Single baseline InSAR orbit error parameterization modeling and estimation method, device, equipment and medium

Abstract

The present invention discloses a single-baseline InSAR orbit error parameterized modeling and estimation method, device, equipment and medium, the method comprising: pre-processing single-baseline interferometric data in a dual-station mode to obtain an interferometric phase; simulating and calculating the interferometric phase based on external DEM data and interferometric geometric parameters to obtain a simulated interferometric phase; differentiating the obtained interferometric phase with the simulated interferometric phase, and then performing an unwrapping process to obtain an unwrapped differential interferometric phase, which is used as an observed value of the orbit error; constructing a parameterized model of the orbit error line by line based on external DEM data and interferometric geometric parameters; solving the unknown parameters in the orbit error parameterized model using a weighted iterative least squares estimation method, and then obtaining the orbit error. The present invention can be applied to the reconstruction of a high-precision, high-resolution digital elevation model (DEM).

Description

Technical Field

[0001] The present invention belongs to the field of image-based geodesy, and specifically relates to a single-baseline InSAR orbit error parameterization modeling and estimation method, device, equipment and medium in a dual-station mode. Background Art

[0002] In recent years, the synthetic aperture radar interferometry (InSAR) technology has been able to carry out large-scale high-precision terrain mapping. For example, high-precision DEM data covering the entire world have been released, such as STRM DEM, TanDEM-X DEM, etc., and have been widely used in large-scale disaster monitoring and resource surveys. Due to the anti-interference ability of the dual-station InSAR system or InSAR networking system to atmospheric delay and time decoherence errors, such as the German Aerospace Agency's TanDEM-X system, my country's Tianhui 2, and the soon-to-be-launched TanDEM-L system, they have made great contributions to the production of global high-precision and high-resolution DEMs by InSAR, as well as the inversion of higher resolution and higher-precision DEMs for local areas.

[0003] Due to the limited positioning and attitude accuracy of the inertial navigation system carried by the SAR system platform, the design deviation between the SAR antenna center and the antenna center of the GPS positioning system, the pulse time delay, the range path delay and the imaging algorithm processing error, all interferometric SAR modes cannot circumvent the influence of orbit (or track) errors, which seriously limits the accuracy of InSAR measurements. In order to estimate and correct the InSAR time-varying orbit (or track) error, a large number of related studies and correction algorithms have been developed. At present, the solution to the orbit error of spaceborne InSAR is mainly based on estimating and correcting the error phase. There are two main methods: the first is based on the fitting of empirical models, such as polynomial fitting method, which is originally intended to fit the phase with trend. Since this method is simple in principle and easy to implement, it is widely used in InSAR orbit error removal. However, in order to cater to the irregular characteristics of the orbit error phase, it is necessary to frequently change the order of the polynomial in the hope of achieving the optimal fit, but this empirical model has more or less fitting errors, so this method is more about solving the overall trend of the orbit error phase, but it is difficult to take into account the local orbit error phase; the second method is mainly based on parametric modeling based on vertical baseline error and parallel baseline error. These methods are related to the initial baseline estimation method, and the vertical baseline and parallel baseline are generally estimated separately through different estimation strategies. For example, the parallel baseline can generally be approximated by the slant range difference or based on the fringe rate method, while the vertical baseline is estimated by the orbit parameter or fringe rate method. Affected by the InSAR geometric relationship approximation or the Fourier FFT (Fast Fourier Transformation) window size and flat ground assumption, the estimated values ​​of the vertical baseline and parallel baseline are no longer constrained by the interferometric geometric relationship, which will introduce the phase error related to the terrain elevation and mix with the orbit error, greatly increasing the difficulty of removing the orbit error phase.

[0004] Therefore, in high-precision and high-resolution terrain mapping applications, it is necessary to design a single-baseline InSAR orbit error parameterization modeling and accurate estimation method. Summary of the invention

[0005] The present invention provides a single-baseline InSAR orbit error parameterization modeling and estimation method, device, equipment and medium, which can ensure that the interferometric synthetic aperture radar (InSAR, Synthetic Aperture Radar Interferometry) technology can achieve high-precision and high-resolution reconstruction of the DEM in high mountain areas.

