Virtual-real fusion method for multi-system structure optimization and surface shape error solution

Abstract

The invention discloses a virtual-real fusion method for multi-system structure optimization and surface shape error solution, which can improve the vibration resistance of a sub-aperture splicing interference method, avoid a complex splicing algorithm, reduce the installation and adjustment difficulty of a system and correct the return error of the system. The method comprises the following steps of: (1) determining an installation and adjustment error which is high in control difficulty or has relatively high influence on image surface wavefront and a possible maximum value of the installation and adjustment error; (2) analyzing the influence of each installation and adjustment error on the wavefront of the image surface; (3) analyzing the influence of each installation and adjustment error on each Zernike coefficient, and determining a calibration scheme of the installation and adjustment error based on the influence size and the independent or coupling condition; (4) the ideal virtual interferometer is optimized to obtain the virtual interferometer after the adjustment error calibration, and the virtual interferometer has the same adjustment error as the actual interference system; and (5) based on the virtual interferometer, reconstructing the surface shape error through a synchronous reverse optimization reconstruction method, and obtaining a final measurement result.

Description

Technical Field

[0001] The present invention relates to the technical field of optoelectronic detection, and in particular to a virtual-real fusion method for optimizing the multi-system structure and solving the surface shape error. Background Art

[0002] In the field of precision detection of large-aperture, small F / # aspheric surfaces, the full-field interference detection method has always been difficult due to the design and processing difficulty of the compensator, the resolution of the interferometer, and the F / # of the spherical lens. The sub-aperture stitching interferometry is a solution proposed in the 1980s. The sub-aperture stitching interferometry divides the measured mirror into several sub-aperture regions within the measurement range of the spherical lens with a large F / # of the interferometer, and measures them sequentially, thus avoiding the limitations of the full-field interference detection method; then the measurement results of different sub-aperture regions are stitched together to obtain the surface shape error distribution of the full aperture. According to the different shapes of the sub-apertures, the sub-aperture stitching interferometry can be divided into the annular sub-aperture stitching interferometry and the circular sub-aperture stitching interferometry. The annular sub-aperture stitching interferometry is designed for rotationally symmetric aspheric surfaces. Its sub-apertures are usually multiple concentric annuli. In the measurement, spherical waves with different radii of curvature are used to match the slopes of different annuli on the measured surface, mainly expanding the longitudinal dynamic range of the interferometer. Zygo Corporation has a commercial instrument that can realize the measurement of the annular sub-aperture stitching interferometry. The circular sub-aperture stitching interferometry uses spherical waves with different curvatures to match the local surface shapes within different circular apertures of the measured surface, greatly expanding the transverse and longitudinal measurement ranges of the interferometer, and can detect planes, spheres, aspheric surfaces and free-form surfaces, and has been commercialized by QED Corporation. It can be seen that the circular sub-aperture stitching interferometry expands the transverse and longitudinal dynamic ranges of the interferometer, and can achieve high-resolution detection of large-aperture, large-slope and even off-axis aspheric surfaces without using a zero compensator.

[0003] However, the sub-aperture stitching interferometry still has deficiencies in terms of vibration resistance, stitching algorithm, alignment difficulty and error correction.

[0004] First of all, the sub-aperture stitching interferometry realizes the full-aperture surface shape measurement by dividing the measured surface into multiple regions and measuring them sequentially. Since this method requires multiple measurements and the whole process takes a long time, environmental vibration is difficult to avoid during this period, so the measurement accuracy will be affected. In addition, the step-by-step measurement of the sub-apertures usually relies on the motion control system to adjust the relative position of the measured surface or the interferometer, and the mechanical motion itself may cause additional vibration disturbances. Therefore, compared with the full-field detection method, the vibration resistance performance of the sub-aperture stitching interferometry is relatively poor. It is necessary to explore ways to improve the vibration resistance of the sub-aperture stitching interferometry.

[0005] Secondly, regardless of the sub-aperture stitching interferometry method, the stitching algorithm that fuses different sub-aperture regions to obtain the full-aperture measurement result plays an important role. Currently, there are several milestone algorithms in the field of sub-aperture stitching, such as the Kwon Thunen method, the synchronous fitting method, the discrete phase method, the multi-aperture overlapping scanning technique, and the sub-aperture stitching and positioning algorithm, etc. However, they all rely on complex mathematical calculations for positioning between sub-apertures and calibration of phase inconsistency, which are prone to introducing positioning and calculation errors. Additionally, the vibration disturbances during the aforementioned measurement process may introduce random and unknown positioning errors, increasing the complexity and errors of the algorithm. It is necessary to explore a data stitching algorithm that does not require calibration.

