A vibration compensation method for correcting interference fringes of atomic interferometer gravimeter
By using the sweep rate sequence and optimization target to calculate the optimal delay coefficient and gain coefficient in the atomic gravity meter, the problem of the vibration compensation effect of the atomic gravity meter deteriorates when the environment changes is changed, and the measurement accuracy is improved.
Patent Information
- Application Number
- CN202210161467.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-02-22
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-02-22
AI Technical Summary
The vibration compensation effect of existing atomic gravity meters decreases relatively when the measurement environment changes, affecting the measurement accuracy.
Using a method that includes a given sweep rate sequence, gravity reference value, pulse interval duration and effective wave vector, by importing actual measured data, calculating and setting optimization targets, segmenting vibration sensor signals, traversal search method or golden segmentation search method Find the optimal delay coefficient and optimal gain coefficient, calculate the vibration speed of the reference mirror and obtain the vibration phase noise sequence, correct the fringes, and finally calculate the RMSE value to achieve the best compensation effect.
The best vibration compensation effect is achieved in different measurement environments, which improves the measurement accuracy of the atomic gravity meter and solves the problem of degradation of vibration compensation effect caused by environmental changes.
Smart Images

Figure CN114527520B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of gravity measurement, and in particular to a vibration compensation method for correcting interference fringes of an atomic interferometer gravimeter. Background Art
[0002] Absolute gravity measurement, that is, the precise measurement of the absolute value of gravitational acceleration, has a wide range of applications in assisted navigation, resource exploration, geophysics, and metrology. Existing high-precision absolute gravimeters generally use the free fall method, mainly including laser interferometer absolute gravimeters with corner cube prisms as falling objects and atomic interferometer absolute gravimeters with cold atomic clusters as falling objects, hereinafter referred to as optical gravimeters and atomic gravimeters, respectively. The former measures the motion trajectory of the corner cube prism falling freely in a vacuum through a laser interferometer, and solves the gravitational acceleration by fitting, while the latter calculates the gravitational acceleration based on the atomic interference fringes formed by the atomic clusters during the free fall process after three Raman laser pulses. Since the actual measurement objects of the two are the motion acceleration of the falling object or the falling atomic cluster relative to a reference mirror (corner cube prism or plane mirror) in the instrument, the acceleration of the reference mirror itself must be coupled to the measurement result. Theoretically, when the vibration of the reference mirror itself is ignored, the measurement accuracy of the existing absolute gravimeter can reach the micro-Gal level; but in reality, if the reference mirror is placed directly on the ground, the measured value dispersion caused by the ground vibration will exceed ten micro-Gal levels, and even reach the milligal or ten milligal level in a complex vibration environment. Therefore, using a vibration measuring instrument to obtain the motion acceleration of the reference mirror itself and correcting its influence when calculating the gravity acceleration is one of the effective means to improve the measurement accuracy of the absolute gravimeter. This correction method is generally called vibration compensation.
[0003] Specifically, the ground vibration noise that affects the atomic gravimeter is mainly environmental noise, which is related to the foundation, coordinates and environment of the measurement time and measurement point, and can be roughly distinguished according to the frequency. The first is human activity noise, which is generally greater than 1Hz, such as the movement of people nearby during instrument testing, vibrations generated by artificial vibration sources such as vehicles, and shaking of buildings. The second is the noise from the earth itself, which is generally between 0.1Hz and 1Hz, mainly including periodic ground pulsations and vibrations caused by internal movements of the earth such as earthquakes. The acceleration power spectrum density of the ground pulsation noise model has peaks at frequencies of 0.2Hz and 3Hz. Finally, there is atmospheric movement noise such as air pressure fluctuations, which is generally less than 0.1Hz, generally changes slowly and has a low amplitude, and the impact on absolute gravity measurement can be ignored. Since absolute gravity measurement only relies on vertical displacement data for calculation, when performing vibration compensation for atomic interference fringes, the horizontal component of ground vibration is generally ignored, and only the vertical vibration signal output by the vibration measuring instrument is used.
