CN103983261B - A fiber optic gyroscope based on vector space analysis and a signal processing method thereof - Google Patents
A fiber optic gyroscope based on vector space analysis and a signal processing method thereof Download PDFInfo
- Publication number
- CN103983261B CN103983261B CN201410174900.8A CN201410174900A CN103983261B CN 103983261 B CN103983261 B CN 103983261B CN 201410174900 A CN201410174900 A CN 201410174900A CN 103983261 B CN103983261 B CN 103983261B
- Authority
- CN
- China
- Prior art keywords
- signal
- signals
- optical fiber
- gyroscope
- adaptive
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 12
- 238000003672 processing method Methods 0.000 title claims abstract description 11
- 239000000835 fiber Substances 0.000 title abstract description 21
- 238000001914 filtration Methods 0.000 claims abstract description 42
- 238000000034 method Methods 0.000 claims abstract description 42
- 239000013307 optical fiber Substances 0.000 claims abstract description 42
- 230000003044 adaptive effect Effects 0.000 claims abstract description 38
- 230000004044 response Effects 0.000 claims abstract description 38
- 238000000513 principal component analysis Methods 0.000 claims abstract description 27
- 238000012360 testing method Methods 0.000 claims abstract description 5
- 238000001514 detection method Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 18
- 230000000694 effects Effects 0.000 claims description 16
- 238000012935 Averaging Methods 0.000 claims description 5
- 238000002474 experimental method Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000001228 spectrum Methods 0.000 claims description 4
- IIPYXGDZVMZOAP-UHFFFAOYSA-N lithium nitrate Inorganic materials [Li+].[O-][N+]([O-])=O IIPYXGDZVMZOAP-UHFFFAOYSA-N 0.000 claims description 3
- 238000005457 optimization Methods 0.000 abstract description 5
- 230000001360 synchronised effect Effects 0.000 abstract description 3
- 230000002596 correlated effect Effects 0.000 abstract description 2
- 230000005764 inhibitory process Effects 0.000 abstract 2
- 230000008569 process Effects 0.000 description 11
- 238000005259 measurement Methods 0.000 description 10
- 230000003287 optical effect Effects 0.000 description 10
- 238000010586 diagram Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 238000004422 calculation algorithm Methods 0.000 description 6
- 230000008649 adaptation response Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000005295 random walk Methods 0.000 description 3
- 230000001629 suppression Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000010355 oscillation Effects 0.000 description 2
- 230000010287 polarization Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 229910013553 LiNO Inorganic materials 0.000 description 1
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000000875 corresponding effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000005426 magnetic field effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000005316 response function Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C19/00—Gyroscopes; Turn-sensitive devices using vibrating masses; Turn-sensitive devices without moving masses; Measuring angular rate using gyroscopic effects
- G01C19/58—Turn-sensitive devices without moving masses
- G01C19/64—Gyrometers using the Sagnac effect, i.e. rotation-induced shifts between counter-rotating electromagnetic beams
- G01C19/72—Gyrometers using the Sagnac effect, i.e. rotation-induced shifts between counter-rotating electromagnetic beams with counter-rotating light beams in a passive ring, e.g. fibre laser gyrometers
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Optics & Photonics (AREA)
- Electromagnetism (AREA)
- Power Engineering (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Gyroscopes (AREA)
Abstract
The invention relates to a fiber optic gyroscope based on vector space analysis and a signal processing method thereof. The method comprises the steps of: firstly, using an open loop test fiber optic gyroscope to conduct orthogonal balanced local signal demodulation, thus forming two orthogonal signals, and using an ordinary method to obtain an expectation signal at the same time; secondly, conducting a predetermined adaptive filtering on the obtained two orthogonal signals and expectation signal, conducting principal component analysis treatment on an error signal output from the adaptive filter, and conducting linear combination on response signals output from the adaptive filter, thus obtaining an optical fiber loop rotation angular velocity; finally, conducting a quasi Kalman filtering, which is based on signal to noise ratio, on the error signals and response signals , to obtain the final optimized angular velocity. The gyroscope and the signal processing method thereof are capable of overcoming problems of filter delay and difficult model parameter setting caused by inhibition of detecting noise in conventional optical fiber gyroscopes, are capable of realizing the inhibition of correlated noise of an optical fiber gyroscope system and signal synchronous optimization.
Description
Technical Field
The invention belongs to the field of optical fiber sensors, and particularly relates to a novel open-loop optical fiber gyroscope based on vector signal detection and signal space analysis and an optical fiber gyroscope signal estimation method.
