Online compensation method for measurement error of space shape sensor under condition that measurement target is unknown

Abstract

The invention provides an online compensation method for a measurement error of a space shape sensor under the condition that a measurement target is unknown, and belongs to the field of space deployable structures. The method comprises the steps that a dynamically updated source domain is constructed based on a geometric model of a to-be-measured target, and the source domain comprises multiple sets of known curvature distributions and corresponding shape reconstruction error distributions; the method comprises the following steps: measuring a to-be-measured target in real time through a shape sensor, constructing curvature distribution obtained by measurement into a target domain, and screening a source domain curve closest to the curvature distribution of the target domain from a source domain as a matching curve; and migrating the reconstruction error distribution of the screened source domain curve to a target domain through affine transformation, and compensating the reconstruction error of the current measurement shape in real time. The problem that the error distribution cannot be directly evaluated under the unknown shape is solved.

Description

Technical Field

[0001] The present invention belongs to the field of space deployable structures, and particularly relates to an on-line compensation method for the measurement error of a space shape sensor under the condition of unknown measurement target. Background Art

[0002] As the "ears" of a satellite, spaceborne antennas are one of the indispensable important devices in the space-based electronic reconnaissance system. The size of the antenna aperture directly determines the signal transceiver and transmission capabilities. To form a long-distance and high-resolution space reconnaissance capability for weak electromagnetic targets on the ground, the antenna aperture is developing towards large size and high precision to achieve high signal gain.

[0003] Space deployable antennas mainly include two categories: spaceborne mesh antennas and spaceborne thin-film antennas, both of which are composed of deployable peripheral trusses and flexible reflectors. The reflector of a spaceborne mesh antenna is formed by tensioning multiple kelvar ropes to support the reflector, and a flexible metal reflector mesh woven from micron-scale metal wires and the rope support together form a parabolic or parabolic cylindrical reflector. The reflector of a thin-film antenna is formed by electrostatic adsorption of two layers of polyimide films about 1 mm thick to form a parabolic reflector. When the size of the space deployable antenna increases to more than 30 meters, non-uniform and time-varying temperature gradients are generated in the structure, and the traditional method of passively controlling the surface accuracy through robust design can no longer guarantee the shape accuracy of the antenna reflector. Therefore, establishing an on-line monitoring method for large-size and flexible reflectors is the basis for regulating the antenna surface.

[0004] On the ground, the deformation of the antenna is mainly measured by the photogrammetry method. The principle is as follows: Target points are pasted on the antenna, and the camera continuously takes pictures at multiple positions along one circle of the antenna. By using the coordinate matching between the target points, the pictures are stitched together to restore the configuration of the antenna. However, there are the following problems in using the photogrammetry method in the space environment: First, the camera light source is weak, and it is difficult to have a wide range of flash capabilities; second, the space light field is uneven, and the quality of the pictures is unstable; third, limited by the camera focal length and depth of field, it is difficult to obtain the overall shape of the antenna even with multiple cameras.

[0005] In recent years, the developed fiber optic shape sensors have the characteristics of anti-electromagnetic interference, light weight, high sensitivity, etc., and have outstanding advantages in the deformation measurement of large-size and flexible structures in space. However, the key difficulty in the current application in space structure measurement is that the shape sensing principle is based on the point-by-point recursion of multi-point discrete curvature, and there is error accumulation in long-distance measurement, and the measurement error changes with the shape; when the shape of the space target to be measured is unknown, it is impossible to evaluate the distribution of the sensor measurement shape error, and it is even more impossible to compensate the sensor shape measurement error. Summary of the Invention

[0006] Technical Problems to be Solved

[0007] To avoid the deficiencies of the prior art, the present invention provides an online compensation method for the measurement error of a spatial shape sensor under the condition of an unknown measurement target. Based on a geometric model, theoretical curvature distributions and corresponding error data under various working conditions are generated, and a dynamically updated source domain set is constructed, breaking through the limitations of traditional static error compensation; by calculating the affine matrix of the curvature distributions of the target domain and the source domain, the error distribution of the matching curve in the source domain is migrated to the current measurement curve, solving the problem that the error distribution under an unknown shape cannot be directly evaluated.