[0004] The most mature atomic gravimeter in the world is the CAG-01 atomic gravimeter developed by the Paris Observatory in France. It has participated in many international gravity comparisons. The vibration processing method adopted is a combination of a passive vertical vibration isolation system and vibration compensation. The specific principle of vibration compensation is as follows: the data acquisition card synchronously records the velocity signal and atomic interference fringes output by the seismometer during the atomic interference process. In the control software, the seismometer signal is processed using an IIR (Infinite Impulse Response) filter and a non-causal low-pass filter to reduce the influence of the inconstant amplitude-frequency characteristic and nonlinear phase-frequency characteristic of the seismometer transfer function, and obtain a more realistic reference mirror vibration signal. Finally, the processed signal is used to obtain the phase noise introduced by the reference mirror vibration and correct the atomic interference fringes. In the measurement environment where the instrument is located, this vibration compensation method itself can increase the sensitivity of the gravimeter by three times. The University of Hannover in Germany proposed to use digital low-pass and high-pass filters to pre-process the output signal of the vibration sensor to obtain a more accurate phase noise introduced by the reference mirror vibration, thereby achieving vibration compensation. The experimental results show that the sensitivity of the gravimeter can be effectively improved by using high-precision commercial seismometers for vibration compensation in a quieter environment, or by using lower-precision commercial accelerometers or their independently developed new optical inertial sensors for compensation in a complex vibration environment.
[0005] In my country, the atomic gravimeter made by Zhejiang University also adopts the vibration compensation method. Its main feature is to use commercial inversion software to correct the influence of the measured transfer function of the seismometer to obtain a more realistic reference mirror vibration signal (such as Figure 1As shown). In its laboratory environment, this method can improve the sensitivity of the gravimeter to a level similar to that when using a passive vibration isolation system. The National University of Defense Technology conducted a detailed analysis of the measurement noise introduced by the vertical vibration and horizontal deflection of the reference mirror, used an accelerometer to measure the vibration of the reference mirror when the atomic gravimeter was working, and calibrated the transfer function between the accelerometer and the reference mirror more accurately. The measured results show that its vibration compensation method can achieve significant vibration compensation effects in a more complex vibration environment. The vibration compensation method proposed by the National Defense Innovation Institute of the Academy of Military Sciences uses a Michelson laser interferometer with a high-precision four-channel phase shift detector to replace the traditional vibration sensor. The interferometer can directly measure the phase shift of the atomic interference fringes, and combine signal shaping, digital phase detection and other methods to determine the change in the phase decimal of the atomic interference fringes before and after the Raman laser pulse, and obtain the atomic interference phase deviation caused by the vibration of the reference mirror. By removing this phase deviation, the transition probability signal measured by the gravimeter can be corrected. In addition, the China Institute of Metrology, the Institute of Precision Measurement of the Chinese Academy of Sciences, Huazhong University of Science and Technology, University of Science and Technology of China, National University of Defense Technology, Stanford University in the United States, Humboldt University of Germany and other institutions have also developed different types of vibration isolation systems to reduce the impact of ground vibration on atomic gravimeters.
[0006] However, the existing atomic gravimeters have poor adaptability. When the measurement environment changes, their vibration compensation effect is significantly reduced, which affects the measurement accuracy of the atomic gravimeter. Therefore, it is necessary to improve them. Summary of the invention
[0007] The object of the present invention is to provide a vibration compensation method for correcting interference fringes of an atomic interferometer gravimeter, which has the best compensation effect in various measurement environments, in view of the above problems.
