A method for defuzzing translation errors of spliced ​​gratings

By using optical right-angle alignment and negative pressure adsorption technology, combined with non-zero order interference fringe image processing, the problems of integer period ambiguity and multi-degree-of-freedom coupling in grating splicing were solved, realizing high-precision alignment and frictionless orientation adjustment of the grating, ensuring the accuracy of splicing and structural integrity.

CN123113166APending Publication Date: 2026-11-17JINAN UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611292432.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-25
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing grating splicing technology suffers from problems such as integer period ambiguity making it impossible to identify absolute translation errors of the grating, and splicing alignment failures caused by multi-spatial degree-of-freedom coupling interference.

Method used

The grating is attached to the movable support stage by air pressure extraction. The edge contour features are extracted by optical right-angle gauge. Combined with non-zero order interference fringe image and comprehensive measurement uncertainty, the in-plane angle deviation and translation error are calculated. Negative pressure adsorption and non-contact displacement compensation are adopted to prevent mechanical assembly stress interference, so as to achieve frictionless directional adjustment and two-dimensional coordinate locking.

Benefits of technology

It achieves absolute translation error identification in grating splicing, improves splicing accuracy, avoids physical collisions and friction interference of gratings, and ensures the integrity of grating structure and measurement accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123113166A_ABST
    Figure CN123113166A_ABST
Patent Text Reader

Abstract

This invention relates to the field of optical precision manufacturing technology and discloses a method for resolving translation errors in spliced ​​gratings. The method includes: attaching two gratings to a movable support stage via air pressure to achieve coplanarity under physical constraints; driving the second grating towards the first grating, bringing the angular deviation and translation error into the interferometric field-of-view measurement capture range; generating a non-zero-order interference fringe image, extracting spatial periodic data to calculate and compensate for in-plane angular deviations until the interference fringes are of equal spatial width; obtaining the real-time edge spacing and phase offset to synthesize an initial translation error estimate; based on a preset comprehensive measurement uncertainty, uniquely determining the integer periodic ambiguity term of the interference phase to zero; calculating the absolute translation error value and completing directional compensation; and finally locking the in-plane two-dimensional coordinates of the grating. This invention solves the phase ambiguity problem in grating splicing and achieves high-precision two-dimensional planar splicing.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of optical precision manufacturing technology, specifically to a method for resolving ambiguity in the translation error of spliced ​​gratings. Background Technology

[0002] With the development of high-resolution spectrometers and high-power laser systems, the demand for large-size diffraction gratings is increasing. Due to the limitations of the manufacturing aperture of a single grating, mechanical splicing technology is an effective way to combine multiple sub-gratings to obtain large-aperture gratings. The core of grating splicing is to align multiple sub-gratings in spatial orientation with high precision to ensure that the spliced ​​grating maintains the equivalent diffraction characteristics of a single grating. This places sub-micron or even nanometer-level requirements on the in-plane angle and translational alignment accuracy between gratings.

[0003] Existing methods for detecting splicing errors largely rely on optical interferometry. However, in practice, interferometry has an inherent integer periodic ambiguity defect. When the relative translation distance between two gratings exceeds the order of the interferometry wavelength or the grating constant, the resulting non-zero order interference fringes will exhibit periodic repetition. Conventional phase unpacking algorithms can only obtain the principal phase and cannot identify the true interference order. This makes it impossible for the measurement system to distinguish between the actual absolute translation error and the translation error that differs by an integer multiple of the grating constant, directly outputting incorrect displacement compensation commands to the control layer, resulting in splicing failure.