[0008] The technical solution of the present invention is: an online compensation method for the measurement error of a spatial shape sensor under the condition of an unknown measurement target, and the specific steps are as follows:

[0009] Based on the geometric model of the target to be measured, a dynamically updated source domain is constructed, and the source domain includes multiple groups of known curvature distributions and their corresponding shape reconstruction error distributions;

[0010] The target to be measured is measured in real time by a shape sensor, the measured curvature distribution is constructed as the target domain, and the source domain curve closest to the curvature distribution of the target domain is selected from the source domain as the matching curve;

[0011] The reconstruction error distribution of the selected source domain curve is migrated to the target domain through affine transformation to perform real-time compensation on the reconstruction error of the current measured shape.

[0012] A further technical solution of the present invention is: the specific process of constructing the dynamically updated source domain is as follows:

[0013] According to the geometric model of the target to be measured, working condition loads are applied to generate multiple deformed surfaces of the reflecting surface;

[0014] Define a theoretical shape curve at the sensor layout position, and extract geometric curvature at the same spatial resolution as the shape sensor;

[0015] Based on the Frenet-Serret formula, reconstruct the shape curve, calculate the ordinate difference between the reconstructed curve and the theoretical shape curve, and obtain the reconstruction error distribution;

[0016] Based on multiple deformed surfaces of the reflecting surface, establish a curvature distribution at the measurement position of the shape sensor and its corresponding reconstruction error data group to obtain the initial source domain;

[0017] When the difference between the curvature distribution of the target domain and all curves in the initial source domain exceeds a preset threshold, the curvature distribution of the target domain and its reconstruction error distribution are dynamically appended to the initial source domain to realize the dynamic update of the source domain.

[0018] A further technical solution of the present invention is: the preset threshold is 0.39, and when the difference is greater than 0.39, it is determined that the difference between the target domain and the source domain curve is significant.

[0019] A further technical solution of the present invention is that the screening process of the closest source domain curve includes:

[0020] Calculate the affine matrix of the curvature distribution of the target domain and the curvature distribution of each curve in the source domain, and the affine matrix is defined as:

[0021]

[0022] Wherein, The subscript h is the number of curvature measurement points corresponding to the curve, and the superscripts sou and tar respectively represent the source domain and the target domain;

[0023] Based on the affine matrix, transform the source domain curve, and calculate the curvature distribution difference between the transformed source domain curve and the target domain curve:

[0024]

[0025] Wherein, is the curvature of the i-th curve in the source domain after being transmitted by the affine matrix at the j-th point; is the curvature of the target curve at the j-th point; h is the number of curvature measurement points corresponding to the curve;

[0026] Calculate the curvature distribution difference between each transformed source domain curve and the target domain curve, and select the source domain curve with the smallest curvature distribution difference as the matching curve; if all the curvature distribution differences exceed the threshold, trigger the dynamic update of the source domain.

[0027] A further technical solution of the present invention is that the expression of the reconstruction error distribution of the matching curve migrating to the target domain is:

[0028]

[0029] Wherein, is the reconstruction error distribution of the matching curve, V is the corresponding affine matrix, and E tar is the predicted value of the reconstruction error distribution of the target domain.

[0030] A further technical solution of the present invention is that the shape sensor is a weak grating array fiber optic sensor, which demodulates the central wavelength of the grating through an optical frequency domain reflectometer to obtain the structural curvature.

[0031] A further technical solution of the present invention is that the object to be measured is a spaceborne large-size flexible antenna reflector, a ground phased array radar array surface, or a space deployable structure using a fiber grating array.

[0032] An on-line compensation system for the measurement error of a space shape sensor under the condition of unknown measurement target includes the following modules:

[0033] A dynamic source domain construction module is used to generate theoretical curvature distributions and corresponding reconstruction error distribution data sets under multiple deformation conditions based on the geometric model of the target to be measured, forming a dynamically updated source domain set;

[0034] The curvature distribution matching module is used to receive the real-time measured curvature data from the shape sensor, calculate the affine matrix of the curvature distribution of the target domain and the curvature distribution of each curve in the source domain, and select the source domain curve with the smallest difference;

[0035] An error migration and compensation module migrates the reconstructed error distribution of the matching curve to the target domain through the affine matrix, and performs real-time error compensation on the measurement results of the current shape sensor;

[0036] The dynamic update module appends the curvature distribution of the target domain and its reconstruction error data set to the source domain set when the difference between the curvature distribution of the target domain and all curves in the source domain exceeds a preset threshold, thereby realizing the dynamic expansion of the source domain set.