[0006] It can be seen that the measurement accuracy of the sub-aperture stitching method is restricted by factors such as the device structure, vibration resistance, and stitching algorithm, and the complexity of the alignment process further exacerbates this problem. To cope with the mechanical errors introduced by multiple adjustment degrees of freedom, reduce the influence of environmental vibration, and improve the accuracy of the stitching algorithm, the alignment process usually requires high-precision calibration of the pose relationship of the system. However, this high-precision calibration places higher requirements on the device accuracy and the technical level of the operator, significantly increasing the difficulty of alignment. The increase in alignment difficulty alleviates the error influence of other factors to a certain extent, but also poses more challenges to the practical application and popularization of the system. Therefore, it is urgent to explore methods to reduce the alignment difficulty while maintaining or even improving the overall measurement accuracy.

[0007] Finally, most sub-aperture stitching interferometry methods are non-zero interference detections, which means that there is a return error introduced by the non-return of the measurement light in the optical path. However, most of the above algorithms do not process the return error, which requires the sub-aperture to be small enough, resulting in an increase in the number of scans, greater cumulative stitching errors and environmental impacts; or the sub-aperture is equipped with a variable compensator, increasing the system complexity. Therefore, in order to obtain high-precision aspherical surface shape error detection results, it is necessary to process the return error. Summary of the Invention

[0008] To overcome the defects of the prior art, the technical problem to be solved by the present invention is to provide a virtual-real fusion method for multi-system structure optimization and surface shape error solution, which can improve the vibration resistance of the sub-aperture stitching interferometry method, avoid complex stitching algorithms, reduce the system alignment difficulty, and correct its return error.

[0009] The technical solution of the present invention is: This virtual-real fusion method for multi-system structure optimization and surface shape error solution includes the following steps:

[0010] (1) Determine the types of alignment errors to be considered in subsequent steps. Based on the structure of the interference system and the specific operations during alignment, determine the alignment errors with high control difficulty or significant impact on the image plane wavefront and their possible maximum values.

[0011] (2) Analyze the influence of each alignment error on the image plane wavefront. Through simulation analysis, study the influence of the alignment errors obtained in step (1) on the image plane wavefront of the interference system.

[0012] on the image plane wavefront of the interference system.

[0013] (3) Based on the influence relationships obtained in step (2), determine the subsequent calibration scheme. Analyze the influence of each alignment error on various Zernike coefficients, and based on the influence magnitude and independent or coupled situations, determine the calibration scheme for alignment errors. This scheme includes the sequence of calibrating each alignment error and the Zernike coefficients to be monitored during the calibration of each alignment error.

[0014] (4) Set the optimization objective in the ideal virtual interferometer as the image plane wavefront W of N sub - apertures obtained from actual interference measurement n , and the optimization variables are the alignment errors of each component.

[0015] According to the alignment error calibration scheme determined in step (3), optimize the ideal virtual interferometer to obtain a calibrated virtual interferometer with the same alignment errors as the actual interference system.

[0016] (5) Based on the virtual interferometer obtained in step (4), reconstruct the surface shape error through the synchronous reverse optimization reconstruction method to obtain the final measurement result.

[0017] The present invention can suppress the return error caused by partial compensation in the sub - aperture stitching interference method, can handle the situation where the structural parameters and alignment errors of multiple sub - aperture measurement systems are different, reduce the influence of alignment errors on measurement errors in sub - aperture scanning measurement, improve the overall vibration resistance of sub - aperture stitching measurement, reduce the alignment difficulty of the system, and avoid the positioning and calculation errors introduced by complex sub - aperture stitching algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flowchart of the virtual - real fusion method for multi - system structure optimization and surface shape error solution according to the present invention.

[0019] Figure 2 is a diagram of the sub - aperture stitching interference system used in a specific embodiment of the present invention.

[0020] Figure 3 is the true value of the surface shape error of the measured surface used in a specific embodiment of the present invention.

[0021] Figure 4It is the sub-aperture layout used in the specific embodiments of the present invention.

[0022] Figure 5 It is the measured surface shape error of the measured surface obtained according to the specific embodiments of the present invention.