[0008] In order to achieve the above object, the technical solution of the present invention is:
[0009] A vibration compensation method for correcting interference fringes of an atomic interferometer gravimeter comprises the following steps:
[0010] S1, given sweep rate sequence, gravity reference value, pulse interval duration, effective wave vector;
[0011] S2, import measured data;
[0012] S3, calculating and setting optimization targets through original stripes;
[0013] S4, dividing the vibration sensor signal according to the trigger signal to obtain the vibration sensor signal corresponding to each transition probability;
[0014] S5, setting the starting range of the delay coefficient and the gain coefficient;
[0015] S6. Obtaining the optimal delay coefficient and the optimal gain coefficient by using a traversal search method or a golden section search method;
[0016] S7. The vibration speed of the reference mirror during each atomic interference is calculated through the optimal delay coefficient and the optimal gain coefficient, and then the vibration phase noise sequence, the atomic phase sequence only affected by the vibration, and the correction fringes are obtained. Finally, the RMSE value when the correction fringes are fitted to the standard cosine signal is calculated.
[0017] Furthermore, in step S2, the measured data includes a measured transition probability sequence corresponding to the frequency sweep rate sequence, a vertical output signal of a vibration sensor for measuring the vibration of a reference mirror during the same period, and a trigger signal before each atomic cluster falls during the same period.
[0018] Furthermore, in step S3, the optimization target is one of the RMSE value when the original fringe is fitted, the coefficient of determination when the original fringe is fitted, the standard deviation of the interference fringe residual, and the correlation coefficient between the original fringe residual and the fringe residual affected only by vibration.
[0019] Furthermore, in step S4, a filter processing operation is performed on the vibration sensor signal using a corrector transfer function.
[0020] Furthermore, in step S6, when the traversal search method or the golden section search method is used to obtain the optimal delay coefficient and the optimal gain coefficient, when the optimization target is the RMSE value of the original fringe fitting, the delay coefficient and the gain coefficient corresponding to the minimum RMSE value of the original fringe fitting are used as the optimal delay coefficient and the optimal gain coefficient;
[0021] When the optimization target is the coefficient of determination during the original fringe fitting, the delay coefficient and gain coefficient corresponding to the minimum value of the coefficient of determination during the original fringe fitting are taken as the optimal delay coefficient and optimal gain coefficient;
[0022] When the optimization target is the standard deviation of the interference fringe residual, the delay coefficient and gain coefficient corresponding to the minimum value of the standard deviation of the interference fringe residual are taken as the optimal delay coefficient and optimal gain coefficient;
[0023] When the optimization target is the correlation coefficient between the original fringe residual and the fringe residual affected only by vibration, the delay coefficient corresponding to the time when the absolute value of the correlation coefficient between the original fringe residual and the fringe residual affected only by vibration reaches the maximum value is taken as the optimal delay coefficient; the gain coefficient corresponding to the time when other optimization targets reach the minimum value is taken as the optimal gain coefficient.
[0024] Furthermore, in step S7, when the reference mirror vibration speed is calculated for each atomic interference by using the optimal delay coefficient and the optimal gain coefficient, when the vibration sensor is a seismic timing, the calculation formula is:
[0025]
[0026] In the formula, v m (t) is the vibration speed of the reference mirror; τ is the optimal delay coefficient; K is the optimal gain coefficient; U s (t) is the output signal of the seismometer; v s is the vibration velocity of the sensitive mass of the seismometer; K s is the nominal sensitivity of the seismometer;
[0027] When the vibration sensor is an accelerometer, the calculation formula is:
[0028]
[0029] In the formula, v m (t) is the vibration speed of the reference mirror; τ is the optimal delay coefficient; K is the optimal gain coefficient; K a is the nominal sensitivity of the accelerometer; U a (t) is the accelerometer output signal.
[0030] Furthermore, in step S7, when the vibration speed of the reference mirror during each atomic interference is calculated by using the optimal delay coefficient and the optimal gain coefficient, when the vibration sensor is an accelerometer, a half-period integration algorithm is used to calculate the vibration speed of the reference mirror.