[0037] An electronic device comprises at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform an online compensation method for measurement errors of a spatial shape sensor when the measurement target is unknown.

[0038] A computer-readable storage medium stores computer instructions, which are used to enable a processor to implement an online compensation method for measurement errors of a spatial shape sensor when the measurement target is unknown.

[0039] Beneficial effects

[0040] The beneficial effects of the present invention are as follows: It proposes an online compensation method for spatial shape sensor measurement errors. This method integrates structural features with monitoring data, selects the curve closest to the currently measured shape from a dynamically updated source domain, and rapidly predicts shape reconstruction errors through affine transformation. This improves the measurement accuracy of long-distance and complex shapes and overcomes the problem that shape sensor error distribution varies with distance and shape. Specific advantages are analyzed as follows:

[0041] 1. This method uses a dynamic source domain to centrally store curvature distributions and error data for various working conditions, combined with affine transformations to accurately match target domain curves. This effectively overcomes the error accumulation problem caused by the limited recursive step size in traditional methods. In an embodiment, the maximum reconstruction error after compensation is reduced by more than 99% (for example, the error is reduced from 33.9mm to 0.186mm at a 14cm measurement point resolution), significantly improving the measurement accuracy of long distances and complex shapes.

[0042] 2. The present invention realizes mathematical closed-loop compensation through an error migration formula, eliminates the random errors introduced by the shape sensor due to environmental interferences (such as temperature gradients, electromagnetic noises), and ensures the high reliability of the measurement results.

[0043] 3. Traditional methods rely on a priori models of the target shape, while the present invention can evaluate the error distribution without prior knowledge of the theoretical geometric shape through real-time matching and updating of the dynamic source domain set. For example, when the difference between the target curve and the source domain exceeds a threshold (such as 0.39), the system automatically incorporates the new data into the source domain set to achieve an adaptive expansion of the error model, solving the problem of error evaluation in scenarios with unknown shapes.

[0044] 4. The dynamic expansion mechanism of the source domain of the present invention ensures that the system can adapt to complex working conditions (such as space temperature changes, dynamic loads on ground radar arrays), avoiding the limitations of traditional static databases. In the embodiment, the system improves the coverage of subsequent matching by adding the 8th significantly different target curve to the source domain set, verifying the self-learning ability of the system.

[0045] 5. The present invention is applicable to various scenarios such as spaceborne large-sized flexible antennas and ground phased array radars, supports the anti-electromagnetic interference characteristics of fiber optic sensors (such as weak grating arrays), and meets the requirements of space and complex electromagnetic environments. Compared with traditional photogrammetry methods (requiring multi-camera collaboration and light source control), the present invention realizes error compensation based on algorithms, without additional hardware investment, significantly reducing the deployment cost and being easily integrated with existing sensing systems.

[0046] 6. The modular design of the system of the present invention (such as the dynamic source domain construction module, error migration and compensation module) enhances the robustness and avoids the impact of single-point failures on the overall performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of a 10m diameter parabolic antenna in the embodiment of the present invention;

[0048] Figure 2 It is a comparison diagram of the reconstructed curve and the theoretical curve under various measurement point spatial resolutions in the embodiment of the present invention;

[0049] Figure 3 It is a relationship diagram between the maximum reconstruction error and the measurement point resolution in the embodiment of the present invention;

[0050] Figure 4 It is a schematic diagram of the source domain curve for migration compensation in the embodiment of the present invention;

[0051] Figure 5 It is a predicted error distribution curve diagram in the embodiment of the present invention;

[0052] Figure 6 It is a comparison diagram of the curves before and after error compensation in the embodiment of the present invention;

[0053] Figure 7 This is the maximum correction error graph under various measurement point spatial resolutions in the embodiments of the present invention;

[0054] Figure 8 This is the maximum error reduction percentage graph under various measurement point spatial resolutions in the embodiments of the present invention;

[0055] Figure 9 This is the updated source domain schematic diagram in the embodiments of the present invention. Detailed implementation manners

[0056] The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention, and should not be construed as limiting the present invention.

