CN101008570A - Method for separating thermal noise, shot noise and intensity noise of optical fiber gyroscope - Google Patents

Abstract

This invention discloses one light fiber top noise, discrete noise and intensity noise separation method, which comprises the following steps: processing phase modulation on system according to selection phase; high speed digital collection circuit processes sample of output signal in different modulation depth; through sample sequence analyzing getting relative noise even and variance; according to different function, computing three kinds of noise value.

Description

The method that optical fibre gyro thermonoise, shot noise and intensity noise separate

Technical field

The present invention relates to the field of Fibre Optical Sensor in the photoelectron technology, particularly relate to the method that a kind of optical fibre gyro thermonoise, shot noise and intensity noise separate.

Background technology

White noise in the optical fibre gyro system is the principal element of system for restricting static accuracy, and it directly translates into the angle random migration performance index of optical fibre gyro.This index ground size has directly reflected the quality of system's static properties.Have three kinds of basic white noise sound sources in optical fibre gyro: detector is striden thermal resistance noise V _Themal, shot noise I _ShotnoiseAnd intensity noise I _RIN, the intensity of these three kinds of white noises can be expressed as respectively:

$V_{\mathrm{themal}}=\sqrt{4\mathrm{kTR}*\mathrm{BW}}$ , perhaps $I_{\mathrm{themal}}=V_{\mathrm{themal}}/R=\sqrt{4\mathrm{kTBW}/R}$

$I_{\mathrm{shotnoise}}=\sqrt{2\mathrm{eIBW}}---\left(1\right)$

$I_{\mathrm{RIN}}=\sqrt{I^2*\mathrm{BW}*{\mathrm{\λ}}^2/\mathrm{c\Δ\λ}}$

Wherein: k is a thermodynamic equilibrium constant, and e is an electron charge, and c is the light velocity in the vacuum, BW is the optical fibre gyro system bandwidth, and I is the detector signal electric current, and R is that photodetector is striden resistance, T is the thermodynamic temperature that photodetector is striden resistance, and λ is the mean wavelength of system source, and Δ λ is the spectral width of light source.

The formation mechanism difference of these three different white noise sound sources, wherein detector is striden the random thermal motion of thermal resistance noise source in resistance internal freedom electronics, and on certain bandwidth and fixing temperature, its amplitude is only determined by the size of resistance and signal intensity I has nothing to do; Flow through carrier number purpose random fluctuation and shot noise derives from PN junction, the square root of its amplitude and signal intensity I is in direct ratio; Additional interferometer that intensity noise is then caused by pad reflected light in the fluctuation of the luminous power of light source, the optical system and backscattering noise superimposed and bring, its amplitude and signal intensity I are the directly proportional relation.Since the mechanism difference that the three produces, the independent random of respectively doing for oneself noise process, this moment, three kinds of white noises were handled according to independent noise source, and system noise is that three kinds of noise power algebraically add, and promptly system's overall noise can be expressed as:

$I_{\mathrm{Total}}^2=I_{\mathrm{Thermal}}^2+I_{\mathrm{ShotNoise}}^2+I_{\mathrm{RIN}}^2=\frac{V_{\mathrm{Thermal}}^2}{R}+I_{\mathrm{ShotNoise}}^2+I_{\mathrm{RIN}}^2---\left(2\right)$

Perhaps: $V_{\mathrm{Total}}^2=V_{\mathrm{Thermal}}^2+V_{\mathrm{ShotNoise}}^2+V_{\mathrm{RIN}}^2=I_{\mathrm{Thermal}}^{2}R+I_{\mathrm{ShotNoise}}^2{R}^2+I_{\mathrm{RIN}}^2{R}^2---\left(3\right)$

In the actual gyrosystem, the decision systems static accuracy be the size of system's overall noise, because the intensity of three kinds of noise sources has nothing in common with each other, contribution to system's static accuracy is also different, to effectively improve the static accuracy of gyro, need to suppress and reduce the size of system's overall noise, and the most effectual way of reduction system noise is to suppress the overriding noise source of gyro itself, promptly reduces the intensity of the noise source of mean square value maximum.Therefore need set up the noise model of system in detail, the noise source of mean square value maximum in the accurate in locating system, thus take to suppress the noise scheme targetedly, accomplish to shoot the arrow at the target.And can only obtain system's overall noise size in the actual test process, can't carry out independent test to the intensity of each white noise sound source, therefore be necessary on test basis system noise, system noise is carried out discrete testing and location, determine the intensity of different noise sources, thereby set up system's white noise model, provide strong analysis tool, point out an improvement direction for improving system's static properties.