Background
A gyroscope is an inertial angular velocity measurement sensor used to determine the rate of rotation of a carrier relative to an inertial reference frame. Gyroscopes are widely used in the fields of guidance and attitude control, and in the scientific fields of time-space precision measurement such as gravitational wave detection. There are mainly five types of gyroscopes currently in use: mechanical gyroscopes, electrostatic gyroscopes, laser gyroscopes, Fiber-optic gyroscopes (FOG), and quantum gyroscopes. The optical gyroscope based on the Sagnac effect is a laser gyroscope and a fiber optic gyroscope. The optical fiber gyroscope has the remarkable advantages of low process complexity, compact structure, small volume, capability of realizing strapdown combination, low maintenance cost and the like, so that the requirement on the optical fiber gyroscope in the field of inertial measurement always exists. Meanwhile, with the improvement of the technological level of photoelectric devices, various performance parameters of the optical gyroscope are continuously improved, and the market share is also continuously improved. However, the optical propagation and the photoelectric detection in the fiber waveguide introduce larger short-time noise and long-time zero offset compared with the conventional gyroscope, so the wandering coefficient and the zero offset stability of the fiber optic gyroscope are inferior to those of some mechanical gyroscopes and quantum gyroscopes.
The optical structure of the optical gyroscope is based on a sagnac type interferometer, and the structure needs to meet reciprocity conditions of polarization reciprocity, mode reciprocity, beam splitter reciprocity and the like. The reciprocity can ensure that the propagation states of clockwise light and anticlockwise light are consistent with a transmission path to the maximum extent, and the effect of common mode rejection is achieved, so that noise introduced by non-reciprocity is eliminated. Fig. 1 shows a minimum reciprocal structure of the optical fiber gyro.
The detection principle of an optical gyroscope is based on the Sagnac effect (Sagnac effect) of relativistic theory. In a closed optical path of the optical gyroscope, two beams of light which are emitted by the same light source and transmitted in a clockwise direction (CW) and a counterclockwise direction (CCW) interfere with each other, and the loop rotation angular velocity of an inertial space reference system is measured by detecting the change of interference fringes. The sagnac effect expression is as follows:
equation (1)
Where ω is the frequency of light, c is the speed of light in vacuum, A is the area enclosed by the light path, and Ω is the angular velocity of rotation of the carrier. The positive and negative of the Sagnac phase are judged by the reference phase. In an Interferometric Fiber Optic Gyroscope (IFOG), a long optical fiber (several hundred meters to several kilometers) needs to be wound into a multi-turn gyro coil. In this case, the sagnac effect can be expressed as:
equation (2)
Wherein L is the length of the optical fiber, D is the diameter of the optical fiber coil, and lambda is the wavelength of the light wave.
According to the theory of interference, the response function of an optical fiber gyroscope is a cosine function I0[1+cos(φs)]. In order to solve the problem of insensitivity of cosine response, sine phase modulation is usually introduced into a fiber optic gyroscope. By applying a phase modulator, e.g. piezo-electric effect modulator or LiNO, to a section of the optical coil3The Y-waveguide, as shown in fig. 2, where the Y-waveguide device incorporates the function of a polarizer. The nonreciprocal modulation of two optical signals is realized by adding a phase modulator at one end of an optical fiber coil (namely an optical fiber ring). The Y waveguide phase modulator makes two light waves receive different phase modulation phi at different timem(t), here, sine modulation is taken as an example, and square wave modulation is not limited to sine modulation. A phase difference is generated as follows
△φ(t)＝φCCW(t)-φCW(t)＝φm(t)-φm(t-τ)
＝φ0sin(ωmt)-φ0sin[ωm(t-τ/2)]Equation (3)
＝2φ0sin(ωmτ/2)cos[ωm(t-τ/2)]
＝φbcos[ωm(t-τ/2)]
Wherein phi0For modulating the amplitude of the signal, phibIs the modulation depth. τ is neffL/c represents the transmission time of light through the entire length of the fiber coil, neffIs the effective refractive index of the fiber. Thus, after modulation is applied, the cosine type interference light intensity signal becomes
Iout＝I0{1+cos[φS+△φ(t)]Equation (4)
At present, the interferometric fiber optic gyroscope mainly has two signal modulation modes, namely sinusoidal phase modulation and square wave phase modulation. Square wave modulation is mainly applied in closed loop detection, and sinusoidal modulation is mainly applied in open loop detection. The invention is mainly discussed based on a sine wave modulated fiber optic gyroscope. However, the method for detecting the open loop of the square wave modulation gyroscope is described in a previous patent (201310392719.X) of the applicant, so that the invention can be applied to the square wave modulation open loop detection optical fiber gyroscope as well.
After modulation, the gyro output signal is expanded into:
equation (5)
Wherein the Sagnac effect is mainly reflected on the first harmonic,
I1(φs)＝2I0J1(φb)sinφsequation (6)
To obtain the desired phase value, φ is first determined from the amplitudes of the 2 nd and 4 th harmonic components of the modulation frequencybI.e. solving the equation
I(4ωm)/I(2ωm)＝J4(φb)/J2(φb) Equation (7)
Then, inverse trigonometric calculation is performed by using the 1, 2-order harmonic wave to obtain the following phase value:
φS＝arctan{I(ωm)J2(φb)/I(2ωm)J1(φb) Equation (8)
The requirements of the interferometric fiber-optic gyroscope on different applications are correspondingly set with different precision levels, and the following table shows the technical requirements of each precision level. The zero bias stability is the detection precision under long-term measurement, and the random walk coefficient is the performance evaluation coefficient under short-term measurement of the gyroscope.