[0057] Based on the problems that in the existing spatial shape sensors in long-distance and large-size flexible structures (such as spaceborne antennas), due to the limited curvature recurrence step size, the reconstruction error accumulates with the distance, affecting the measurement accuracy; when the theoretical shape after the target deformation is unknown, the traditional method cannot predict the error distribution and it is difficult to achieve real-time error compensation; in the space environment, factors such as temperature gradient and electromagnetic interference cause the traditional photography method to fail, and the existing sensor technology cannot meet the high-precision dynamic monitoring requirements, etc., the present invention provides an online compensation method for the measurement error of a spatial shape sensor under the condition that the measurement target is unknown, and the specific steps are as follows:

[0058] Step 1: Based on the geometric model of the target to be measured, construct a dynamically updated source domain, and the source domain includes multiple groups of known curvature distributions and their corresponding shape reconstruction error distributions;

[0059] Specifically, the specific process of constructing the dynamically updated source domain is as follows:

[0060] Step 1.1 Apply working condition loads according to the geometric model of the target to be measured to generate multiple deformed surfaces of the reflecting surface;

[0061] Step 1.2 Define the theoretical shape curve at the sensor layout position, and extract the geometric curvature with the same spatial resolution as the shape sensor;

[0062] Step 1.3 Reconstruct the shape curve based on the Frenet-Serret formula, calculate the ordinate difference between the reconstructed curve and the theoretical shape curve, and obtain the reconstruction error distribution;

[0063] Step 1.4 Based on multiple deformed surfaces of the reflecting surface, establish a curvature distribution at the measurement position of the shape sensor and its corresponding reconstruction error data group to obtain the initial source domain;

[0064] Step 1.5: When the difference between the curvature distribution of the target domain and all curves in the initial source domain exceeds a preset threshold, the curvature distribution of the target domain and its reconstruction error distribution are dynamically appended to the initial source domain to achieve dynamic update of the source domain.

[0065] Step 2: Use a shape sensor to measure the target in real time, construct the curvature distribution obtained from the measurement into a target domain, and select a source domain curve that is closest to the curvature distribution of the target domain from the source domain as a matching curve;

[0066] Step 3: The reconstruction error distribution of the filtered source domain curve is transferred to the target domain through affine transformation, and the reconstruction error of the current measured shape is compensated in real time.

[0067] Specifically, the shape sensor is a weak grating array optical fiber sensor, which demodulates the grating center wavelength through an optical frequency domain reflectometer to obtain the structural curvature.

[0068] Specifically, the target to be measured is a large-sized flexible antenna reflector surface on board a satellite, an array surface of a ground-based phased array radar, or a spatially deployable structure using a fiber grating array.

[0069] The present invention provides an online compensation system for measurement errors of a spatial shape sensor when the measurement target is unknown, comprising the following modules:

[0070] A dynamic source domain construction module is used to generate theoretical curvature distributions and corresponding reconstruction error distribution data sets under multiple deformation conditions based on the geometric model of the target to be measured, forming a dynamically updated source domain set;

[0071] The curvature distribution matching module is used to receive the real-time measured curvature data from the shape sensor, calculate the affine matrix of the curvature distribution of the target domain and the curvature distribution of each curve in the source domain, and select the source domain curve with the smallest difference;

[0072] An error migration and compensation module migrates the reconstructed error distribution of the matching curve to the target domain through the affine matrix, and performs real-time error compensation on the measurement results of the current shape sensor;

[0073] The dynamic update module appends the curvature distribution of the target domain and its reconstruction error data set to the source domain set when the difference between the curvature distribution of the target domain and all curves in the source domain exceeds a preset threshold, thereby realizing the dynamic expansion of the source domain set.

[0074] The present invention provides an electronic device, characterized in that it includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute an online compensation method for measurement errors of a spatial shape sensor when the measurement target is unknown.