[0031] Compared with the prior art, the present invention has the following advantages and positive effects:
[0032] The present invention has excellent adaptability to the environment and can achieve the best vibration compensation effect at that time and place when the measurement environment changes, solving the problem of relatively decreased vibration compensation effect of the existing vibration compensation method when the measurement environment changes, and effectively improving the vibration compensation effect of the atomic gravimeter in a vibration environment, thereby improving the measurement accuracy of the atomic gravimeter, and further making a certain contribution to the measurement of absolute gravity. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0034] Figure 1 This is the flow chart of the vibration compensation method of Zhejiang University of Technology;
[0035] Figure 2 Schematic diagram of atomic interference fringes;
[0036] Figure 3 It is a frame relationship diagram of the seismometer output signal and the real vibration of the reference mirror;
[0037] Figure 4 This is a flow chart of the vibration compensation algorithm of Example 1;
[0038] Figure 5 This is a flow chart of a vibration compensation algorithm of Example 2;
[0039] Figure 6 This is a flow chart of the vibration compensation algorithm of Example 10. DETAILED DESCRIPTION
[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work, any modifications, equivalent substitutions, improvements, etc., should be included in the protection scope of the present invention.
[0041] 1. General principle of vibration compensation for interference fringes of atomic interferometer gravimeter
[0042] The atomic gravimeter is based on the principle of cold atomic matter wave interference. It applies laser pulses to cold atomic clusters and realizes the atomic interference process through two-photon stimulated Raman transition or multi-photon Bragg diffraction process. Taking the atomic gravimeter based on stimulated Raman transition as an example, the cold atomic cluster falls freely in the vacuum cavity. The plane reference mirror below the vacuum cavity reflects the vertically downward Raman laser along the original path, so that two opposing laser beams act on the atomic cluster at the same time. In this process, three consecutive laser pulses realize the splitting, inversion and combination of atomic wave packets, which are the first π / 2 pulse, π pulse and the second π / 2 pulse respectively. The time interval between adjacent pulses is T. After the combination, atoms in any energy state are composed of the superposition of atoms on two paths. The transition probability P of the atomic cluster after the pulse can be detected through the fluorescence signal, that is, the interference fringes, which satisfy
[0043]
[0044] Among them, ΔФ is the atomic phase, which is related to the local gravitational acceleration g; the coefficients A and B representing the fringe contrast can be obtained by cosine fitting the interference fringes. Ideally, the atomic phase ΔФ is equal to its theoretical value ΔФ th ,Right now
[0045] ΔΦ th =(k eff g-α)T 2 . (2)
[0046] Among them, k eff is the effective wave vector, and α represents the sweep rate of the laser. The control software of the gravimeter adjusts the linear increase of α in equation (2), and the atomic phase ΔФ will also change linearly, making the interference fringes appear in the form of standard cosine signals. At this time, P should be consistent with the standard signal P fitted by P fit Exactly the same,
[0047]
[0048] But in reality, affected by various phase noises, the atomic phase ΔФ is equal to its actual value ΔФ m ,have
[0049]
[0050] Among them, Δφ vib is the phase noise introduced by the reference mirror vibration, hereinafter referred to as vibration phase; Δφothers is the sum of the phase noise introduced by other noise sources, hereinafter referred to as other phase noise. At this time, P is composed of P fit becomes
[0051]
[0052] When the vibration phase dominates the phase noise, the vibration compensation process of the modified interference fringe P can be determined according to the above principle.
[0053] The first step is to measure the original measured fringe P(ΔФ th ) to perform cosine fitting (Note: the stripe is the curve, the vertical coordinate of the curve is outside the brackets, and the horizontal coordinate of the curve is inside the brackets), and the fitted theoretical stripe P is obtained. fit (ΔФ th ) and the coefficients A and B in equation (1).
[0054] The second step is to obtain the actual vibration speed v of the reference mirror within a period of time t through the vibration sensor. m , and then according to
[0055]
[0056] Calculate the actual vibration phase Δφ vib , where S(t) is the sensitivity function of the gravimeter, T1 and T3 are the action times of the first π / 2 pulse and the second π / 2 pulse, respectively, both of which are determined by the known parameters of the gravimeter control software. It should also be noted that although the gravity value g is an unknown quantity to be measured, the approximate rough value g can be used in the initial stage of actual measurement. th Substitute into equation (2) to calculate the theoretical atomic phase ΔФ thTo obtain the fitting fringe P fit , and then obtain the accurate measured value g through subsequent measurement and signal processing.