By first three expression formula (1) of planting the white noise sound source as can be known, three kinds of white noise sound sources have different funtcional relationships with signal intensity I, and wherein detector is striden thermal resistance noise power V _Thermal ²Irrelevant with signal intensity I, and shot noise power V _ShotNoise ²I is proportional with signal intensity, intensity noise power V _RIN ²Then be and signal intensity square, also promptly and I ²Be directly proportional, by changing the size of signal intensity, can obtain one group of different equation about signal intensity and noise amplitude, the intensity by system noise under the test unlike signal intensity can obtain three kinds of noises intensity separately by the simultaneous solution equation.Only need the noise intensity under three varying strengths of test in theory, obtaining three equations is that solvable equation obtains three noise sizes, but this method precision is limited, and credible result is poor.

Summary of the invention

In the research of present optical fibre gyro through engineering approaches, still there is not the present situation that test separates the inner various different white noise sound source effective ways of optical fibre gyro itself, the method that the object of the present invention is to provide a kind of optical fibre gyro thermonoise, shot noise and intensity noise to separate.

Inventive principle:

According to the principle of Sagnac interferometer, on system's photodetector output signal can be expressed as:

I＝I ₀[1+cos(φ _m+φ _Ω)] (4)

φ wherein _mBe the depth of modulation of system, and φ _ΩBe the Saganc phase shift that system's input speed is produced, φ under the system quiescence state _Ω=0, signal intensity can be reduced on the detector at this moment:

I＝I ₀(1+cosφ _m) (5)

Signal is depth of modulation φ _mFunction.With (5) formula substitution (2) formula, the expression that associating (1) can get system noise is:

$I_{\mathrm{Total}}^2=\frac{V_{\mathrm{Thermal}}^2}{R}+I_{\mathrm{ShotNoise}}^2+I_{\mathrm{RIN}}^2=\frac{V_{\mathrm{Thermal}0}^2}{R}+I_{\mathrm{ShotNoise}0}^2(1+{\cos \mathrm{\φ}}_m)+I_{\mathrm{RIN}0}^2{(1+{\cos \mathrm{\φ}}_m)}^2---\left(6\right)$

According to three kinds of noises is separate zero-mean white-noise process, and the principle according to the power stack is expressed as with variance:

${\mathrm{STD}}_{\mathrm{totalNoise}}^2={\mathrm{STD}}_{\mathrm{Thermal}}^2+{\mathrm{STD}}_{\mathrm{ShotNoise}0}^2(1+{\cos \mathrm{\φ}}_m)+{\mathrm{STD}}_{\mathrm{RIN}0}^2{(1+\cos {\mathrm{\φ}}_m)}^2$

$=a+b\left({1+\cos \mathrm{\φ}}_m\right)+c{(1+\cos {\mathrm{\φ}}_m)}^2---\left(7\right)$

Wherein: $a={\mathrm{STD}}_{\mathrm{Thermal}}^2,$ $b={\mathrm{STD}}_{\mathrm{ShotNoise}0}^2,$ $c={\mathrm{<figure-callout\; id="0"\; label="STD\; RIN"\; filenames="A20071006697000111.png"\; state="\{\{state\}\}">STD</figure-callout>}}_{\mathrm{RIN}}^0,$ Represented the mean square value of three kinds of white noise sound sources in the system respectively, directly corresponding to the intensity of three kinds of white noise sound sources, for needing the amount to be asked of discrete testing among the present invention, and STD _TotalNoise ²Being the system noise mean square value, is can be by the direct amount that obtains of test and analytical calculation in the reality.