[0075] The present invention provides a computer-readable storage medium, characterized in that: the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the online compensation method for the measurement error of a spatial shape sensor when the measurement target is unknown.

[0076] The present invention proposes a high-precision detection method for large spatial dimensions and flexible surfaces. By arranging the shape sensor with the function of self-compensating reconstruction error in the present invention on the reflector of a spaceborne antenna, the geometric shapes of multiple key positions can be obtained in real time, overcoming the problem that the surface of a deployable space antenna cannot be measured during on-orbit operation, providing technical support for on-orbit regulation of the surface. This method has prominent advantages such as low cost, anti-electromagnetic interference, high measurement accuracy, large measurement size, and easy conformal.

[0077] The above technical solutions will be further described below with reference to the drawings and examples:

[0078] In one embodiment, referring to Figure 1 As shown, in this example, a parabolic cable-net antenna with an optical aperture of 10 m is selected, and the focal length f1 of the parabolic surface is 5 m. To ensure the high-precision working state of the antenna, it is necessary to monitor its deformation in real time and compensate for errors. On the parabolic surface structure of the antenna, in order to unify the coordinate reference for shape reconstruction analysis, a Cartesian coordinate system is established. The plane where the vertex of the parabolic antenna is located is used as the reference plane, and the vertex is set as the coordinate origin, and the xoy coordinate system shown in the figure is defined.

[0079] Specifically, two weak grating array optical fibers are symmetrically pasted and fixed on both sides of a flexible silicone substrate with a thickness of 100 microns to form a shape sensor. The grating length is 1 cm, and the central wavelength is 1540 nm. The central wavelength demodulation of the grating is performed by an optical frequency domain reflectometer and converted into structural curvature. This part of the technology is relatively mature and will not be elaborated here.

[0080] Shape sensors with four different measuring point spacings are fixed on the antenna reflector, and the spatial resolutions are 2 cm, 6 cm, 10 cm, and 14 cm respectively. Based on the arc length and the measured curvature data, shape reconstruction is performed according to the Frenet–Serret formula:

[0081]

[0082] Among them, s is the arc length of the curve, r(s) is the spatial Cartesian coordinate vector, T(s) is the unit tangent vector, N(s) is the unit normal vector, k(s) is the curvature of the shape sensor at position s, and Δs represents the recurrence step size. The initial tangent vector is set to (1, 0), pointing in the positive x-axis direction, that is, extending outward from the origin along the unfolding direction of the antenna structure; the initial normal vector is set to (0, 1), along the positive y-axis direction, perpendicular to the tangent vector and pointing to the focus.

[0083] Refer to Figure 2 As shown, the curve is reconstructed to obtain the reconstructed curves at the spatial resolution of each measurement point. In Figure 2 The enlarged figure shows the reconstructed curves at the measurement point resolutions of 14 cm, 10 cm, 6 cm, and 2 cm from left to right in sequence. Among them, the reconstructed curve with the highest measurement point resolution of 2 cm is the closest to the theoretical curve. Extract the relationship between their maximum reconstruction error and the corresponding measurement point resolution, as Figure 3 shown. It can be seen that as the measurement point resolution increases, the reconstructed curve becomes more accurate, and the maximum reconstruction error continuously decreases, gradually dropping from 33.9177 mm to 5.0357 mm.

[0084] The specific process for realizing the migration compensation of this method is as follows. First, a set of curves containing 7 curves with known curvature distributions and corresponding reconstruction error distributions is constructed, that is, the source domain, as Figure 4 shown. In this example, the curve reconstructed at the measurement point resolution of 2 cm is used as the target curve for the following migration compensation. To find the source domain curve closest to the target curve, each curve in the source domain is subjected to the following operations with the target curve to find the source domain curve with the smallest curvature distribution difference D:

[0085]

[0086] where V is the defined affine matrix,

[0087]

[0088] h is the number of curvature measurement points corresponding to the curve, is the curvature of the i-th transferred curve at the j-th point, is the curvature of the target curve at the j-th point.

[0089] After traversing all the source domain curves, it is found that curve 4 has the smallest curvature distribution difference D, and it is determined to be the curve closest to the target curve.