[0004] Furthermore, traditional grating stitching systems involve the coupling and coordination of multiple spatial degrees of freedom during the adjustment process. The out-of-plane tilt and out-of-plane height deviation of the grating introduce additional parasitic phases. These parasitic phases are mixed and superimposed with the phases generated by in-plane translation. Directly extracting interference features without physical dimensionality reduction and decoupling results in impure spatial periodic distribution data. This causes the calculated results of in-plane angle deviation and translation error to deviate significantly from the true values, reducing the final stitching accuracy. At the same time, in the process of driving the gratings to approach each other to the interference capture range, conventional control methods lack anti-collision braking mechanisms based on actual physical spacing and closed-loop accuracy. During micron-level stroke fine-tuning, physical collisions of the gratings often occur, resulting in damage to the substrate structure. Moreover, after completing the pose adjustment, there is a lack of stress-free locking means for the final in-plane two-dimensional coordinate position, causing the alignment state to slip and fail after the driving force is removed. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for de-ambiguing translation errors in splicing gratings. This method solves the problems of existing interferometry techniques being unable to identify absolute translation errors of gratings due to integer periodic ambiguity, and splicing alignment failures caused by multi-spatial degree-of-freedom coupling interference.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for deblurring translation errors of spliced ​​gratings, the method comprising the following steps: The first grating and the second grating are attached to the corresponding movable bearing surfaces of the first movable bearing platform and the second movable bearing platform by air pressure. The first movable bearing platform outputs the rotation of the first grating around the normal direction, and the second movable bearing platform outputs the in-plane translation and rotation around the normal direction of the second grating. Drive the second grating to move in the direction of the first grating, so that the in-plane angle deviation and translation error are brought into the measurement capture range of the interferometric field of view; Generate a non-zero order interference fringe image corresponding to the splicing region of the first grating and the second grating. The non-zero order interference fringe image includes a first interferometric sub-image and a second interferometric sub-image. Extract the spatial periodic distribution data and principal phase distribution data from the non-zero order interference fringe image. Based on the spatial periodic distribution data, the in-plane angular deviation is calculated, and the second grating is driven to perform in-plane angular displacement compensation until the non-zero order interference fringes present a state of equal spatial width. The real-time edge spacing and the offset of the two edge phase references are algebraically synthesized to obtain the initial translation error estimate in the vector direction. Based on the real-time edge spacing, the second grating is driven to approach the first grating, so that the sum of the absolute value of the initial translation error estimate and the preset comprehensive measurement uncertainty is less than the grating constant of the spliced ​​grating divided by twice the absolute value of the diffraction order, so that the integer periodic ambiguity term corresponding to the interference phase is uniquely determined to be zero. The absolute translation error value in the vector direction is calculated, and the second grating is driven to perform planar translation orientation compensation to lock the in-plane two-dimensional coordinate positions of the first grating and the second grating.

[0007] This method establishes a mathematical correlation between mechanical parameter extraction and optical interferometry demodulation to construct an initial translation error estimate. It introduces comprehensive measurement uncertainty as a judgment boundary and determines that the integer periodic fuzzy term is zero when specific conditions are met, directly outputting the absolute translation error. Simultaneously, it combines negative pressure adsorption clamping and non-contact displacement compensation to prevent mechanical assembly stress from interfering with the optical surface shape, thus completing frictionless orientation adjustment and two-dimensional coordinate locking.

[0008] Furthermore, the step of driving the second grating to move towards the first grating so that the in-plane angular deviation and translation error enter the measurement capture range of the interferometric field of view includes: Obtain the relative position image of the edge of the first grating and the edge of the optical right-angle guide, extract the edge contour features in the relative position image and calculate the in-plane angle deviation; The first grating is driven to perform planar rotation based on the in-plane angle deviation until the edge of the first grating is parallel to the edge of the optical right angle guide. Acquire the relative position image between the second grating and the first grating, and extract the spacing data and parallelism data from the relative position image; The second grating is driven to move towards the direction of the first grating based on the spacing and parallelism data.

[0009] By using an optical right-angle gauge as a physical reference to extract the geometric edge contour, the initial spatial position alignment of the two gratings is completed, and the macroscopic initial assembly deviation is converged to the micrometer level corresponding to the interferometer measurement capture range, ensuring that the diffracted return light enters the interferometer field of view.

[0010] Furthermore, the step of driving the second grating to move towards the first grating based on the spacing data and parallelism data includes: The system's underlying layer presets a collision safety distance threshold in the range of 10 to 50 micrometers, corresponding to the closed-loop positioning accuracy. When the spacing data decreases to the level of the collision avoidance safety distance threshold, the motion controller outputs a braking command to ensure that the second grating and the first grating are physically suspended and isolated throughout the process.

[0011] By setting a collision avoidance buffer at the position boundary at the bottom layer of the system, the path of mechanical collision during the nanoscale feeding process is cut off, preventing the grating substrates from being squeezed together and generating contact stress or physical damage, thus ensuring the safety of the closed-loop debugging process and the integrity of the grating structure.