[0006] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0007] A parameterized modeling and estimation method for single baseline InSAR orbit error in dual-station mode, including:

[0008] Step 1, preprocessing the single baseline interferometric data in the dual-station mode to obtain the interferometric phase;

[0009] Step 2, simulating and calculating the interference phase based on external DEM data and interference geometric parameters;

[0010] Step 3, the obtained interference phase is differentiated with the simulated interference phase, and then unwrapped to obtain the unwrapped differential interference phase, which is used as the observation value of the orbit error;

[0011] Step 4: Based on the external DEM data and interferometric geometric parameters, a parameterized model of the track error is constructed row by row;

[0012] Step 5: construct the observation equation based on the observed value of the orbit error and the parameterized model, solve the unknown parameters in the orbit error parameterized model, and then obtain the orbit error.

[0013] Furthermore, the unknown parameters in the orbit error parameterization model only include the baseline length error and the baseline inclination error.

[0014] Furthermore, the expression of the parameterized model of the orbit error is:

[0015]

[0016]

[0017] In the formula, represents the orbit error; p represents the interference mode, p = 2 in the heavy orbit mode, and p = 1 in the dual antenna mode; λ represents the microwave wavelength; represents the differential interferometric phase; B represents the measured value of the baseline length, ΔB is the baseline length error, that is, the difference between the measured value and the actual value of the baseline length; α represents the measured value of the baseline inclination, Δα is the baseline inclination error, that is, the difference between the measured value and the actual value of the baseline inclination; R 10 Represents the measurement of the slant range from the center of the main antenna to the ground target; Represents the slant distance from the center of the secondary antenna to the ground target calculated based on geometric parameters and external DEM data; "±" or " " represents the relative position change of the main and sub antennas in the interference geometry. The positive sign represents that the sub antenna is on the right side of the main antenna, while the negative sign represents that the sub antenna is on the left side of the main antenna. θ represents the viewing angle when the ground target is at a height of h, which is the sum of the design viewing angle θ0 of the center of the main antenna and the angle Δθ between the ground target and the line of sight of the radar wave.

[0018] Furthermore, the calculation formula for the simulated interference phase is:

[0019]

[0020] In the formula, represents the simulated interference phase, p represents the interference mode, p = 2 in the heavy track mode, and p = 1 in the dual antenna mode; λ represents the microwave wavelength; R 10 Represents the measurement of the slant range from the center of the main antenna to the ground target; Represents the slant range from the center of the secondary antenna to the ground target calculated based on geometric parameters and external DEM data.

[0021] Furthermore, the viewing angle θ is estimated based on the interferometric geometry parameters:

[0022]

[0023] In the formula, H r is the distance between the main radar antenna position and the center of the earth in the case of spaceborne; R e is the radius of the Earth at the ground target.

[0024] Furthermore, the single-baseline interferometric data in the dual-station mode uses the X-band single-baseline interferometric data of the TanDEM-X dual-station system; and the external data uses the TanDEM-X DEM.

[0025] Furthermore, in step 5, the weighted iterative least squares method is used to solve the unknown parameters in the orbit error parameterization model.

[0026] A single baseline InSAR orbit error parameterization modeling and estimation device in dual-station mode, comprising:

[0027] The interferometric phase acquisition module is used to: pre-process the single baseline interferometric data in the dual-station mode to obtain the interferometric phase;

[0028] The interference phase simulation calculation module is used to: simulate and calculate the interference phase based on external DEM data and interference geometric parameters to obtain a simulated interference phase;

[0029] The difference module is used to: perform differentiation of the obtained interference phase with the simulated interference phase, and then perform unwrapping processing to obtain the unwrapped differential interference phase, which is used as the observation value of the orbit error;

[0030] Track error model building module, used to: build a parametric model of track error line by line based on external DEM data and interferometric geometric parameters;

[0031] The orbit error solving module is used to: construct the observation equation according to the observed value of the orbit error and the parameterized model, solve the unknown parameters in the orbit error parameterized model, and then obtain the orbit error.

[0032] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor implements any one of the methods described above.

[0033] A computer-readable storage medium stores a computer program, which implements any of the above methods when executed by a processor.

[0034] The beneficial effects of the present invention include:

[0035] (1) The composition and error characteristics of orbit error were determined, and a parameterized model of orbit error was constructed to ensure accurate estimation and correction of orbit error under complex terrain conditions;

[0036] (2) External DEM refers to DEM data released free of charge worldwide. For the TanDEM-X system and the Tianhui-2 system, the selection of TanDEM-X DEM can ensure the improvement of the accuracy of the orbit error estimation of the two interferometric systems.