[0057] The third step is to calculate the atomic phase ΔФ when it is only affected by vibration. v and transition probability P v ,Right now
[0058]
[0059] At this time, re-calculate ΔФ v The data obtained by taking P as the horizontal axis and P as the vertical axis is P(ΔФ v ) is the corrected interference fringe, and the compensation process ends here. Figure 2 As shown, the original fringe P(ΔФ th ) compared to the modified fringe P(ΔФ v ) has moved a little, and theoretically should be closer to the standard cosine signal, that is, the fitted theoretical fringe P fit (ΔФ th That is to say, for the corrected stripe P(ΔФ v )The goodness of fit when performing cosine fitting again should be higher than that of the original fringe P(ΔФ th ) is used to measure the goodness of fit.
[0060] 2. Specific principles of the vibration compensation method described in the present invention
[0061] The method for implementing the above vibration compensation process based on the simplified transfer function model includes two aspects: hardware and software.
[0062] In terms of hardware, an independent seismometer or accelerometer is placed next to or below the reference mirror to measure its vibration. Taking the seismometer as an example, due to the limitations of the mechanical structure and the performance of the seismometer itself, the seismometer output signal U s (t) cannot fully and truly reflect the vibration speed v of the reference mirror m (t), the relationship between the two is as follows Figure 3 As shown, G a represents the transfer function from the ground vibration velocity to the reference mirror vibration velocity, G b represents the transfer function from the ground vibration velocity to the vibration velocity of the seismometer sensitive mass, G c G represents the transfer function from the velocity of the sensitive mass to the output voltage of the seismometer. Since there must be horizontal and vertical differences between the actual position of the sensitive mass inside the seismometer and the reference mirror, G a With G b cannot be equal, and the two transfer functions themselves may also change when the foundation material changes.
[0063] On the other hand, the transfer function of the seismometer itself G c It is not an ideal transfer function with constant amplitude-frequency characteristics and linear phase-frequency characteristics, and it may change with the measurement time and environment (this is because environmental parameters such as temperature, humidity, and air pressure will affect the parameters of the electromechanical modules inside the seismometer, and these parameters may drift over time). In summary, the transfer function H between the true vibration velocity of the reference mirror and the voltage signal output by the seismometer satisfies
[0064]
[0065] Although in reality G a , G b , G c It is difficult to measure and may change, which makes it impossible for surveyors to obtain the exact value of the total transfer function H. However, since the measurement environment will not change significantly during a single gravity measurement task and the measurement time generally does not exceed 24 hours, a simplified model consisting of a delay coefficient τ and a gain coefficient K can be used to estimate H, and it is assumed that the values of these two coefficients remain unchanged during the current measurement process. Figure 3 The simplified model is
[0066]
[0067] Among them, K s is the nominal sensitivity of the seismometer, provided by the manufacturer.
[0068] Theoretically, when the delay coefficient τ and the gain coefficient K are closest to the true value, the output voltage U from the seismometer is s , the reference mirror vibration velocity v calculated according to formula (9) m It is also closest to the true value and has the best correction effect on the stripes. Therefore, the specific idea of vibration compensation is: given the value range of τ and K, select a parameter value that represents the goodness of fit as the optimization target, and find the optimal delay coefficient τ corresponding to the optimal value of the parameter. opt and the optimal gain coefficient K opt After the search is completed, set the delay to τ opt , gain is K opt In the case of v ), recalculate the parameter value used as the optimization target.
[0069] Since the transfer function of the seismometer itself is G c It is not an ideal form, that is, the amplitude-frequency characteristic is not constant and the phase-frequency characteristic is nonlinear. Therefore, the output signal U s (t) may be distorted. Therefore, before searching for the delay factor τ and the gain factor K, a correction filter F can be used to correct the signal Us (t) is processed to reduce the degree of distortion.
