wherein _BoBe beat length of polarization maintaining optical fiber, F is a stress intensity, and α is a stress direction, and l is the action length of power.

The spatial resolution L of sensor-based system _RFor:

Δ λ is the light source spectrum width in the formula, and λ is the light source center wavelength.Under the situation that the optical fiber bat is grown and the light source center wavelength is certain, the spatial resolution of system depends primarily on the light source spectrum width.

The concrete steps of Stress Profile for Polarization-Maintaining sensing data demodulation method provided by the invention are following:

Step1 uses the Stress Profile for Polarization-Maintaining sensing testing system to gather interference signal, obtains the photovoltage data.

In the embodiment of the invention, light source is selected the SLD-561 light source of SUPERLUM for use, and can export with 1310nm is the light signal of centre wavelength.Use the DH1718D-4 two-way to follow the tracks of the steady road of voltage stabilizing power supply and 6V is provided DC voltage, carry out data acquisition with the USB6251 of NI as modulator.Regulate SLD output photocurrent at 136.0nA; The dc terms of the photovoltage that obtains is 0.74V; Photodetector output connects the analog differential input port of capture card, and sweep velocity is set to 0.2mm/s, and the capture card sampling rate is set to 200 points of each step motor drive pulse collection.Use diameter 125 μ m, the optical fiber of clapping long 3.1mm experimentizes, and the action direction of power is 15 °, and action length is 12.5mm, and the signal to noise ratio (S/N ratio) of trying to achieve the signal noise section that collects is 56.28dB.

Step2 carries out the data pre-service, asks and obtains new photovoltage data after average for per 100, and as shown in Figure 4, Fig. 5 is its corresponding stiffness of coupling.

Step3 has become 16 IMF components and a surplus with the empirical modal method with signal decomposition, and Fig. 6 has shown preceding 7 natural mode of vibration component IMF1-IMF7 and surplus res.

Step4 finds out the substrate component and discerns little Coupling point.

With each IMF component respectively with surplus addition, i.e. x _i(t)=c _i(t)+r _M(t), obtain separately stiffness of coupling distribution h _i(t), Fig. 7 has shown the first six component h ₁(t)-h ₆(t).From figure, can obviously find out h ₁(t) comprising three little Coupling points, is the useful signal that identifies at first, is basal signal so IMF1 adds res, h ₂(t) compare h ₁(t) comprised a new Coupling point, the scope that can be got new Coupling point by algorithm routine is the 14254-14744 point.Remaining h _i(t) all no longer comprise new Coupling point.So reconstruction signal is IMF1, res and the index of signal length is that the IMF2 summation of 14254-14744 gets final product.

Step5 asks reconstruction signal x ' stiffness of coupling (t), and is as shown in Figure 8.Stress intensity can calculate through the each point stiffness of coupling, and Fig. 9 causes the situation of change of Coupling point stiffness of coupling with the size of power for polarization maintaining optical fibre power under this performance.Stress distribution and size that Figure 10 comes out for sensor-based system demodulation under this performance.Calculating signal to noise ratio (S/N ratio) is 72.04dB.Demodulation method of the present invention can effectively improve the signal to noise ratio (S/N ratio) of Stress Profile for Polarization-Maintaining sensor-based system.Under the high-noise environment, stress position and size that this Stress Profile for Polarization-Maintaining sensor-based system still can be exerted oneself demodulation to cause Coupling point have increased the recognition capability of weak coupling point, have improved the coupling measurement sensitivity.

After using this stress sensing demodulating algorithm, the weak coupling point that is difficult to detect can be detected, and can be found out by Fig. 5 and Fig. 9; Experiment causes Coupling point with containing four power in the polarization maintaining optical fibre; Because ambient noise interference, the signal that hardware system collects is as shown in Figure 5, does not promptly detect to exert oneself to cause the existence of Coupling point; Through obtaining Fig. 9 behind the demodulation method according to the invention, can the size and the position probing of four Coupling points be come out.Be submerged in the stronger noise even signal is described, this algorithm also to a certain extent the power that obtains in the polarization maintaining optical fibre of demodulation cause Coupling point.

Claims (2)

1. the data demodulation method of a Stress Profile for Polarization-Maintaining sensing is characterized in that, this method comprises:

1st, use the Stress Profile for Polarization-Maintaining sensing testing system to gather interference signal, obtain photovoltage data x (t);

2nd, photovoltage data x (t) is carried out the data pre-service, every n point obtains new photovoltage data x after doing on average ₀(t), 0＜n＜q, q are the data volume size of x (t), and purpose is the random noise of effectively eliminating in the acquired signal, and can data volume be dwindled n doubly;

3rd, with the empirical modal method with photovoltage data-signal x ₀(t) be decomposed into M IMF component and a surplus, promptly

Wherein, x ₀(t) be photovoltage data after average, c _i(t) be the IMF component, reflected different frequencies and amplitude composition, M is the number of IMF component, 0＜M＜q ₀, q ₀Be x ₀(t) data volume size, r _M(t) be a surplus, the IMF component that the empirical modal algorithm can be high with frequency at first extracts;

4th, from the IMF component that decomposites, find out substrate component and identify little Coupling point, concrete grammar is following: each IMF component that will decomposite respectively with surplus addition, i.e. x _i(t)=c _i(t)+r _M(t), obtain separately stiffness of coupling distribution h _i(t) be used to judge substrate and little Coupling point; The stiffness of coupling computing method that generally adopt are: h=10log (I _Cf/ I _Main) ², I wherein _CfAnd I _MainBeing respectively Coupling point interferes envelope and principal maximum to interfere the amplitude of envelope; Because surplus is the trend term of signal, the DC component of representation signal; In conjunction with Stress Profile for Polarization-Maintaining sensor-based system characteristic, the substrate component is the useful signal that identifies at first, is defined as b (t), thinks to identify first x of little Coupling point _i(t) component; Physical characteristics by Coupling point; Stiffness of coupling is higher than average level 10dB and promptly is judged to be little Coupling point; Therefore can from figure, identify little Coupling point and reach thereof easily, the scope that contains little Coupling point in this IMF component is superimposed to the substrate component with reconstruction signal, promptly

Wherein It is the medium and small Coupling point part of each IMF;

5th, reconstruction signal x ' stiffness of coupling (t), stress intensity calculates through the each point stiffness of coupling, thereby has realized the Distributed Detection of system.

2. method according to claim 1 is characterized in that the computing method of described stress intensity of the 5th step are following:

Can set up stiffness of coupling and the optic fibre force size and Orientation has following relation according to the theoretical model of three sections polarization maintaining optical fibres of polarization coupled serial connection strain gauge:

$h=F^2{\sin }^2\left(2\mathrm{\&alpha;}\right)\&CenterDot;{\left\{\frac{\sin \left[\mathrm{\&pi;}\sqrt{1+F^2+2F\cos \left(2\mathrm{\&alpha;}\right)}(l/L_{\mathrm{bo}})\right]}{\sqrt{1+F^2+2F\cos \left(2\mathrm{\&alpha;}\right)}}\right\}}^2---\left(1\right)$

L wherein _BoBe beat length of polarization maintaining optical fiber, F is a stress intensity, and α is a stress direction, and l is the action length of power, can obtain the optic fibre force size by the stool and urine of polarization coupled point stiffness of coupling.

Images