[0012] Furthermore, the step of generating a non-zero order interference fringe image corresponding to the stitching region of the first and second gratings includes: The reference module carrying the first and second gratings is driven to perform a rotation around the axis according to the set target incident angle. Receive the return light of the same non-zero diffraction order generated by the first grating and the second grating in the splicing area; The beam splitter assembly combines the return light of the same non-zero diffraction order with the reference beam formed by the reflection of the reference mirror. The detection module processes the beam after the return light of the same non-zero diffraction order is combined with the reference beam to generate a non-zero order interference fringe image.

[0013] This step involves separating the return light of a specific diffraction order and interfering it with the reference light. This process converts the minute displacements and angular deflections within the grating surface into information on the periodic distribution and phase changes of the interference fringes, thereby enabling the extraction of splicing errors.

[0014] Furthermore, the step of calculating the in-plane angular deviation based on the spatial periodic distribution data includes: Based on the spatial frequency difference or phase gradient difference between the first and second interferometric images, and combined with the optical magnification of the system imaging relay component, physical scale conversion is performed to obtain the object-side equivalent fringe spacing value. The main control computer calculates and outputs the absolute value of the inward angle deviation under the current state based on the set wavelength value and the extracted object equivalent fringe spacing value.

[0015] This step is based on Fourier optics theory, establishing a quantitative mapping relationship between the spatial domain characteristics of interference fringes and the physical rotation angle of the grating. The equivalent fringe spacing in the object space is calculated by combining the spatial frequency difference and phase gradient difference with the optical magnification, and the angular deviation value is output to support the servo system to perform reverse angular displacement compensation.

[0016] Furthermore, the preset comprehensive measurement uncertainty is composed of the positioning error of the multi-axis high-precision piezoelectric displacement stage, the measurement error of the capacitive displacement sensor, the calibration error of the edge phase reference offset, and the dimensional change error caused by temperature changes.

[0017] By setting a tolerance boundary for comprehensive measurement uncertainty, when the sum of the predicted deviation and the system uncertainty limit is within the range of the interference period, multiple solutions are eliminated, the fuzzy term of the interference phase is determined to be zero, and the multiple uncertainties of the system solution are eliminated.

[0018] Furthermore, the step of calculating the absolute translation error value of the vector direction includes: Based on the unique determining condition that the periodic ambiguity term is zero, the main control computer performs phase demodulation on the first and second interferometric sub-images respectively; The absolute translation error of the first grating and the second grating in the vector direction is calculated directly using the principal phase difference between the first grating region corresponding to the first interferometer image and the second grating region corresponding to the second interferometer image.

[0019] The relevant steps simplify the phase unpacking operation under the condition that the fuzzy term is zero. Only the principal phase of the two grating regions is extracted for differential calculation, and the absolute translation error is directly output, which reduces the computational load and improves the real-time performance of the error output.

[0020] Furthermore, the step of attaching the first grating and the second grating to the corresponding movable bearing surfaces of the first movable bearing platform and the second movable bearing platform by air pressure includes: The microporous array negative pressure adsorption structure, which is respectively set inside the first movable support platform and the second movable support platform, is activated. Negative pressure air chambers are constructed between the bottom surfaces of the first and second gratings and their corresponding movable bearing surfaces by air pressure extraction. When the air pressure in the negative pressure chamber drops and enters the first negative pressure range of -80kPa to -50kPa, the external atmospheric pressure applies a sufficient vertically downward uniformly distributed load through the top surfaces of the first and second gratings. Under uniformly distributed load, the first grating and the second grating undergo elastic micro-deformation and attach the first grating and the second grating to the corresponding movable bearing surface.

[0021] This step utilizes fluid dynamics principles to establish a uniformly distributed air pressure load instead of mechanical clamping. External atmospheric pressure provides a large-area uniform downward pressure, causing the grating substrate to undergo controlled micro-elastic deformation to conform to the movable bearing surface, eliminating stress concentration points caused by local mechanical clamping, and maintaining the flatness of the grating surface microstructure.

[0022] Furthermore, the step of locking the in-plane two-dimensional coordinate positions of the first grating and the second grating includes: The uncured adhesive material is extruded upward through microfluidic air pressure channels arranged at the bottom of the bridging adhesive groove of the corresponding movable bearing surface; The upward pressure difference generated by the extrusion pressure in the microfluidic gas channel is strictly less than the absolute value of the downward adsorption pressure difference generated by the vacuum adsorption pressure in the negative pressure chamber, so that the top surface of the uncured adhesive material is in full contact with the bottom surfaces of the first grating and the second grating. The bonding material at the bottom of the first and second gratings is cured by ultraviolet light irradiation, thereby locking the in-plane two-dimensional coordinate positions of the first and second gratings.