Definition fits independent variable x=1+COS φ _m, then from formula (7), can see: STD _TotalNoise ²Be the quadratic function of variable x, and three kinds of white noise intensity is respectively to three coefficients that should quadratic function.By changing depth of modulation, obtain a series of x values as variable, and the mean square value of the system noise under the corresponding depth of modulation of test calculating, thereby obtain corresponding to fitting independent variable x (k) and systematic variance STD _TotalNoise ²(k) two sequences utilize least square to fit this two sequences, obtain three and fit coefficient, promptly obtain the intensity ground numerical value corresponding to three kinds of white noises of system, thereby realize the separation of optical fibre gyro white noise sound source and the test of intensity.

The step of this method of the technical solution adopted in the present invention is as follows:

(1) at 0～φ _Max, φ _MaxIn＞2 π phase place zones, choose some N+1 number and form a phase place system degree of depth sequence φ _m(k);

(2) according to a phase modulation (PM) degree of depth of the phase modulation (PM) degree of depth sequence φ in (1) _m, optical fibre gyro system is carried out phase modulation (PM) by numeral or analog waveform;

(3) to being in the output signal of optical fibre gyro under this phase modulation (PM) degree of depth, sample, obtain corresponding sample sequence, obtain noise mean square value STD under this phase modulation (PM) degree of depth by analyzing and processing to this sample sequence by the high-speed figure Acquisition Circuit ²(φ _m);

(4) to each the phase modulation (PM) depth value in the phase modulation (PM) degree of depth sequence that contains the N+1 number, repeat above (2), (3) step, method with (2) is carried out phase modulation (PM) to system, and obtain noise mean square value, thereby obtain the corresponding noise mean square value sequence of phase modulation (PM) velocity series STD in corresponding and (1) corresponding to the respective phase depth of modulation according to the method in (3) ²[φ _m(k)];

(5) according to STD ²[φ _m(k)]=a+bx (k)+cx ²(k), to obtaining noise mean square value sequence STD in (4) ²[φ _m(k)], and by phase modulation (PM) sequence φ _mWhat (k) generate fits independent variable sequence x (k)=1+cos φ _m(k), method with least square fits calculating, obtain fitting the constant term coefficient a of independent variable and the coefficient c of a coefficient b, quadratic term once, correspond respectively to thermonoise, shot noise and intensity noise mean square value, thereby the test that realizes three kinds of noises separates.

Because optical fibre gyro is a τ time delay differential modulation, τ is the transit time of optical fibre gyro fiber optic loop.Square wave is produced by signal generator, the main devices of signal generator is digital to analog converter DAC, digital to analog converter DAC is responsible for the digital signal of from processor cpu is converted to simulating signal, deliver to the numerical code of digital to analog converter DAC by change, thereby adjust the amplitude of square wave, modulation is equivalent to the size that changes depth of modulation through time delay.

Choosing of depth of modulation sequence can be according to following two kinds of distinct methods or its equivalent method:

A, depth of modulation are uniformly-spaced chosen, promptly ${\mathrm{\&phi;}}_m\left(k\right)=(k-1)\frac{{\mathrm{\&phi;}}_{\max }}{N},k=\mathrm{1,2},...N+1,$ Then fit variable:

$x\left(k\right)=1+\cos \left[(k-1)\frac{{\mathrm{\&phi;}}_{\max }}{N}\right],k=\mathrm{1,2},...N+1$

B, fit variable and uniformly-spaced choose, promptly $x\left(k\right)=(k-1)\frac{2}{N},k=\mathrm{1,2},...N+1,$ Corresponding depth of modulation sequence is: ${\mathrm{\&phi;}}_m\left(k\right)=a\cos \left[(k-1)\frac{2}{N}\right],k=\mathrm{1,2}...N+1$