[0070] When the hardware part is replaced by the seismometer, the output signal U of the accelerometer can be directly a (t) is integrated once, and the reference mirror vibration signal v is obtained by combining the delay coefficient and the gain coefficient. m (t).
[0071] The corresponding calculation formula is
[0072]
[0073] Among them, K a is the nominal sensitivity of the accelerometer, provided by the manufacturer; the integration time is set according to actual needs, at least including the entire time period of free fall of the atomic group. Other integration algorithms such as half-period integration can also be used to obtain the accelerometer signal U a (t) Calculate the reference mirror velocity signal v m (t).
[0074] 3. Embodiment
[0075] (1) Embodiment 1
[0076] A vibration compensation method for correcting interference fringes of an atomic interferometer gravimeter, characterized in that it comprises the following steps:
[0077] S1, given sweep rate sequence, gravity reference value, pulse interval duration, effective wave vector;
[0078] S2, importing the measured transition probability sequence corresponding to the frequency sweep rate sequence, the vertical output signal of the vibration sensor measuring the vibration of the reference mirror in the same period, and the trigger signal before each atomic group falls in the same period;
[0079] S3, calculating and setting the RMSE value of the original fringe fitting to the optimization target;
[0080] S4, dividing the seismometer signal according to the trigger signal to obtain the seismometer signal corresponding to each transition probability;
[0081] S5, setting the starting range of the delay coefficient and the gain coefficient;
[0082] S6. Obtain the optimal delay coefficient and the optimal gain coefficient by traversal search method;
[0083] S7. The vibration speed of the reference mirror during each atomic interference is calculated through the optimal delay coefficient and the optimal gain coefficient, and then the vibration phase noise sequence, the atomic phase sequence only affected by the vibration, and the correction fringes are obtained. Finally, the RMSE value when the correction fringes are fitted to the standard cosine signal is calculated.
[0084] In this embodiment, the hardware sensor is a seismometer, and the output signal is U s (t). Taking the RMSE value (Root Mean Square Error) when fitting the stripes as the optimization target, first search for the delay coefficient τ, and then search for the gain coefficient K. The delay coefficient and gain coefficient corresponding to the minimum RMSE value are taken as τ. opt and K opt The specific algorithm flow is as follows: Figure 4 shown.
[0085] (2) Embodiment 2
[0086] The steps are the same as those in Example 1, and the RMSE value when fitting fringes is still used as the optimization target. Example 2 uses the golden section method to simultaneously search for the delay coefficient τ and the gain coefficient K. The specific algorithm flow is as follows Figure 5 shown.
[0087] (3) Embodiment 3
[0088] right Figure 4 Seismometer signal U after "data preprocessing" s (t) Add a correction step, that is, filter the signal with a set corrector transfer function F. The specific filtering function can be independently written or use the function library of the software itself. Then execute the process of "traversing and searching for optimal parameters" and "calculating results" in Example 1 on the corrected signal.
[0089] (4) Embodiment 4
[0090] The R-square (coefficient of determination) when fitting fringes is taken as the optimization target, and the delay coefficient and gain coefficient corresponding to the minimum R-square value are taken as τ opt and K opt The method for searching the delay coefficient and the gain coefficient K is the same as that in Embodiment 1, that is, traversal search.
[0091] (5) Embodiment 5
[0092] The R-square (coefficient of determination) when fitting the stripes is taken as the optimization target, and the delay coefficient and gain coefficient corresponding to the minimum value of R-square are taken as τ opt and K opt The search delay coefficient and the search gain coefficient K are the same as those in Embodiment 2, that is, the golden section method is adopted.
[0093] (6) Embodiment 6
[0094] The RMSE value has a simple and intuitive mathematical meaning, but its physical meaning is not obvious. Therefore, another variable with more physical meaning can be used as the optimization target, that is, the change of the interference fringe fitting residual. The steps of searching the coefficients are the same as those in Example 1, which is still to first traverse and search the delay coefficient τ, and then traverse and search the gain coefficient K.