TABLE 1 accuracy class and specification of interferometric fiber optic gyroscopes
However, in the existing fiber-optic gyroscope measurement process, there are temperature Shupe effect, faraday magnetic field effect, light source intensity noise, device thermal noise, and polarization state coupling in the fiber-optic gyroscope and scintillation noise of the detector, which results in long-term fluctuation of data, i.e., zero-bias stability degradation and increase of short-term noise.
For the noise suppression problem of the gyroscope, there are two main solutions. One type is a redundant configuration, namely a plurality of gyros are adopted for simultaneous and same-ground measurement, noise pollution of measurement can be inhibited to a certain degree, however, the scheme increases the complexity and debugging difficulty of the system, and the hardware cost is greatly increased. Another solution is to use kalman filtering algorithm for optimization. However, kalman filtering requires accurate modeling of a system, and meanwhile, a model is susceptible to failure due to environmental changes, so that the kalman filtering faces a problem of unconvergence of processing results in the field of fiber optic gyroscopes greatly affected by environments.
Disclosure of Invention
The invention provides a novel optical fiber gyroscope based on vector space analysis and a signal processing method, aiming at solving the problems that filter time delay and model parameters are difficult to set and the like introduced for suppressing detection noise in the conventional optical fiber gyroscope. The processing method can realize the relevant noise suppression and the signal synchronization optimization of the optical fiber gyroscope system.
The technical scheme adopted by the invention is as follows:
a fiber-optic gyroscope signal processing method based on vector space analysis comprises the following steps:
1) carrying out quadrature balanced local signal demodulation by using an optical fiber gyroscope for open loop detection to obtain two paths of quadrature signals, and simultaneously obtaining one path of expected signals by using a common method;
2) performing preset adaptive filtering on the obtained two paths of orthogonal signals and an expected signal, performing principal component analysis processing on an error signal output by an adaptive filter, and performing linear combination on a response signal output by the adaptive filter to obtain the rotation angular speed of the optical fiber loop;
3) and performing quasi-Kalman filtering based on the signal-to-noise ratio on the error signal and the response signal to obtain the required optimized final angular speed.
In the above-mentioned solution of the present invention, the local orthogonal sine and cosine signal and the square wave modulated or sine wave modulated output signal of the optical fiber gyroscope are correlated, so as to separate and extract the harmonic component in the optical fiber gyroscope signal, thereby reducing the angular velocity. Note that, here, a fiber optic gyro for square wave modulation may be selected, and a fiber optic gyro for sine wave modulation may be selected.
Further, in the step 1), demodulation adopts an orthogonal demodulation method to obtain two paths of detection signals of I and Q, which are respectively used as input ends to be input into the two adaptive filters, and simultaneously, one path of reference signal is obtained according to a traditional method and is used as an expected signal to be input into expected ends of the two adaptive filters.
Further, step 2) carries out averaging processing and principal component analysis processing on the two paths of response signals and the two paths of error signals output by the adaptive filter respectively.
Furthermore, the two processed signals are orthogonal signals, the response value is used as the predicted value of the system model, the error signal is used as the innovation, and the optimized estimation of the system is realized by utilizing the Kalman filtering principle.
In one or more examples of the above aspects, the modulation frequency is a native frequency of the optical fiber gyroscope.
The optical fiber gyroscope adopting the method comprises a light source, a coupler, a phase modulator and an optical fiber ring which are sequentially connected, wherein the coupler is also connected with a photoelectric detector, the phase modulator is also connected with a signal generator, the optical fiber gyroscope is characterized by also comprising an acquisition card which is connected with the photoelectric detector and the phase modulator, and a digital signal processor which is connected with the acquisition card and the signal generator, and the digital signal processor comprises:
the orthogonal demodulation module is used for demodulating the orthogonal balanced local signal to obtain two paths of orthogonal signals and simultaneously obtaining one path of expected signal by adopting a common method;
the adaptive filtering module is connected with the orthogonal demodulation module and is used for carrying out preset adaptive filtering on the two paths of orthogonal signals and the expected signal;
the principal component analysis module is connected with the self-adaptive filtering module and is used for carrying out principal component analysis processing on the error signal output by the self-adaptive filter;
and the Kalman filtering module is connected with the principal component analysis module and the self-adaptive filtering module and is used for carrying out quasi Kalman filtering on the error signals and the response signals based on the signal-to-noise ratio so as to obtain the optimized final angular speed.
Further, the light source is a wide-spectrum light source, and the phase modulator is LiNO3And a Y-shaped waveguide. However, the present invention is not limited to this, and a piezoelectric effect modulator may be used, that is, the present invention is also applicable to a PZT-modulated fiber optic gyroscope.