[0037] (3) The present invention can be used for accurate estimation of interference baseline parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 Flowchart of the parameterized modeling and precise estimation method of single-baseline InSAR orbit error in dual-station mode.

[0039] Figure 2 It is the geographical location and topographical conditions of the test area in the implementation case.

[0040] Figure 3 The orbit error phase correction result of the experimental area estimated in step V of the implementation case. (a) Original differential interferometry phase; (b) Orbit error phase estimated by polynomial fitting method considering elevation term; (c) Orbit error phase estimated by the method of the present invention; (d) Differential phase corrected by polynomial fitting method; (e) Differential phase corrected by the method of the present invention; (f) is the difference between (d) and (e).

[0041] Figure 4 The residual phase after orbit error correction is a local enlarged view after correction by (a) polynomial fitting method and (b) the method of the present invention.

[0042] Figure 5 The horizontal and vertical profiles of the differential phase after correction by (a) the polynomial fitting method and (b) the new method.

[0043] Figure 6 The composition of orbit errors in the experimental area. (a) Residual flat ground phase error; (b) Terrain phase error caused by orbit error

[0044] Figure 7 is the baseline parameter error estimated in step V of the implementation case. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present invention clearer, high-resolution interferometric data obtained by the German Aerospace Engineering Agency's TanDEM-X dual-station system are used. The specific parameters are shown in Appendix 1. Since the spatial distance (i.e., baseline length) between the two satellites TerraSAR-X / TanDEM-X is not large, about 120m to 500m, it helps to offset the atmospheric delay error. In addition, the two satellites receive the surface echo signal almost simultaneously, and the temporal decorrelation can be ignored. The verification experimental area is located in You County, Zhuzhou City, Hunan Province. The terrain in this area is significantly undulating, and the elevation range is 40 to 1400m. See Appendix 1. Figure 2 In summary, the selection of interference data and experimental area is very suitable for verifying the orbit error correction effect of the new method in the high-resolution interference map in the terrain undulating area. The specific implementation cases described here are only used to explain the present invention and are not used to limit the present invention.

[0046] Table 1 Geometric parameters of TanDEM-X dual-station interferometry data

[0047]

[0048] Based on the X-band single baseline interferometric data of the TanDEM-X dual-station system used above, this embodiment provides a single baseline InSAR orbit error parameterization modeling and estimation method in dual-station mode. Figure 1 As shown, the following steps are included:

[0049] Step 1: Preprocess the single baseline interferometric data in dual-station mode to obtain the interferometric phase.

[0050] Since the dual-station interferometric data have been registered and resampled, interferometric processing and multi-view processing can be directly performed to obtain the final multi-view interferometric phase The row and column numbers are 4448x2621, and the final spatial resolution is 12x12m (azimuth x distance).

[0051] In dual-station InSAR mode (such as TanDEM-X, Tianhui-2, etc.), atmospheric delay, temporal decorrelation and surface deformation signals can be ignored, so the interferometric phase The phase composition can be written as:

[0052]

[0053] In the formula, is the terrain phase; It is the flat ground phase; is the thermal noise phase.

[0054] According to the interference geometry, the function expression of the interference phase is written as the slant range difference between the center of the main and secondary radar antennas to the ground target:

[0055]

[0056] Wherein, R1 is the slant distance from the center of the main antenna to the ground target; R2 is the slant distance from the center of the secondary antenna to the ground target; λ represents the microwave wavelength; p represents the interference mode, p = 2 in the heavy track mode, and p = 1 in the dual antenna mode.

[0057] Step 2, simulating and calculating the interference phase based on external DEM data and interference geometric parameters to obtain a simulated interference phase.

[0058] In this embodiment, the external DEM data uses the global TanDEM-X DEM data (elevation is h), combined with the interferometric geometric parameters in the attached table, to perform interferometric phase simulation operations to prepare for differential processing. Specifically, it includes:

[0059] According to the interferometric geometric relationship, the interferometric geometric parameters and external DEM data are used to calculate the slant distance from the center of the secondary antenna to the ground target.