[0095] Before running the vibration compensation algorithm, calculate the original fringe P(ΔФ th ) compared with the theoretical fringe P fit (ΔФ th ) m (ΔФ th ) m When running the vibration compensation algorithm, given the current delay coefficient τ, the interference fringes P only affected by vibration are obtained according to equations (6), (7), and (9). v Compared with the theoretical fringe P fit The residual R v At the same time, record the original stripe P and the stripe P only affected by vibration v The difference is R c Obviously R c That is, the original fringe residual R m The fringe residual R is only affected by vibration. v The difference between Δφ and Δφ is equivalent to the fringe residual after removing the vibration effect. Theoretically, the better the vibration compensation effect, the smaller the vibration phase Δφ. vib The more accurate the calculation, the smaller the fringe residual R v The closer it is to the original stripe residual R m At this time R c will approach the value consisting only of the other phase noise Δφ others The introduced fringe residual, its standard deviation σ c should also tend to the minimum value. Therefore, Example 6 sets σ c Set as the optimization target, with σ c When the minimum value is reached, the corresponding delay coefficient and gain coefficient are τ opt and K opt The search method for the delay coefficient and the gain coefficient is the same as that in Example 1, that is, they are searched separately.
[0096] (7) Embodiment 7
[0097] The interference fringe residual standard deviation σ in Example 6 is c To optimize the target, the golden section method in Example 2 is used to search for the delay coefficient τ and the gain coefficient K simultaneously.
[0098] (8) Embodiment 8
[0099] As described in Example 6, theoretically, the better the vibration compensation effect, the more the fringe residual R calculated only due to vibration. v The closer it is to the original stripe residual R m Therefore, Example 8 uses R m and R v The correlation coefficient C R To optimize the target, we traverse the search delay coefficient τ, with C R When the absolute value of reaches the maximum value, the corresponding delay coefficient is τ opt .
[0100] Since the original data needs to be normalized when calculating the correlation coefficient, C R It cannot be used as an optimization target when searching for the gain coefficient. Here, the RMSE value in Example 1 can be used as the optimization target for traversing and searching for the gain coefficient K.
[0101] (9) Embodiment 9
[0102] Using the correlation coefficient C in Example 8 R To optimize the target, the delay coefficient τ is searched traversally. Then, the interference fringe residual standard deviation σ in Example 6 is used c To optimize the target, the gain coefficient K is searched traversally.
[0103] (10) Embodiment 10
[0104] The hardware sensor of vibration compensation is replaced by an accelerometer, and the output signal is U a (t). At this time, a single integration is used and the delay coefficient and gain coefficient are combined to obtain U a (t) Get the vibration velocity signal v of the reference mirror m (t). The steps are still the same as those in Example 1. The RMSE value when fitting the fringes is used as the optimization target, and the delay coefficient τ and the gain coefficient K are searched separately. The specific algorithm flow is as follows: Figure 6 shown.
[0105] (11) Embodiment 11
[0106] The hardware sensor for vibration compensation is still an accelerometer. In Embodiment 10, the acceleration signal U a (t) Calculate the velocity signal v m (t) method adopts the half-cycle integration method. The RMSE value when fitting fringes is still used as the optimization target, and the delay coefficient τ and the gain coefficient K are searched separately.
[0107] The present invention has excellent adaptability to the environment and can achieve the best vibration compensation effect at that time and place when the measurement environment changes, solving the problem of relatively decreased vibration compensation effect of the existing vibration compensation method when the measurement environment changes, and effectively improving the vibration compensation effect of the atomic gravimeter in a vibration environment, thereby improving the measurement accuracy of the atomic gravimeter, and further making a certain contribution to the measurement of absolute gravity.