[0023] Wherein: 1 - light source, 2 - linear polarizer, 3 - λ / 2 wave plate, 4 - collimating lens, 5 - polarization beam splitter prism, 6 - first λ / 4 wave plate, 7 - reference mirror, 8 - second λ / 4 wave plate, 9 - compensation mirror, 10 - polarization grating, 11 - third λ / 4 wave plate, 12 - measured mirror, 13 - fourth λ / 4 wave plate, 14 - lens, 15 - polarization camera. Specific embodiments

[0024] As Figure 1 shown, this virtual-real fusion method for multi-system structure optimization and surface shape error solution includes the following steps:

[0025] (1) Determine the types of alignment errors that need to be considered in subsequent steps. According to the structure of the interference system and the specific operations during alignment, determine the alignment errors with high control difficulty or significant influence on the image plane wavefront and their possible maximum values.

[0026] (2) Analyze the influence of each alignment error on the image plane wavefront. Through simulation analysis, analyze the influence of the alignment errors obtained in step (1)

[0027] on the image plane wavefront of the interference system.

[0028] (3) Based on the influence relationship obtained in step (2), determine the subsequent calibration scheme. Analyze the influence of each alignment error on each Zernike coefficient, and determine the calibration scheme for the alignment errors based on the influence magnitude and independent or coupled situations. This scheme includes the order of calibration of each alignment error and the Zernike coefficients to be monitored during the calibration of each alignment error.

[0029] (4) Set the optimization target in the ideal virtual interferometer as the image plane wavefront W n of N sub-apertures obtained from actual interference measurement, and the optimization variables are the alignment errors of each component;

[0030] According to the alignment error calibration scheme determined in step (3), optimize the ideal virtual interferometer to obtain a virtual interferometer after alignment error calibration, which has the same alignment errors as the actual interference system.

[0031] (5) Based on the virtual interferometer obtained in step (4), reconstruct the surface shape error through the synchronous reverse optimization reconstruction method to obtain the final measurement result.

[0032] The present invention can suppress the return error caused by partial compensation in the sub-aperture stitching interferometry method, can handle the situation where the structural parameters and alignment errors of multiple sub-aperture measurement systems are different, reduce the influence of alignment errors on measurement errors in sub-aperture scanning measurement, improve the overall vibration resistance of sub-aperture stitching measurement, reduce the alignment difficulty of the system, and avoid the positioning and calculation errors introduced by complex sub-aperture stitching algorithms.

[0033] Preferably, in step (3), according to the structural parameters of the interference system, two interference system models with exactly the same structure are established in the ray tracing program; then, the alignment errors obtained in step (1) are successively added to one of them to obtain a virtual interferometer with alignment errors, and the other remains in an ideal state, defined as an ideal virtual interferometer; the image plane wavefronts of the above two systems are respectively obtained and analyzed to obtain the influence of alignment errors on the image plane wavefront.

[0034] Preferably, steps (1)-(3) are carried out before actual measurement, and steps (4)-(5) are carried out after the image plane wavefronts W of N sub-apertures are obtained in actual interference measurement, where n = 1, 2,... N. n

[0035] Preferably, in step (1), the maximum values are all estimated values of the visual errors during the optical path alignment process. For some systems that may use computer-aided alignment, the maximum values here can be the errors caused by computer-aided alignment.

[0036] Preferably, in step (2), the Zernike coefficients of the image plane wavefronts of the actual interferometer and the ideal virtual interferometer are subtracted to obtain △Z i as the influence of the modeling error on the image plane wavefront, where i still represents the number of Zernike polynomial terms; each error is successively added independently to the virtual interferometer to obtain the Zernike coefficients of the image plane wavefronts of each sub-aperture caused by the alignment errors of each component when the measured spherical surface has no surface shape error. It should be noted that in step (2), the Zernike polynomial may not necessarily be used for characterization, that is, the influence of the modeling error may not necessarily be represented by the Zernike coefficients.

[0037] Preferably, in step (3), only the axial offset of the measured surface has a significant influence on the defocus terms of each sub-aperture; the axial offset of the measured surface has a significant influence on the tilt terms of the off-axis sub-apertures; the eccentricity of the measured surface only has a significant influence on the tilt terms of each sub-aperture; each alignment error of the polarization grating only has a significant influence on the tilt terms of each sub-aperture. This is the situation in the specific embodiment mentioned here.