Compared with the prior art, the invention has the beneficial effects that:
through synchronous orthogonal demodulation, adaptive filtering and principal component analysis, synchronous demodulation and optimization of the interferometric fiber-optic gyroscope for open-loop detection are realized, and noise introduced by asynchronous difference and long delay and complex model estimation in the filtering process in the traditional demodulation method are solved. Compared with the traditional finite step filter, the invention adopts an iterative algorithm with faster throughput, and meanwhile, the self-adaptive filtering does not need to be estimated in advance. More importantly, due to the adoption of vector space analysis, the processing method ensures the orthogonality of data, and further can fully meet the requirement of Kalman filtering, and the condition cannot be ensured in the traditional filter. Compared with the traditional filter, the method has smaller delay effect, does not have extremely high requirements on a prior model of the system like a Kalman filter, only needs a covariance matrix of the system, can realize real-time update by using a register, and eliminates errors possibly brought by environment change. The invention realizes the statistical minimum variance demodulation of the system, simultaneously does not introduce new hardware on the hardware, does not bring delay and divergence on the algorithm, and has obvious advantages and wide application prospect.
Drawings
FIG. 1 is a schematic diagram of a minimum reciprocal structure of a fiber optic gyroscope;
FIG. 2 is a structural diagram of a fiber optic gyroscope incorporating a Y-type waveguide for modulation;
fig. 3 is a process flow chart showing a detection method of the optical fiber gyro according to the embodiment of the present invention;
fig. 4 is a schematic structural diagram of an optical fiber gyro according to an embodiment of the present invention;
FIG. 5 is a flow chart of the digital signal processor internal to FIG. 4;
FIG. 6 is a vector space diagram during signal processing of the method of the present invention;
FIG. 7 is a signal convergence process diagram of the method of the present invention;
FIG. 8 is a graph of Allan standard deviation after treatment according to the method of the present invention.
Detailed Description
The invention is further illustrated by the following specific examples and the accompanying drawings.
Fig. 3 is a schematic diagram of the present invention, fig. 4 is an overall structural diagram according to the present invention, and fig. 5 is a process flow diagram inside the digital signal processor. In the following description, the device structure is based on fig. 4, and the processing is based on fig. 3 and 5. For the output signal of the optical fiber gyroscope modulated by the sine wave, the orthogonal trigonometric function signal is used as a local signal, and the harmonic signal is extracted through relevant demodulation, so that the angular velocity information is obtained.
The quadrature local oscillation signals LO selected are:
equation (9)
Wherein, ω is the modulation frequency, θ is the initial phase, and the local signal contains a constant phase difference of 90 °.
The two paths of orthogonal local signals and the signal output by the gyroscope are respectively sent to a multiplier, and then the average value of the two paths of orthogonal local signals and the signal output by the gyroscope is respectively obtained, so that the in-phase signal and the orthogonal signal can be obtained, and the whole of the two paths of signals contains the amplitude and phase information of the output signal of the optical fiber gyroscope. The initial phase of the local signal is 45 degrees, the invention is to contain sin (phi)s) Carrying out equalization detection demodulation on the first harmonic to obtain a related first harmonic signal orthogonal signal I, Q:
equation (10)
Equation (11)
Wherein I0Is the light source intensity. Phi is asIs the sagnac phase. 2 nd and 4 th harmonics are only used to extract the modulation depth phibAnd only quadrature demodulation is needed without equalization detection. J. the design is a square1Is a first order bessel function. According to the formula, the inverse trigonometric function can be solved to obtain the omega to which the fiber-optic gyroscope is sensitive:
equation (11)
Equation (12)
Wherein J2Is a second order bessel function. The rotation speed signal omega demodulated by different local oscillation frequencies n omeganI.e. the rotational speed information in the different harmonic components, which approximately obey the following relationship:
Ωn～N(Ω,σn) Equation (13)
Where N is a Gaussian distribution, Ω is the mean of the distribution, σnIs the variance value of the distribution. The reason that the channel follows a gaussian distribution is that in a fiber optic gyroscope for a high power broad spectrum light source, the Relative Intensity Noise (RIN) dominates. At present, the power of a mainstream wide-spectrum light source is in the magnitude of 10dBm, and the assumption of Gaussian noise is met.
For the input signal, the view angle used by the present invention is vector space, so we use the sequence of the input signal as a column vector, as follows:
(equation 14)
M＝T,I,Q,IA,QA,IE,QE,IE',QE',KP,KN,K,A,E',S
Wherein T refers to the conventional quadrature demodulation result, and I and Q are quadrature channels; IA and QA are response signals after adaptive filtering; IE and QE are error signals after self-adaptation; IE ', QE' is the error signal after Principal Component Analysis (PCA); KP is a state prediction signal in the classic Kalman filtering; KN is an innovation signal in the classical Kalman filtering; k is a state estimation signal after the classical Kalman filtering; a is a total response signal obtained after IA and QA are averaged; e ' is IE ', QE ' is the averaged total error signal; and S is a final estimation signal obtained after Kalman gain is adopted. These signals all follow a gaussian distribution.