By the high-speed figure sample circuit output signal of optical fibre gyro is sampled, obtain corresponding to the sample sequence s under this depth of modulation _{φ m}(k), wherein k is the sequence number of the sampled value of sample sequence, and sample length is n.The s that this moment, test obtained _{φ m}(k) comprising the noise and the signal of system, is a stochastic variable, its DC quantity $\mathrm{mean}\left({\mathrm{\&phi;}}_m\right)=\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}{s}_{{\mathrm{\&phi;}}_m}\left(k\right)$ Corresponding to the signal of system, of ac is corresponding to the noise of system.This sequence is carried out data processing, and its mean square deviation is corresponding to the mean square value of system noise under this depth of modulation, that is:

${\mathrm{STD}}_{\mathrm{TotalNoise}}^2\left({\mathrm{\&phi;}}_m\right)=\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}{[s_{{\mathrm{\&phi;}}_m}\left(k\right)-\mathrm{mean}\left(s_{{\mathrm{\&phi;}}_m}\left(k\right)\right)]}^2.$

The beneficial effect that the present invention has is: proposed the method for test and three kinds of white noise sound sources of separation optical fibre gyro first, this method need not increase any additional firmware, is entirely on-line testing, carries out in the gyro working environment of reality; Method of testing is simple and convenient, and operation has very high practical value easily; Data processing method reliability and precision that system is used are all high, and can further improve the precision of actual test result by increasing number of test points n; Whole method of testing has the advantage of low cost, high precision and high reliability; Do not have other at present and can realize the method for testing of same target.Utilize this technology, can set up the noise model of optical fibre gyro static properties, the bottleneck of location system for restricting static properties is to improve its static accuracy.

Description of drawings

Fig. 1 is an exemplary test system theory of constitution block diagram;

The square-wave frequency modulation waveform that Fig. 2 is to use and the sequential chart of Optical Fiber Gyroscope sample mode;

Fig. 3 is the white noise of optical fibre gyro system below different depth of modulation;

Fig. 4 utilizes this technical method certain optical fibre gyro system to be carried out the result of noise separation match.

Embodiment

The invention will be further described below in conjunction with drawings and Examples:

Fig. 1 is an exemplary test system theory of constitution block diagram, wherein the light that sends of light source 11 is divided into two bundles by first beam splitter 13 of system, wherein a branch of by second beam splitter 14 of optical fiber arrival, be divided into two again, and be injected in the two ends of optical fiber 17, respectively along the propagating clockwise and counterclockwise of fiber optic loop 17, thereby and the time delay modulation affix that is subjected to modulator 16 depth of modulation by the modulation waveform decision is arranged, modulated waveform is by 18 generations.Light through the time delay modulation, after propagating respectively, converge again and interference mutually, interference light is reversed beam splitting from beam splitter 14, wherein a part arrives first beam splitter 13 along optical fiber, and be divided into two once more, part light reaches detector 12, realizes that by the detector opto-electronic conversion light signal is converted to electric signal, electric signal then carries out the signal filtering amplification by follow-up processing of circuit part 19 and sample conversion is a digital signal, and uploads to host computer for analyzing usefulness by communication port.

The step of this method is as follows:

At 0～φ _Max, φ _Max＞2 π phasetophases are chosen N+1 test point and are formed depth of modulation sequence φ _m(k), wherein N is not less than 2, tests simultaneously and calculate to fit independent variable sequence x (k) accordingly, and choosing of depth of modulation sequence can be according to following two kinds of distinct methods or other similar methods:

A, depth of modulation are uniformly-spaced chosen, promptly ${\mathrm{\&phi;}}_m\left(k\right)=(k-1)\frac{{\mathrm{\&phi;}}_{\max }}{N},k=\mathrm{1,2},...N+1,$ Then fit variable:

$x\left(k\right)=\cos \left[(k-1)\frac{{\mathrm{\&phi;}}_{\max }}{N}\right],k=\mathrm{1,2},...N+1$

B, fit variable and uniformly-spaced choose, promptly $x\left(k\right)=(k-1)\frac{2}{N},k=\mathrm{1,2},...N+1,$ Corresponding modulation is dark

The degree sequence is: ${\mathrm{\&phi;}}_m\left(k\right)=a\cos \left[(k-1)\frac{2}{N}\right],k=\mathrm{1,2},...N+1$

Phase place according to any point in the depth of modulation sequence of choosing in (1) is modulated system, realizes the phase modulation (PM) function by integrated electro-optic modulator YWG or piezoelectric modulator PZT device in the practical application.