[0090] Finally, the reconstruction error distribution of curve 4 is migrated based on:

[0091]

[0092] Among them, is the reconstruction error distribution of curve 4. The predicted error distribution E is calculated tar , and the predicted error distribution curve is as shown in Figure 5 . As the measurement distance x increases, the curve reconstruction error also increases. The maximum reconstruction error is: 4.9831 mm. The comparison diagram of the curves before and after the error compensation using this error distribution is as shown in Figure 6 . The curve after compensation is closer to the theoretical curve than the curve before compensation. The maximum error is reduced by 5.02 mm, and the maximum error compared to the theoretical curve is only 0.05 mm.

[0093] For the above-mentioned multiple spatial measurement point resolutions, this method is used for error correction and compensation. The curve error after correction and compensation is as shown in Figure 7 . As the measurement point resolution increases, the corresponding reconstruction error decreases. The maximum correction errors for 2 cm, 6 cm, 10 cm, and 14 cm are: 0.0525 mm, 0.0843 mm, 0.1323 mm, and 0.186 mm respectively.

[0094] To evaluate the accuracy of this method in relative percentages:

[0095]

[0096] After calculation, the relationship between each spatial measurement point resolution and the percentage reduction of its maximum error is obtained, as shown in Figure 8 . It can be seen that after this migration compensation, the error is significantly reduced. Among them, the errors of the reconstructed curves with measurement point resolutions of 6 cm, 10 cm, and 14 cm are all reduced by more than 99.4%, while the error of the reconstructed curve with a measurement point resolution of 2 cm is reduced by 98.96%. It is speculated that because the reconstruction accuracy of the 2 cm measurement point resolution itself is relatively high compared to other measurement point resolutions, the percentage reduction of the error after compensation is relatively lower than that of others.

[0097] To further verify the dynamic update ability of the source domain set in this method, a target curve that is significantly different from all the curves in the current source domain set is specifically selected as a supplement. The curvature distribution difference D i between this target curve and all the curves in the source domain exceeds the set threshold of 0.39, so it is determined as a failed curve that fails to complete curve matching. The system automatically adds its error distribution and curvature distribution to the source domain set, becoming the 8th curve. After the source domain is updated, it is as shown in Figure 9 . As more curves are accessed subsequently, the source domain set is dynamically expanded and is considered together in the subsequent target curve matching process, thereby continuously improving the coverage and prediction accuracy of error migration.

[0098] The above method realizes high-precision error compensation in the case of unknown target shape, which can greatly improve the accuracy of shape measurement and lay a theoretical and technical foundation for the high-precision measurement of shape sensors in spatial structures.

[0099] In one embodiment, the online compensation method for the measurement error of the spatial shape sensor in the case of unknown measurement target is used for the deformation monitoring and compensation of the ground phased array radar front.

[0100] Example 3:

[0101] In one embodiment, the fiber optic shape sensor demodulates wavelengths based on a fiber grating array and a fast-band spectral light source.

[0102] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention without departing from the principles and purposes of the present invention.

Claims

1. An online compensation method for the measurement error of a spatial shape sensor when the target is unknown, characterized in that The specific steps are as follows: Based on the geometric model of the target to be measured, a dynamically updated source domain is constructed, wherein the source domain includes multiple sets of known curvature distributions and their corresponding shape reconstruction error distributions; The target to be measured is measured in real time by a shape sensor, the curvature distribution obtained by the measurement is constructed as a target domain, and a source domain curve closest to the curvature distribution of the target domain is selected from the source domain as a matching curve; The reconstruction error distribution of the filtered source domain curve is transferred to the target domain through affine transformation, and the reconstruction error of the current measured shape is compensated in real time.