The correlation between signals is the important relationship of the signals, and in the vector space, the correlation is regarded as an included angle and is defined as
Equation (15)
Where ρ is the correlation coefficient and i, j is the direction vector.
As shown in FIG. 5, for the demodulated I, Q channels, the rotation speed values Ω I and ΩQRespectively as input signals, performing adaptive filtering processing, wherein the desired signal of the adaptive filter is the rotation speed value omega obtained by the traditional demodulation modeT。
After passing through the self-adaptive filter, a double path omega is outputIAAnd ΩQASignal as a response signal, and an error signal omegaIEAnd ΩIE. In which the response signal is relatively stationaryGood, but slow response time, the response of the error signal reflects the noisy observation intact, but jitter is large. The conventional low pass filter method inevitably slows down the response time and may introduce detection errors while suppressing the jitter caused by noise. If a recursive structure of kalman filtering is desired, modeling of the system and estimation noise and measurement noise with orthogonality are required, which is difficult to implement in a real complex system, thus affecting the estimation accuracy and possibly causing divergence. The method can take a flat response signal as a prediction signal and an error signal as innovation by utilizing the iterative tracking characteristic of the adaptive filter, and the subsequent vector space analysis shows the orthogonality of the two paths of signals, so that the Kalman gain is adopted for combined estimation.
The response signal expression is:
equation (16)
Equation (17)
The lower subscripts IA and QA refer to the filter output signals, respectively. The adaptive coefficient is adjusted by the formula
Equation (18)
Equation (19)
Where e denotes the filter error signal, which can be obtained by differentiating the reference value and the output value, and the expression is as follows:
equation (20)
Equation (21)
In the signal estimation theory, the requirement of the optimal lower bound (CRB) in the averaging process is that there is no correlation between channels participating in the estimation, so that for error signals containing zero mean values with high correlation, we adopt the PCA method to decorrelate and obtain the optimal estimation of the error signals, i.e. the optimized innovation, according to the relationship between signal included angles in the vector space. For the adaptive response channel with a certain degree of correlation, even if only average processing is adopted, the wandering is also superior to the adaptive filtering with the same order of one-step delay of the traditional demodulation value.
For the original IQ two-path signals, the correlation coefficient is theoretically
Equation (22)
The I channel and the Q channel can be viewed as two paths of signals obtained by projecting the T channel obtained by the traditional method to the two orthogonal axes respectively.
For the response signal after passing the adaptive filtering
Equation (23)
Since the adaptive filtering process can be viewed as a re-projection process onto the desired signal, i.e., the T channel, the relationship between the projected IA and QA channels is the product of the cosine of the included angle.
For adaptively filtered error signals
Equation (24)
Where the direction of the adaptively filtered error signal is perpendicular to the desired signal path when the adaptive filter converges, such that the directions of IE and QE are theoretically the same direction.
After the error signal has been subjected to PCA, it will be orthogonalized again with a correlation coefficient of 0, i.e.
Equation (25)
For the averaged adaptive response signal and the averaged adaptive PCA error signal, the correlation coefficient is
Equation (26)
Since the error signal is already orthogonal to the desired signal and the direction of the response signal is the desired signal direction as the error signal decreases during the adaptive filtered error signal minimization, the subsequent error signal is still orthogonal to the response signal.
For the principle component analysis to eliminate the correlation, the correlation is projected to two new channels with eigenvalues of 1+ rho and 1-rho theoretically through space simultaneous projection. Thus, the relevant parts of the two signals are eliminated in the averaging process. The covariance matrix required for principal component analysis can be obtained by statistical results of multiple tests of the gyroscope, or can be updated in real time through a real-time register in one experiment. Meanwhile, the error signal completely meets the zero mean value required by the principal component analysis.
The specific formula is as follows:
equation (27)
Wherein,a likelihood function representing a vector of error signals;is a 2-dimensional vector error signal composed of quadrature error signals;
equation (27) for the signal ΩIEAnd ΩQEThe probability distribution of the formed error vector satisfies the Gaussian characteristic, rhoIE,QEIs the correlation coefficient of the two. After principal component analysis
Equation (28)
Equation (29)
Where a, b are coefficients used to normalize the feature vector, cijAre the elements in the matrix formed by the normalized eigenvectors. Characteristic values are respectivelyAndthus, the mean post-variance of the new vector after linear transformation is
Equation (30)
Therefore, the invention realizes the optimal lower bound of the error signal obtained after the adaptive filtering by optimizing the error signal.