[0038] Preferably, in step (3), the following alignment error calibration scheme is formulated for this interference optical path (this is the situation in the specific embodiment mentioned here):

[0039] ​Set the defocus terms of each sub-aperture as the optimization objective, and optimize the axial position of the measured surface.

[0040] Based on the optical path structure obtained in step (1), set the tilt terms of each sub-aperture as the optimization objective, and optimize the eccentricity, tilt of the measured surface, and various alignment errors of the polarization grating.

[0041]

[0042]

[0043] Preferably, in step (4), the Zernike polynomial is used to characterize the wavefront of the image plane, and the optimization objective in the ideal virtual interferometer is the Zernike coefficient Z obtained after fitting the wavefront W of the image plane. n ni In step (4), it is not necessarily characterized by the Zernike polynomial. There are other methods to characterize the wavefront of the image plane. What is mentioned here is the form of the Zernike coefficient used in the specific embodiment.

[0044] Preferably, in step (5), the surface shape error of the measured surface is characterized by the Zernike polynomial. In step (5), it is not necessarily characterized by the Zernike polynomial. What is mentioned here is the form of the Zernike coefficient used in the specific embodiment.

[0045] To better illustrate the purpose and advantages of the present invention, the following further describes the content of the invention with reference to the drawings and examples.

[0046] Build a sub-aperture stitching interferometric measurement system as shown in Figure 2 to measure the surface shape error of the measured mirror, including a light source 1, a linear polarizer 2, a λ / 2 wave plate 3, a collimating lens 4, a polarization beam splitter prism 5, a first λ / 4 wave plate 6, a reference mirror 7, a second λ / 4 wave plate 8, a compensating mirror 9, a polarization grating 10, a third λ / 4 wave plate 11, a measured mirror 12, a fourth λ / 4 wave plate 13, a lens 14, and a polarization camera 15. The true value of the surface shape error of the measured mirror is as shown in Figure 3 Figure 4 shown. After collecting the sub-aperture data, perform multiple system structure optimization and surface shape error solution according to the method proposed in the present invention. The steps are as follows:

[0047] 1. Determine the alignment error, and determine the types of alignment errors that need to be considered in the subsequent steps.

[0048] According to the structure of the interference system and the specific operations during the alignment process, determine the alignment errors with large control difficulty or significant influence on the wavefront of the image plane and their possible maximum values.

[0049] ​​​​Most of the components in the optical path are located in the planar optical waveguide, which has low alignment difficulty and has little influence on the image plane wavefront. Therefore, the components in the spherical optical path are analyzed. In the experiment, the polarization grating and the third λ / 4 wave plate are installed in the sleeve through a snap ring. Therefore, the alignment errors of the two are considered together. The eccentricity and tilt of the measured surface will compensate for each other's influence on the interference pattern. In other words, their influence on the image plane wavefront is similar, and they will be optimized simultaneously during the subsequent alignment error calibration process. Therefore, only the eccentricity error is considered here.

[0050] From the inference of the optical path structure and the alignment process in the experiment, the alignment errors of the components in the optical path and their possible maximum values are shown in Table 1, where the maximum values are all estimated values of the visual errors during the optical path alignment process.

[0051] Table 1 Alignment errors of each component and their possible maximum values

[0052]

[0053] 2. Analysis of the influence of each alignment error on the image plane wavefront. Through simulation, analyze the influence of the alignment errors obtained in step 1 on the image plane wavefront of the interference system.

[0054] According to the structural parameters of the interference system, two interference system models with exactly the same structure are established in the ray tracing program. Then, the alignment errors obtained in step 1 are successively added to one of them to obtain a virtual interferometer with alignment errors, and the other is still in an ideal state, defined as an ideal virtual interferometer. The image plane wavefronts of the above two systems are obtained and analyzed respectively to obtain the influence of the alignment error on the image plane wavefront.

[0055] To quantitatively analyze and unify the optimization criteria, the embodiment uses Zernike polynomials to characterize the image plane wavefront. The specific process of the above analysis is to fit the image plane wavefronts of the two systems to obtain the Zernike coefficients under the influence of alignment errors and in the ideal state. The difference between the two can obtain the influence of the alignment error on the image plane wavefront, which is quantified in the form of Zernike coefficients.