Get modulation waveform with square wave SQUART (t) as system among Fig. 2 a, because it is optical fibre gyro is the time delay differential modulation, actual that the phase modulation (PM) depth size is φ _m(k)=and SQUART (t)-SQUART (t-τ), τ is the transit time of optical fibre gyro fiber optic loop.Square wave is produced by signal generator, the main devices of signal generator is digital to analog converter DAC, digital to analog converter DAC is responsible for the digital signal of from processor cpu is converted to simulating signal, deliver to the numerical code of digital to analog converter DAC by change, thereby adjust the amplitude of square wave, after time delay τ modulation, be equivalent to the size that changes depth of modulation.

When method is modulated in according to (2), the output signal of optical fibre gyro is sampled, obtain corresponding to the sample sequence s under this depth of modulation by the high-speed figure sample circuit _{φ m}(k), wherein k is the sequence number of the sampled value of sample sequence, and sample length is n.The s that this moment, test obtained _{φ m}(k) comprising the noise and the signal of system, is a stochastic variable, its DC quantity $\mathrm{mean}\left({\mathrm{\&phi;}}_m\right)=\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}{s}_{{\mathrm{\&phi;}}_m}\left(k\right)$ Corresponding to the signal of system, of ac is corresponding to the noise of system.This sequence is carried out data processing, and its mean square deviation is corresponding to the mean square value of system noise under this depth of modulation, that is: ${\mathrm{STD}}_{\mathrm{TotalNoise}}^2\left({\mathrm{\&phi;}}_m\right)=\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}{[s_{{\mathrm{\&phi;}}_m}\left(k\right)-\mathrm{mean}\left(s_{{\mathrm{\&phi;}}_m}\left(k\right)\right)]}^2.$

For the depth of modulation sequence φ that is generated in (1) _m(k), realize phase modulation (PM), obtain corresponding STD with the method for (3) according to (2) method _TotalNoise ²(φ _m) sequence, with STD _TotalNoise ²(φ _m) be variable, be independent variable with x (k), according to (7) formula, carry out second order with least square method and fit, obtain a, b, the value of c is separated the intensity size that obtains three kinds of noises.With the phase modulated signal of square wave as system, realize the control of phase modulation (PM) change in depth by changing square wave amplitude, as shown in Figure 2,21 is square-wave frequency modulation waveform Ф (t), its peak-to-peak value is φ _m(k), square-wave cycle is τ eigenperiod of twice gyro, and 22 is time delay difference debug signal, and 22 waveform is directly determined according to following formula by 2a1: φ _m(t)=and Ф (t)-Ф (t-τ), can find depth of modulation φ in the real system _m(t) only get ± φ _m(k) numerical value is by control modulated square wave peak-to-peak value φ _m(k), can change the depth of modulation of system.And 23 be the output of optical fibre gyro system under this modulation waveform, and wherein burr s is because the limited rise time on modulated square wave 21 edges brings.S is the sampling pulse of system in 24, because system is a dynamic bias, depth of modulation at peak-to-peak value is ± φ _m(k) switch between, during sampling only to φ _m(k) gyro of positively-modulated correspondence output P part signal is sampled, and for-φ _mWhether (k) gyro of negatively-modulated correspondence output N part signal is not sampled, control high speed acquisition circuit by sampling pulse s and sample.

Among Fig. 3,31 is the time dependent changing value of the system modulation degree of depth, and ordinate is a radian, and horizontal ordinate is the time, and 32 is the sequence that the noise samples result corresponding to 31 depth of modulation forms, and peak-to-peak value changes with the variation of depth of modulation.

According to system noise mean square value sequence STD ²[φ _mAnd fit independent variable x (k)=1+cos φ (k)] _m(k) two sequences, the equivalent method with least square method fits by function, calculates the method for separating the intensity that obtains three kinds of different noises.Among Fig. 4,41 lines are mean square values of the system noise that obtains according to the test of described method, and 42 dotted lines are for to isolate three kinds of noises according to the described method of fitting, and with the fitted results as the system noise mean square value of synthesizing of three kinds of noises.