[0060]

[0061] In the formula, “±” or “ " represents the relative position change of the main and sub antennas in the interference geometry. The positive sign represents that the sub antenna is on the right side of the main antenna, while the negative sign represents that the sub antenna is on the left side of the main antenna. Represents the slant range from the center of the secondary antenna to the ground target calculated based on geometric parameters and external DEM data; R 10 represents the measured value of the slant distance from the center of the main antenna to the ground target; B and α represent the measured values ​​of the baseline parameters (baseline length and baseline inclination, respectively); θ represents the viewing angle when the ground target is at a height of h, which is the sum of the design viewing angle θ0 of the center of the main antenna and the angle Δθ between the ground target and the line of sight of the radar wave. According to the interference geometry,

[0062]

[0063] In the formula, H r is the distance between the main radar antenna position and the center of the earth in the case of spaceborne; R e is the radius of the Earth at the ground target.

[0064] Therefore, the simulated interference phase It can be written as,

[0065]

[0066] Step 3: perform differential and unwrapping processing on the obtained interference phase and the simulated interference phase to obtain the unwrapped differential interference phase, which is used as the observed value of the orbit error. The details are as follows:

[0067] (1) Obtaining differential interferometry phase In fact, it is the process of removing the simulated interference phase from the interference phase, that is,

[0068]

[0069] At this time, the composition of the dual-station InSAR differential phase includes the orbit error Terrain height residual phase and the noise phase, i.e.

[0070]

[0071] Note that the use of external DEM data determines the terrain height residual phase In order to eliminate the influence of this part on the orbit error estimation, the height deviation between the external DEM and the elevation measured by the interferogram should be avoided as much as possible. Therefore, the use of external DEM data is key.

[0072] For the dual-station interferometric data obtained by TanDEM-X and Tianhui-2 satellite, the global TanDEM-X DEM data can be used as the external DEM to reduce the impact of elevation error on orbit error.

[0073] For vegetation areas and ice and snow areas, the difference in phase center height in the differential interferometry phase is almost negligible. Therefore, for areas with high penetration such as vegetation areas and ice and snow areas, external DEM data can use other DEM data with inconsistent bands.

[0074] (2) The Goldstein filter algorithm is used to filter the differential interferometry phase. The filter window is set to 32×32 pixels and the filter factor is 0.35, thereby obtaining the filtered differential interferogram. Then, the minimum network cost flow algorithm is used to perform phase unwrapping operations on the filtered differential phase. The weighting factor used for unwrapping is the coherence map before or after filtering. In order to avoid the influence of low coherence areas on unwrapping, the pixels with coherence lower than 0.3 are masked off during the unwrapping process. Finally, the unwrapped differential interferogram is obtained as the observation value of the parameterized model. See Appendix. Figure 3 (a).

[0075] Step 4: Based on the external DEM data and interferometric geometric parameters, a parameterized model of the track error is constructed row by row.

[0076] According to equations (2), (5), and (6), the differential phase can be further written as:

[0077]

[0078] Due to the finiteness of the observed geometric parameters and the existence of observation errors, the differential phase shown in equation (8) is used to obtain the partial derivative of the interferometric geometric parameters (slant range of the main antenna, baseline length, baseline inclination, and viewing angle) to construct a parameterized model of the orbit error relative to the geometric parameters, namely:

[0079]

[0080] In the formula, represents the partial derivative of the differential phase with respect to the slant range of the main antenna; represents the partial derivative of the differential phase with respect to the baseline length; represents the partial derivative of the differential phase with respect to the baseline inclination; Represents the partial derivative of the differential phase with respect to the viewing angle; ΔR 10 ,ΔB,Δα,Δθ0 represent the slant range error, baseline length error, baseline inclination error and design viewing angle error respectively.

[0081] However, directly estimating the error of the interferometric geometric parameters in model (9) is inaccurate, no matter which parameter estimation algorithm is used, because the interferometric geometric parameters have errors, which implies that the parameters to be estimated are mutually coupled. Obviously, the interferometric geometric relationship is uncertain. Therefore, assuming that the measured slant distance R 10 and the design view angle θ0 are accurate, and the main track position is fixed. Then, by fixing the ground elevation, the estimation of the orbit error can be completely transferred to the solution of the baseline parameter error. Therefore, the orbit error parameterization model can be derived as:

[0082]

[0083] Where:

[0084]

[0085] It is easy to see from equations (10) and (11) that θ implies that the orbit error phase consists of two parts. One part is related to the interferometric imaging geometry (i.e., the design viewing angle θ0 corresponding to the reference surface), which is called the residual flat-Earth phase The other part is related to terrain elevation and is considered to be a terrain phase error caused by orbit error. Right now

[0086]

[0087] Note that only when the terrain elevation is close to the reference surface elevation, that is, Δθ approaches 0, the orbit error phase can be considered to contain only the residual flat-earth phase error.