Averaging the two-path response signals after the self-adaptive filtering to obtain a total response signal
Equation (31)
Equation (32)
Due to the orthogonality of the error channel and the response channel, a Kalman gain coefficient is adopted for estimation. Where the coefficient a is the autocorrelation term, σPAnd (3) referring to system estimation noise obtained by performing principal component analysis after the mean value of an original channel is removed, and obtaining the same result by adopting maximum likelihood estimation based on Gaussian distribution. And K is a stable value of the standard Kalman filtering gain, and is not a direct Kalman gain of a total response channel and a total error channel. This is because this process does not change the physical properties of the system, so the gain factor needs to be still the gain calculated by the standard kalman filter. Meanwhile, if a delay factor is introduced in consideration of the delay effect of the total response channel, a similar effect can be obtained. Therefore, the vector detection and optimization processing of the split-ring gyroscope is realized.
Fig. 4 is a schematic diagram showing a structure of novel signal processing of the optical fiber gyro according to the embodiment of the present invention. Which comprises a wide-spectrum light source, a polarization-maintaining coupler,LiNo3y-type waveguide, 2 kilometer long optical fiber ring, signal generator capable of being controlled by serial port, wherein sampling rate of acquisition card is set to be 2M per second, sampling number is 200000, namely average time is 0.1s, and bandwidth is 10 Hz.
Fig. 6 is a plan view of the optical fiber gyroscope, again after a day-night test, with a smooth 45000 o' clock plot of the quadrature channel signal taken therefrom. The experimental measurement object is the rotational angular velocity of the earth, and the theoretical value to be measured in the laboratory dimension (39.9 degrees north latitude) is 9.67 degrees/hour. The first group, the upper left graph, is the original unprocessed IQ channel with correlation coefficient of approximately 0. The upper right graph is the adaptive response channel with a correlation coefficient of approximately 0.5. The left center is the adaptively filtered non-PCA error signal with a correlation coefficient of approximately 1. The right middle part is an error signal after PCA processing, and the correlation coefficient is 0. The lower left is the estimation signal and innovation signal of traditional Kalman filtering, the correlation coefficient is about-0.5, a is 0.999343, and the state noise is 0.0293[ (°/h)2]Observation noise of 0.0630[ (°/h)2]. The averaged total response channel and the PCA total error channel are shown at the bottom right with correlation coefficients of approximately 0. Here we introduce the traditional kalman filtering method as a comparison with the effect of the present invention. Thus, experimental test results indicate that the correlation fits our previous inference. The detailed parameters have been shown in table 2. Wherein sigma2 c[(°/h)2]Is the error value after combining the corresponding IQ channels.
TABLE 2 noise and correlation coefficient for each channel
In the experiment, the algorithm of the adaptive filter is a Least Mean Square (LMS) algorithm, the order is 40, and the step length is 0.00001. The invention is not limited in this regard and the step size and order of the adaptive filter may be varied as desired. The algorithm may employ other RLS or gradient descent or other convergence methods. At the same time, the total response averaged by the I, Q response channels of the adaptive filterVariance of 0.0028[ (°/h)2]Less than 0.0029[ (°/h) from one step delay of a conventional channel at the same filter setting2]。
Fig. 7 is a detailed view of the convergence of the present processing method at the initial time, and the data length is 500 points. At the data 0 point, the original data omega are sequentially arranged from high to lowTError signal Ω 'after PCA processing'EFinally, the signal Ω is estimatedSTotal response value omega after adaptive filteringA. Therefore, the real-time convergence effect of the estimated signal is influenced by the step length and the order of the self-adaptive filter set by experiments, and the convergence speed can be adjusted according to actual needs. The convergence rate can be considered as the response rate of the filter or the gyro tracking capability. At the same time, the variance of the total response averaged by the adaptive filter I, Q response channels is 0.0028[ (°/h)2]Less than 0.0029[ (°/h) from one step delay of a conventional channel at the same filter setting2]。
Fig. 8 is a comparative graph obtained by performing an Allan standard deviation analysis on the data of fig. 6. The initial part is the original data omega from high to lowTThe conventional kalman filter signal ΩKAnd the estimated value omega of the present inventionS. It can be seen that the principal component analysis processing method can improve short-time parameters of the system, such as wandering and quantization, and the like, without affecting the zero offset stability of the system, and the new method realizes noise suppression under the condition of ensuring the quick tracking response capability. Specific Random Walk Coefficients (RWC) and zero bias stability are shown in table 3.
TABLE 3 Allan standard deviation coefficients for raw and processed data
The method has the characteristic of improving the system variance, so that the random walk parameters are improved, and the reasonability of the gyroscope signal analysis is also described.
While the foregoing disclosure shows illustrative embodiments of the invention, it should be noted that various changes and modifications could be made herein without departing from the scope of the invention as defined by the appended claims. The functions, steps and/or actions of the method claims in accordance with the inventive embodiments described herein need not be performed in any particular order. Furthermore, although elements of the invention may be described or claimed in the singular, the plural is contemplated unless limitation to the singular is explicitly stated.