[0056] By subtracting the Zernike coefficients of the image plane wavefronts of the actual interferometer and the ideal virtual interferometer, △Z can be obtained. iTo model the influence of the modeling error on the image-plane wavefront, where i still represents the number of Zernike polynomial terms. Each error item in Table 1 is independently added to the virtual interferometer in sequence. When the measured spherical surface has no surface shape error, the influence of the alignment error of each component on the Zernike coefficients of the image-plane wavefront of each sub-aperture is shown in Table 2. Since the influence of the same alignment error on the central and off-axis sub-apertures is not the same, and the change amounts introduced to the same Zernike coefficient of the image-plane wavefront of different off-axis sub-apertures are not exactly equal, but of the same order of magnitude. For the sake of concise expression, Table 2 lists the order-of-magnitude changes in the Zernike coefficients of the image-plane wavefront of the central and off-axis sub-apertures caused by the alignment error, without listing the specific change values, and is represented by the central △Z i and the off-axis △Z i respectively. In addition, items 5 - 36 of the Zernike coefficients of the image-plane wavefront are basically all below 1 / 10 compared with the items listed in the table, so they are not listed.

[0057] Table 2 Influence of the alignment error of each component on the Zernike coefficients of the image-plane wavefront and its order of magnitude (λ)

[0058]

[0059] Note: The "-" item in the table indicates that the influence of this alignment error on the coefficient of this Zernike term is below 1 / 10 compared with other terms

[0060] below

[0061] 3. Determination of the alignment error calibration scheme. Based on the influence relationship obtained in step 2, determine the subsequent calibration scheme.

[0062] Analyze the influence of each alignment error on each Zernike coefficient, and determine the calibration scheme for the alignment error based on the influence magnitude and the independent or coupled situation. This scheme includes but is not limited to the sequence of calibrating each alignment error, the Zernike coefficients that should be monitored during the calibration of each alignment error, etc.

[0063] It can be obtained from Table 2 that:

[0064] ① Analyzing the defocus terms in the last two columns, it can be seen that only the axial offset of the measured surface has a significant impact on the defocus terms of each sub-aperture;

[0065] ② Further analyzing the axial offset of the measured surface, it can be seen that the axial offset of the measured surface will also have a significant impact on the tilt terms of the off-axis sub-apertures;

[0066] ③ Analyzing the eccentricity of the measured surface, it can be seen that it only has a significant impact on the tilt terms of each sub-aperture;

[0067] ④ Analyzing the alignment errors of the polarization gratings in the first four rows, it can be seen that they all only have a significant impact on the tilt terms of the off-axis sub-apertures.

[0068] Based on the above conclusions, the alignment error calibration scheme for this interference optical path is formulated as follows:

[0069] ① Set the defocus terms of each sub-aperture as the optimization objective, and optimize the axial position of the measured surface.

[0070] ② Based on the optical path structure obtained in ①, set the tilt terms of each sub-aperture as the optimization objective, and optimize the eccentricity and tilt of the measured surface, as well as the alignment errors of each polarization grating.

[0071] 4. Alignment error calibration: Based on the calibration scheme determined in step 3, calibrate the alignment errors through reverse optimization in the ray tracing program.

[0072] Set the optimization objective in the ideal virtual interferometer as the image plane wavefront W of N sub-apertures obtained by actual interference measurement n , and the optimization variables are the alignment errors of each component. According to the alignment error calibration scheme determined in step 3, optimize the ideal virtual interferometer to obtain a virtual interferometer after alignment error calibration, that is, a virtual interferometer with the same alignment errors as the actual interference system.

[0073] As mentioned above, for quantitative analysis and unified optimization criteria, the embodiment uses Zernike polynomials to characterize the image plane wavefront. Therefore, the optimization objective in the ideal virtual interferometer in the embodiment is the Zernike coefficients Z n obtained after fitting the image plane wavefront W ni .

[0074] 5. Surface shape error reconstruction: Based on the virtual interferometer obtained in step 4, reconstruct the surface shape error through synchronous reverse optimization reconstruction method to obtain the final measurement result.

[0075] Similarly, for quantitative analysis and unified optimization criteria, the surface shape error of the measured surface in this article is also characterized by Zernike polynomials, and the obtained surface shape error of the measured surface is as Figure 5 shown. It can be seen that it is consistent with the true distribution of the surface shape error shown in Figure 3 , and the PV difference is only 0.0889λ, and the RMS difference is only 0.0027λ.