2. The online compensation method for the measurement error of a spatial shape sensor under the condition of an unknown target according to claim 1, wherein: The specific process of constructing the dynamically updated source domain is as follows: According to the geometric model of the target to be measured, the working load is applied to generate multiple reflective surface deformation surfaces; A theoretical shape curve is defined at the sensor placement location to extract the geometric curvature with the same spatial resolution as the shape sensor; Reconstruct the shape curve based on the Frenet-Serret formula, calculate the vertical coordinate difference between the reconstructed curve and the theoretical shape curve, and obtain the reconstruction error distribution; Based on multiple reflective surface deformation surfaces, the curvature distribution of the shape sensor measurement position and its corresponding reconstruction error data set are established to obtain the initial source domain; When the difference between the curvature distribution of the target domain and all curves in the initial source domain exceeds a preset threshold, the curvature distribution of the target domain and its reconstruction error distribution are dynamically appended to the initial source domain to achieve dynamic update of the source domain.

3. The online compensation method for the measurement error of the spatial shape sensor under the condition of unknown target according to claim 2, characterized in that: The preset threshold is 0.

39. When the difference is greater than 0.39, it is determined that the target domain and the source domain curves are significantly different.

4. An on-line compensation method for the measurement error of a spatial shape sensor in the case of an unknown target according to claim 2, characterized in that: The closest source domain curve screening process includes: Calculate the affine matrix of the curvature distribution of the target domain and the curvature distribution of each curve in the source domain. The affine matrix is defined as: Among them, The subscript h represents the number of curvature measurement points corresponding to the curve, and the superscripts sou and tar represent the source domain and the target domain respectively; The source domain curve is transformed based on the affine matrix, and the curvature distribution difference between the transformed source domain curve and the target domain curve is calculated: Among them, is the curvature of the i-th curve after passing through the affine matrix in the source domain at the j-th point; is the curvature of the target curve at the j-th point; h is the number of curvature measurement points corresponding to the curve; The curvature distribution difference between each transformed source domain curve and the target domain curve is calculated, and the source domain curve with the smallest curvature distribution difference is selected as the matching curve; if all curvature distribution differences exceed the threshold, the source domain dynamic update is triggered.

5. The online compensation method for the measurement error of a spatial shape sensor under the condition of an unknown target according to claim 4, characterized in that: The expression for migrating the reconstruction error distribution of the matching curve to the target domain is: Among them, is the reconstruction error distribution of the matching curve, V is the corresponding affine matrix, and E tar is the predicted value of the reconstruction error distribution in the target domain.

6. The online compensation method for the measurement error of a spatial shape sensor under the condition of an unknown target according to claim 1, wherein: The shape sensor is a weak grating array optical fiber sensor, which demodulates the grating center wavelength through an optical frequency domain reflectometer to obtain the structural curvature.

7. An on-line compensation method for the measurement error of a spatial shape sensor under the condition of an unknown target to be measured, characterized in that: The target to be measured is a large-sized flexible antenna reflector surface on board a satellite, a ground phased array radar array, or a spatially deployable structure using a fiber grating array.

8. An on-line compensation system for the measurement error of a spatial shape sensor under unknown measurement targets, which is used to implement the on-line compensation method for the measurement error of a spatial shape sensor under unknown measurement targets according to any one of claims 1-7; characterized in that Includes the following modules: A dynamic source domain construction module is used to generate theoretical curvature distributions and corresponding reconstruction error distribution data sets under multiple deformation conditions based on the geometric model of the target to be measured, forming a dynamically updated source domain set; The curvature distribution matching module is used to receive the real-time measured curvature data from the shape sensor, calculate the affine matrix of the curvature distribution of the target domain and the curvature distribution of each curve in the source domain, and select the source domain curve with the smallest difference; An error migration and compensation module migrates the reconstructed error distribution of the matching curve to the target domain through the affine matrix, and performs real-time error compensation on the measurement results of the current shape sensor; The dynamic update module, when the difference between the curvature distribution of the target domain and all curves in the source domain exceeds a preset threshold, appends the curvature distribution of the target domain and its reconstructed error data group to the source domain set, realizing the dynamic expansion of the source domain set.

9. An electronic device, characterized in that: It includes at least one processor and a memory communicatively connected to the at least one processor; wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the online compensation method for the measurement error of the spatial shape sensor under the condition that the measurement target is unknown as recited in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the processor to execute the online compensation method for the measurement error of the spatial shape sensor under the condition that the measurement target is unknown as recited in any one of claims 1-7 when executed.

Images