Although the present invention has been disclosed in connection with the preferred embodiments thereof as shown and described in detail, it will be understood by those skilled in the art that various modifications may be made in the method and apparatus for vector probing a fiber optic gyroscope herein before described without departing from the spirit of the invention. Therefore, the scope of the present invention should be determined by the contents of the appended claims.
Claims (7)
1. A fiber-optic gyroscope signal processing method based on vector space analysis comprises the following steps:
1) carrying out quadrature balanced local signal demodulation by using an optical fiber gyroscope for open loop detection to obtain two paths of quadrature signals, and inputting the two paths of quadrature signals into two adaptive filters as input ends respectively; meanwhile, a common method is adopted to obtain a path of expected signal which is input to the expected ends of the two adaptive filters;
2) performing preset adaptive filtering on the obtained two paths of orthogonal signals and the expected signal by adopting the two adaptive filters, performing principal component analysis processing on an error signal output by the adaptive filters, and performing linear combination on response signals output by the adaptive filters to obtain the rotation angular speed of the optical fiber loop;
3) and performing quasi-Kalman filtering based on the signal-to-noise ratio on the error signal and the response signal to obtain the optimized final angular speed.
2. The method of claim 1, wherein: step 1) processing a local orthogonal sine and cosine signal and a square wave modulated or sine wave modulated optical fiber gyroscope output signal to obtain two paths of orthogonal signals.
3. The method of claim 1, wherein: and 2) carrying out averaging processing and principal component analysis processing on the two paths of response signals and the two paths of error signals output by the adaptive filter respectively.
4. The method of claim 1, wherein: and (3) carrying out the covariance matrix of the principal component analysis in the step 2) by adopting the statistical result of multiple tests of a gyroscope, or carrying out real-time updating in one experiment through a real-time register.
5. The method of claim 1, wherein: and 3) taking the response signal as a prediction signal, taking the error signal as an innovation, and realizing optimized estimation by utilizing a Kalman filtering principle.
6. An optical fiber gyroscope adopting the method of claim 1, comprising a light source, a coupler, a phase modulator and an optical fiber ring connected in sequence, wherein the coupler is further connected with a photodetector, the phase modulator is further connected with a signal generator, the optical fiber gyroscope is characterized by further comprising an acquisition card connecting the photodetector and the phase modulator, and a digital signal processor connecting the acquisition card and the signal generator, and the digital signal processor comprises:
the orthogonal demodulation module is used for demodulating the orthogonal balanced local signal to obtain two paths of orthogonal signals and simultaneously obtaining one path of expected signal by adopting a common method;
the adaptive filtering module is connected with the orthogonal demodulation module and is used for carrying out preset adaptive filtering on the two paths of orthogonal signals and the expected signal;
the principal component analysis module is connected with the self-adaptive filtering module and is used for carrying out principal component analysis processing on the error signal output by the self-adaptive filter;
and the Kalman filtering module is connected with the principal component analysis module and the self-adaptive filtering module and is used for carrying out quasi Kalman filtering on the error signals and the response signals based on the signal-to-noise ratio so as to obtain the optimized final angular speed.
7. The optical fiber gyroscope of claim 6, wherein: the light source is a wide-spectrum light source, and the phase modulator is LiNO3Y-waveguides or piezoelectric effect modulators.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410174900.8A CN103983261B (en) | 2014-04-28 | 2014-04-28 | A fiber optic gyroscope based on vector space analysis and a signal processing method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410174900.8A CN103983261B (en) | 2014-04-28 | 2014-04-28 | A fiber optic gyroscope based on vector space analysis and a signal processing method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103983261A CN103983261A (en) | 2014-08-13 |
| CN103983261B true CN103983261B (en) | 2017-02-15 |
Family
ID=51275326
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410174900.8A Expired - Fee Related CN103983261B (en) | 2014-04-28 | 2014-04-28 | A fiber optic gyroscope based on vector space analysis and a signal processing method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103983261B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106404016A (en) * | 2016-08-29 | 2017-02-15 | 深圳艾瑞斯通技术有限公司 | Fiber sampling signal filtering method and device and fiber sensing system |
| CN106771600B (en) * | 2017-01-07 | 2019-07-30 | 中国人民解放军陆军勤务学院 | Adaptive unbiased phase difference estimation method under the conditions of frequency mismatch |