[0076] The above are only the preferred embodiments of the present invention, and do not impose any form of limitation on the present invention. Any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A virtual-real fusion method for multi-system structural optimization and surface error solution, characterized by: It includes the following steps: (1) Determine the types of alignment errors that need to be taken into consideration in the subsequent steps. According to the structure of the interferometer system and the specific operations in the alignment process, determine the alignment errors that are difficult to control or have a greater impact on the image wavefront and their possible maximum values; (2) Analysis of the influence of various adjustment errors on the image wavefront, by simulating and analyzing the influence of the adjustment errors obtained in step (1) on the image wavefront of the interferometer system; (3) determining a subsequent calibration scheme based on the influence relationship obtained in step (2), analyzing the influence of each installation error on each Zernike coefficient, and determining a calibration scheme for the installation error based on the influence size and the independent or coupled situation, the scheme including the order of calibration of each installation error and the Zernike coefficient to be monitored during the calibration of each installation error; (4) The optimization target in the ideal virtual interferometer is set to the image wavefront W of the N sub-apertures obtained by actual interferometric measurement. n , the optimization variable is the adjustment error of each component; according to the adjustment error calibration scheme determined in step (3), the ideal virtual interferometer is optimized to obtain a virtual interferometer after adjustment error calibration, which has the same adjustment error as the actual interference system; (5) Based on the virtual interferometer obtained in step (4), the surface error is reconstructed by a synchronous inverse optimization reconstruction method to obtain the final measurement result.

2. The virtual-real fusion method for multi-system structure optimization and surface error solution according to claim 1, characterized in that: In the step (3), two interference system models with completely identical structures are established in a ray tracing program according to the interference system structural parameters; then the adjustment errors obtained in step (1) are added to one of them in turn to obtain a virtual interferometer with adjustment errors, and the other one is still in an ideal state and is defined as an ideal virtual interferometer; the image wavefronts of the two systems are respectively obtained and analyzed to obtain the influence of the adjustment errors on the image wavefront.

3. The virtual-real fusion method for multi-system structure optimization and surface error solution according to claim 1, characterized in that: The steps (1) to (3) are performed before the actual measurement, and the steps (4) to (5) are performed after the actual interferometric measurement to obtain the image wavefronts of the N sub-apertures. W n Then proceed, n=1,2,…N.

4. The virtual-real fusion method for multi-system structure optimization and surface error solution according to claim 3 is characterized in that: In the step (1), the maximum values ​​are all estimated values ​​of the visual errors during the optical path adjustment process.

5. The virtual-real fusion method for multi-system structure optimization and surface error solution according to claim 4, characterized in that: In step (2), the image wavefront Zernike coefficients of the actual interferometer and the ideal virtual interferometer are subtracted to obtain ΔZ i To model the influence of the error on the image wavefront, i still represents the number of Zernike polynomial terms; each error is independently added to the virtual interferometer in turn, and the Zernike coefficient of each component adjustment error on the image wavefront of each sub-aperture is obtained when the measured spherical surface does not contain surface error.

6. The virtual-real fusion method for multi-system structure optimization and surface error solution according to claim 5, characterized in that: In the step (3), the only factor that significantly affects the defocus term of each sub-aperture is the axial offset of the measured surface; the axial offset of the measured surface significantly affects the tilt term of the off-axis sub-aperture; the eccentricity of the measured surface only significantly affects the tilt term of each sub-aperture; and the adjustment errors of the polarization grating only significantly affect the tilt term of each sub-aperture.

7. The virtual-real fusion method for multi-system structure optimization and surface error solution according to claim 6, characterized in that: In step (3), a calibration scheme for the adjustment error is formulated for the interference optical path as follows: The defocus items of each sub-aperture are set as the optimization target, and the axial position of the measured surface is Based on the optical path structure obtained in step (1), the tilt items of each sub-aperture are set The optimization goal is to adjust the eccentricity and tilt of the measured surface and the adjustment errors of the polarization grating. Difference optimization.

8. The virtual-real fusion method for multi-system structure optimization and surface error solution according to claim 7, characterized in that: In step (4), the image wavefront is characterized by the Zernike polynomial, and the optimization target in the ideal virtual interferometer is the image wavefront W n The Zernike coefficient Z obtained after fitting ni .

9. The virtual-real fusion method for multi-system structure optimization and surface error solution according to claim 8, characterized in that: In the step (5), the surface error of the measured surface is characterized by using Zernike polynomials.

Images