Claims
1. A vibration compensation method for correcting interference fringes of atomic interferometer gravimeter, Features: The following steps are involved: S1, given sweep rate sequence, gravity reference value, pulse interval duration, effective wave vector; S2, import measured data; S3, calculating and setting optimization targets through original stripes; S4, dividing the vibration sensor signal according to the trigger signal to obtain the vibration sensor signal corresponding to each transition probability; S5, setting the starting range of the delay coefficient and the gain coefficient; S6. Obtaining the optimal delay coefficient and the optimal gain coefficient by using a traversal search method or a golden section search method; S7, calculating the reference mirror vibration speed at each atomic interference by the optimal delay coefficient and the optimal gain coefficient, and then obtaining the vibration phase noise sequence, the atomic phase sequence only affected by the vibration, and the correction fringes, and finally calculating the RMSE value when the correction fringes are fitted to the standard cosine signal; In step S3, the optimization target is one of the following: the RMSE value when the original fringe is fitted, the coefficient of determination when the original fringe is fitted, the standard deviation of the interference fringe residual, and the correlation coefficient between the original fringe residual and the fringe residual affected only by vibration; In step S6, when the traversal search method or the golden section search method is used to obtain the optimal delay coefficient and the optimal gain coefficient, when the optimization target is the RMSE value during the original fringe fitting, the delay coefficient and the gain coefficient corresponding to the minimum RMSE value during the original fringe fitting are used as the optimal delay coefficient and the optimal gain coefficient; When the optimization target is the coefficient of determination during the original fringe fitting, the delay coefficient and gain coefficient corresponding to the minimum value of the coefficient of determination during the original fringe fitting are taken as the optimal delay coefficient and optimal gain coefficient; When the optimization target is the standard deviation of the interference fringe residual, the delay coefficient and gain coefficient corresponding to the minimum value of the standard deviation of the interference fringe residual are taken as the optimal delay coefficient and optimal gain coefficient; When the optimization target is the correlation coefficient between the original fringe residual and the fringe residual only affected by vibration, the delay coefficient corresponding to the time when the absolute value of the correlation coefficient between the original fringe residual and the fringe residual only affected by vibration reaches the maximum value is taken as the optimal delay coefficient; The gain coefficient corresponding to when other optimization objectives reach the minimum value is taken as the optimal gain coefficient.
2. The vibration compensation method for correcting interference fringes of an atomic interferometer gravimeter as claimed in claim 1, Features: In step S2, the measured data includes the measured transition probability sequence corresponding to the sweep rate sequence, the vertical output signal of the vibration sensor measuring the vibration of the reference mirror in the same period, and the trigger signal before each atomic group falls in the same period.
3. The vibration compensation method for correcting interference fringes of an atomic interferometer gravimeter as claimed in claim 1, Features: In step S4, the vibration sensor signal is filtered using the corrector transfer function.
4. The vibration compensation method for correcting interference fringes of an atomic interferometer gravimeter as claimed in claim 2, Features: In step S7, when the reference mirror vibration speed is calculated for each atomic interference by using the optimal delay coefficient and the optimal gain coefficient, when the vibration sensor is a seismic timing, the calculation formula is: In the formula, v m (t) is the vibration speed of the reference mirror; τ is the optimal delay coefficient; K is the optimal gain coefficient; U s (t) is the output signal of the seismometer; v s is the vibration velocity of the sensitive mass of the seismometer; K s is the nominal sensitivity of the seismometer; When the vibration sensor is an accelerometer, the calculation formula is: In the formula, v m (t) is the vibration speed of the reference mirror; τ is the optimal delay coefficient; K is the optimal gain coefficient; K a is the nominal sensitivity of the accelerometer; U a (t) is the accelerometer output signal.
5. The vibration compensation method for correcting interference fringes of an atomic interferometer gravimeter as claimed in claim 2, Features: In step S7, when the vibration speed of the reference mirror during each atomic interference is calculated by using the optimal delay coefficient and the optimal gain coefficient, when the vibration sensor is an accelerometer, a half-period integration algorithm is used to calculate the vibration speed of the reference mirror.