| CN107727113B (en) * | 2017-09-13 | 2020-06-23 | 北京小鸟看看科技有限公司 | Method and device for testing gyroscope in equipment |
| US10746550B2 (en) | 2018-08-01 | 2020-08-18 | Honeywell International Inc. | Resonator fiber-optic gyro with quadrature error reducer |
| CN109084743B (en) * | 2018-08-24 | 2022-04-19 | 中国科学院光电技术研究所 | Method for separating target information output by fiber-optic gyroscope and disturbance signal of photoelectric tracking system |
| CN115420272B (en) * | 2022-08-24 | 2024-04-26 | 北京航空航天大学 | Method for realizing adaptive suppression of relative intensity noise of optical fiber gyro light source |
| CN117109450B (en) * | 2023-07-25 | 2024-05-10 | 广东工业大学 | High-resolution spectrum confocal interference film thickness measuring method for self-adaptive parameter spectrum estimation |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101221046A (en) * | 2008-01-22 | 2008-07-16 | 南京航空航天大学 | Error processing method for output signal of optic fiber gyroscope component |
| CN102508278A (en) * | 2011-11-28 | 2012-06-20 | 北京航空航天大学 | Adaptive filtering method based on observation noise covariance matrix estimation |
| CN102607547A (en) * | 2012-02-17 | 2012-07-25 | 北京大学 | Digital domain balanced detecting method and device for optical fiber gyroscope |
| CN103471578A (en) * | 2013-09-02 | 2013-12-25 | 北京大学 | Orthogonal detection-based measuring method using multi-dimensional signal of optical fiber gyroscope |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS5958886A (en) * | 1982-09-28 | 1984-04-04 | Fujitsu Ltd | Single mode semiconductor laser diode |
| JPH0658228B2 (en) * | 1987-04-07 | 1994-08-03 | 株式会社日立製作所 | Light fiber gyro |
-
2014
- 2014-04-28 CN CN201410174900.8A patent/CN103983261B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101221046A (en) * | 2008-01-22 | 2008-07-16 | 南京航空航天大学 | Error processing method for output signal of optic fiber gyroscope component |
| CN102508278A (en) * | 2011-11-28 | 2012-06-20 | 北京航空航天大学 | Adaptive filtering method based on observation noise covariance matrix estimation |
| CN102607547A (en) * | 2012-02-17 | 2012-07-25 | 北京大学 | Digital domain balanced detecting method and device for optical fiber gyroscope |
| CN103471578A (en) * | 2013-09-02 | 2013-12-25 | 北京大学 | Orthogonal detection-based measuring method using multi-dimensional signal of optical fiber gyroscope |
Non-Patent Citations (1)
| Title |
|---|
| 基于LMS的光纤陀螺数据处理;杨峰 等;《重庆工学院学报(自然科学)》;20091031;第23卷(第10期);第2节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103983261A (en) | 2014-08-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103983261B (en) | A fiber optic gyroscope based on vector space analysis and a signal processing method thereof | |
| Lefevre | The fiber-optic gyroscope | |
| Culshaw | The optical fibre Sagnac interferometer: an overview of its principles and applications | |
| CN108534798B (en) | Polarization nonreciprocal error elimination method in dual-polarization fiber-optic gyroscope and dual-polarization fiber-optic gyroscope | |
| WO2021135171A1 (en) | Multi-phase modulation and demodulation-based fiber-optic gyroscope multi-closed-loop method | |
| JP6404656B2 (en) | Method and apparatus for tracking / rocking the free spectral range of a resonator and its application to a resonator fiber optic gyroscope | |
| US8085407B2 (en) | Resonator optical gyroscope having input beam modulation optimized for high sensitivity and low bias | |
| Paturel et al. | One nautical mile per month FOG-based strapdown inertial navigation system: a dream already within reach? | |
| US20160146607A1 (en) | Dynamically monitoring the instantaneous zero rotation rate voltage of interferometric fiber optic gyroscope (ifog) | |
| EP2226612B1 (en) | Cavity length modulation in resonator fiber optic gyroscopes | |
| CN103471578A (en) | Orthogonal detection-based measuring method using multi-dimensional signal of optical fiber gyroscope | |
| NO340647B1 (en) | Method and power assisted optical fiber measurement device and gyro | |
| CN102607589A (en) | Method and device for measuring angular speed of optical fiber gyroscope based on double-frequency modulating signals | |
| Meshkovsky et al. | Three-axis fiber-optic gyroscope: Development and test results | |
| US20150025856A1 (en) | Simulator for simulating the operation of a fiber optic gyroscope | |
| Pei et al. | Initial Self‐Alignment for Marine Rotary SINS Using Novel Adaptive Kalman Filter | |
| CN104729493A (en) | Novel detection method of optical fiber gyroscope | |
| CN113587914B (en) | Spectrum separation method for inhibiting back reflection error of resonant integrated optical gyroscope | |
| Celikel | Application of the vector modulation method to the north finder capability gyroscope as a directional sensor | |
| TWI448662B (en) | Optical interferometric apparatus | |
| Zhang et al. | Fiber optic gyroscope dynamic north-finder algorithm modeling and analysis based on Simulink | |
| Chamoun | A laser-driven fiber optic gyroscope for inertial navigation of aircraft | |
| Wang et al. | Lock‐In Amplifier Technology in Laser Gyroscope North Finder of Constant Rate Biasing | |
| CN113804175A (en) | Dual-polarization interference type optical fiber gyroscope | |
| Nasiri-Avanaki et al. | Comparative Assessment on the performance of Open-loop and Closed-loop IFOGs |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170215 Termination date: 20190428 |