[0088] Considering that the orbit error varies in time and frequency, the above parameterized model is constructed based on the range line. Then the baseline parameter error of each range line is phase reconstructed along the azimuth direction, which is the orbit error phase to be determined.

[0089] In this embodiment, there are more than 2000 pixels in each distance direction, which constitute more than 2000 observation equations to solve two unknowns X(ΔB, Δα). It can be seen that the equation group is overdetermined.

[0090] Step 5: construct the observation equation based on the observed value of the orbit error and the parameterized model, solve the unknown parameters in the orbit error parameterized model, and then obtain the orbit error.

[0091] The weighted iterative least squares estimation method is used to solve the model parameters, that is,

[0092] [ΔB,Δα] T =pinv(C T PC)·C T PL (13)

[0093] Where pinv(*) represents the pseudo-inverse operation; C and L are the coefficient matrix and observation matrix of the model, that is,

[0094]

[0095] Where n represents the number of available pixels in each distance row; P represents the weight matrix (n×n order), and the weight matrix uses the weight matrix adjustment factor. The weight function can be designed using Zhou Jiang grammar as

[0096]

[0097] In the formula, σ 2 =(v b ) T ·P b-1 ·v b / (n-1 is the unit weight variance, b represents the number of iterations; v i Represents the correction factor of the observation value; ε is a minimum value. In order to avoid the denominator being 0, it can be set to 10 -6 ; k1 and k2 are two empirical values, here set k1 = 1, k2 = 2. Moreover, the weight matrix and parameter estimation need to be iteratively calculated, that is,

[0098]

[0099] Among them, the initial weight matrix P 0 Expressed by coherence; note that the iteration termination condition is b>100 or ξ is the threshold for two iterative parameter estimates (take 10 -6 ).

[0100] The baseline parameter error (ΔB, Δα) estimated in step 5 is used to compensate the initial baseline parameters (B, α) to obtain the refined baseline parameters (B new ,α new ),Right now

[0101]

[0102] The baseline parameters estimated in this embodiment are used to compensate the initial baseline parameters in Appendix 1, and the accurately estimated baseline parameters are B new =76.148m,αn ew =18.030°.

[0103] In order to verify the reliability of orbit error estimation, the method of the present invention is compared with the currently widely used polynomial fitting method. The estimation results are shown in the attached Figure 3 After the orbit error correction, it is obvious that the residual phase after polynomial fitting correction is Figure 3 (d) The differential phase after correction by the method of the present invention (see Appendix Figure 3 (e)) is more inclined to 0 as a whole, and the differential phase corrected by the polynomial fitting method shows inconsistency in multiple local areas. Figure 3 (e), the phase difference between the two is concentrated between [-1rad, 1rad], as shown in the attached Figure 3 (f) is shown. According to the phase height conversion calculation, attached Figure 3 The phase difference in (f) will cause a height error of about [-10m, 8m]. Figure 3 In (d) and (e), a small local area (such as the black rectangle in the figure) is selected for enlarged analysis, and two horizontal and vertical profile lines (such as pf_r1 and pf_c1) are also selected for consistency test. Figure 4 After correction by the method of the present invention, the terrain details such as ridge lines and valley lines can be clearly identified. However, there are many residual orbit error phases in the phase results corrected by the polynomial fitting method. Figure 5 The profile results in show that the differential phase profile line corrected by the new method is closer to 0, regardless of the transverse or longitudinal profile, and the RMSE of the differential phase corrected by the new method is the smallest (i.e., 0.23 rad and 0.18 rad), indicating that the orbit error phase correction effect is better. In addition, the contribution of the two components of the orbit error phase is quantified here, see Appendix Figure 6 Attached Figure 6 (a) is the residual flat-earth phase error, which is completely a trend phase, and the stripes are very regular, which contributes to the vast majority of the orbit error phase. Figure 6(b) is the phase error related to terrain height. The contribution of this term is very small and can be easily masked by the residual flat ground phase error, but its altimetry impact is very significant and increases the uncertainty of the orbit error phase. It can be seen that this term can only be obtained when the prior elevation is known.