Claims (7)

1, the method for optical fibre gyro thermonoise, shot noise and intensity noise separation is characterized in that the step of this method is as follows:

(1) at 0～φ _Max, φ _MaxIn＞2 π phase place zones, choose some N+1 number and form a phase place system degree of depth sequence φ _m(k);

(2) according to a phase modulation (PM) degree of depth of the phase modulation (PM) degree of depth sequence φ in (1) _m, optical fibre gyro system is carried out phase modulation (PM) by numeral or analog waveform;

(3) to being in the output signal of optical fibre gyro under this phase modulation (PM) degree of depth, sample, obtain corresponding sample sequence, obtain noise mean square value STD under this phase modulation (PM) degree of depth by analyzing and processing to this sample sequence by the high-speed figure Acquisition Circuit ²(φ _m);

(4) to each the phase modulation (PM) depth value in the phase modulation (PM) degree of depth sequence that contains the N+1 number, repeat above (2), (3) step, method with (2) is carried out phase modulation (PM) to system, and obtain noise mean square value, thereby obtain the corresponding noise mean square value sequence of phase modulation (PM) velocity series STD in corresponding and (1) corresponding to the respective phase depth of modulation according to the method in (3) ²[φ _m(k)];

(5) according to STD ²[φ _m(k)]=α+bx (k)+cx ²(k), to obtaining noise mean square value sequence STD in (4) ²[φ _m(k)], and by phase modulation (PM) sequence φ _mWhat (k) generate fits independent variable sequence x (k)=1+cos φ _m(k), method with least square fits calculating, obtain fitting the constant term coefficient a of independent variable and the coefficient c of a coefficient b, quadratic term once, correspond respectively to thermonoise, shot noise and intensity noise mean square value, thereby the test that realizes three kinds of noises separates.

2, the method for optical fibre gyro thermonoise according to claim 1, shot noise and intensity noise separation is characterized in that: at 0～φ _Max, φ _MaxIn the phase range of＞2 π, increase progressively principle, choose N+1 point and form phase modulation (PM) degree of depth sequence φ according to equal difference _m(k), the numerical value of N is not less than 2; Simultaneously according to relation: x (k)=1+cos φ _m(k) obtain fitting independent variable sequence x (k).

3, the method for optical fibre gyro thermonoise according to claim 1, shot noise and intensity noise separation, its feature comprises: in 0～2 numerical range, for 1+cos φ _m(k) value increases progressively principle according to equal difference, chooses N+1 point as fitting variable sequence x (k), and the numerical value of N is not less than 2, and by concerning: x (k)=1+cos φ _m(k) obtain phase modulation (PM) degree of depth sequence φ _m(k)=and acos[1+x (k)], wherein acos is an inverse cosine function.

4, the method for optical fibre gyro thermonoise according to claim 1, shot noise and intensity noise separation, it is characterized in that:, realize the control of phase modulation (PM) change in depth by changing square wave amplitude with square wave or its equivalent waveform phase modulated signal as system.

5, the method for optical fibre gyro thermonoise according to claim 1, shot noise and intensity noise separation, it is characterized in that: use sine wave or its equivalent modulation waveform phase modulated signal, by changing the control that sinusoidal wave period or amplitude realize the phase modulation (PM) degree of depth as system.

6, the method for optical fibre gyro thermonoise according to claim 1, shot noise and intensity noise separation, it is characterized in that: utilize the high-speed digital signal Acquisition Circuit that Optical Fiber Gyroscope is sampled, obtain sample sequence s (k), and calculate noise mean square value STD this moment according to this sampling corresponding to this phase modulation (PM) degree of depth ²[φ _m(k)] method.

7, the method for optical fibre gyro thermonoise according to claim 1, shot noise and intensity noise separation is characterized in that: according to system noise mean square value sequence STD ²[φ _mAnd fit independent variable x (k)=1+cos φ (k)] _m(k) two sequences with least square method and equivalent method thereof, fit by function, calculate the method for separating the intensity that obtains three kinds of different noises.

Images