[0104] The baseline parameter errors estimated in this step are shown in the attached Figure 7 It can be seen that the time-frequency error of the baseline length estimated by the model fluctuates from -33.04m to -33.02m, while the time-frequency error of the baseline inclination changes from 0.057rad to 0.068rad.

[0105] The above embodiments are preferred embodiments of the present application. Ordinary technicians in this field can also make various changes or improvements on this basis. Without departing from the overall concept of the present application, these changes or improvements should fall within the scope of protection required by the present application.

Claims

1. A single baseline InSAR orbit error parameterization modeling and estimation method in dual-station mode, characterized in that: include: Step 1, preprocessing the single baseline interferometric data in the dual-station mode to obtain the interferometric phase; Step 2, simulating and calculating the interference phase based on external DEM data and interference geometric parameters; Step 3, the obtained interference phase is differentiated with the simulated interference phase, and then unwrapped to obtain the unwrapped differential interference phase, which is used as the observation value of the orbit error; Step 4, construct the parameterized model of track error line by line based on external DEM data and interferometric geometric parameters; Step 5: construct the observation equation based on the observed value of the orbit error and the parameterized model, solve the unknown parameters in the orbit error parameterized model, and then obtain the orbit error.

2. The method according to claim 1, characterized in that The unknown parameters in the orbit error parameterization model only include the baseline length error and the baseline inclination error.

3. The method according to claim 2, characterized in that The expression of the parameterized model of the orbit error is: In the formula, Indicates orbit error; p represents the interference mode, p = 2 in the heavy orbit mode, and p = 1 in the dual antenna mode; λ represents the wavelength of microwaves; represents the differential interferometric phase; B represents the measured value of the baseline length, and ΔB is the baseline length error, i.e., the difference between the measured value and the actual value of the baseline length; α represents the measured value of the baseline inclination, Δα is the baseline inclination error, that is, the difference between the measured value and the actual value of the baseline inclination; R 10 Represents the measurement of the slant range from the center of the main antenna to the ground target; Represents the slant distance from the center of the secondary antenna to the ground target calculated based on geometric parameters and external DEM data; "±" or Represents the relative position change of the main antenna and the sub-antenna in the interference geometry. The positive sign represents that the sub-antenna is on the right side of the main antenna, while the negative sign represents that the sub-antenna is on the left side of the main antenna. θ represents the viewing angle when the ground target is at a height of h. It is the sum of the design viewing angle θ0 of the main antenna center and the angle Δθ between the ground target and the line of sight of the radar wave.

4. The method according to claim 1, characterized in that: The calculation formula for the simulated interference phase is: In the formula, represents the simulated interference phase, p represents the interference mode, p = 2 in the heavy track mode, and p = 1 in the dual antenna mode; λ represents the wavelength of microwaves; R 10 Represents the measurement of the slant range from the center of the main antenna to the ground target; Represents the slant range from the center of the secondary antenna to the ground target calculated based on geometric parameters and external DEM data.

5. The method according to claim 3 or 4, characterized in that: The viewing angle θ is estimated based on the interferometric geometry parameters: In the formula, H r is the distance between the main radar antenna position and the center of the earth in the case of spaceborne; R e is the radius of the Earth at the ground target.

6. The method according to claim 1, characterized in that The single-baseline interferometric data in the dual-station mode uses the X-band single-baseline interferometric data of the TanDEM-X dual-station system; the external data uses the TanDEM-X DEM.

7. The method according to claim 1, characterized in that Step 5 uses the weighted iterative least squares method to solve the unknown parameters in the orbit error parameterization model.

8. A single baseline InSAR orbit error parameterization modeling and estimation device in dual-station mode, characterized in that: include: The interferometric phase acquisition module is used to: pre-process the single baseline interferometric data in the dual-station mode to obtain the interferometric phase; The interference phase simulation calculation module is used to: simulate and calculate the interference phase based on external DEM data and interference geometric parameters to obtain a simulated interference phase; The difference module is used to: perform differentiation of the obtained interference phase with the simulated interference phase, and then perform unwrapping processing to obtain the unwrapped differential interference phase, which is used as the observation value of the orbit error; Track error model building module, used to: build a parametric model of track error line by line based on external DEM data and interferometric geometric parameters; The orbit error solving module is used to: construct the observation equation based on the observed value of the orbit error and the parameterized model, solve the unknown parameters in the orbit error parameterized model, and then obtain the orbit error.

9. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the computer program is executed by the processor, the processor is caused to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images