[0023] By controlling the bottom extrusion air pressure to be lower than the surface vacuum adsorption pressure during the bonding and fixing process, the adhesive material fills the contact gap in the opposite direction, blocking the upward thrust that would disrupt the negative pressure adsorption state. Combined with UV irradiation to cure the contact state, the grating assembly posture is locked.

[0024] Furthermore, the step of maintaining the spatially equal width of the non-zero order interference fringes includes: The spatial equal width state is defined as follows: after normalization processing based on the system preset non-zero constant, the fringe spatial frequency difference, fringe direction difference, and phase gradient difference between the first interferometric sub-image and the second interferometric sub-image are all less than the corresponding system-set dimensionless threshold of 0.01 to 0.05, and the absolute value of the in-plane angle deviation updated in real time is less than the system-set limit angle threshold. When the space is equal in width, the motion controller outputs a voltage holding command to determine that the in-plane angle alignment is complete.

[0025] By formulating a closed-loop convergence termination criterion, setting an error tolerance for the spatial dimension parameters of the stripe image, and converting the optical parameters into a servo system trigger signal, the output voltage is maintained when the threshold condition is met, the displacement adjustment of the servo motor is stopped, and the state is solidified.

[0026] This invention provides a method for resolving translation errors in spliced ​​gratings. It offers the following advantages: 1. This invention constructs an initial translation error estimate by algebraically synthesizing the real-time edge spacing and edge phase reference offset. Using a preset comprehensive measurement uncertainty as the judgment boundary, the integer periodic fuzzy term corresponding to the interference phase is uniquely determined to be zero. The system eliminates the uncertainty of multiple solutions in large-span translation error interferometry based on the limited boundary conditions. It does not require complex phase unpacking and iteration operations. The absolute translation error value can be linearly calculated directly using the principal phase difference, reducing the computational load of the main control computer and improving the real-time output of error data.

[0027] 2. This invention utilizes a microporous array negative pressure adsorption structure and external atmospheric pressure to construct a uniformly distributed load, causing the grating substrate to undergo controlled elastic micro-deformation and adhere to the bearing surface, replacing the traditional mechanical rigid clamping. In the locking process, the upward extrusion pressure of the microfluidic channel is strictly controlled to be less than the vacuum adsorption pressure of the negative pressure chamber, allowing the uncured adhesive material to fill the contact gap without disrupting the negative pressure balance. The fluid dynamic coordination between the microfluidic air pressure channel and the microporous array negative pressure adsorption structure, combined with the ultraviolet curing operation, eliminates stress concentration caused by local clamping, maintains the flatness of the grating microstructure, and achieves displacement-free physical locking of the in-plane two-dimensional coordinates.

[0028] 3. This invention uses an optical right-angle guide to extract edge contour features for initial spatial alignment and sets a collision safety distance threshold at the system's bottom layer as the braking trigger condition for the motion controller. The system guides the servo actuator to perform displacement compensation through spacing and parallelism data, ensuring that the grating is always in a physically suspended and isolated state during the optimization and adjustment of non-zero order interference fringes. The bottom-layer braking mechanism is combined with non-contact closed-loop control to cut off the mechanical collision path during the assembly and adjustment process, prevent the grating substrate from being squeezed together and causing damage, and eliminate the interference of positioning residuals caused by mechanical friction on the accuracy of interferometric measurements. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the system timing interaction of the present invention; Figure 2 This is a schematic diagram of the method flow of the present invention; Figure 3 This is a schematic diagram of the homologous substrate segmentation and the fabrication of the first and second gratings according to the present invention; Figure 4 This is a schematic diagram of the five-degree-of-freedom splicing error coordinates of the first and second gratings of the present invention; Figure 5 This is a schematic diagram of the grating initial alignment and interference field-of-view acquisition process of the present invention; Figure 6 This is a schematic diagram of the light intensity distribution of the interference fringes in this invention. Figure 7This is a schematic diagram of the error convergence polygonal line for the in-plane angular displacement closed-loop compensation of the present invention. Figure 8 This is a schematic diagram of the in-plane error correction process for the non-zero order interference fringes of the present invention; wherein, (a) is a schematic diagram of the state of the non-zero order interference fringes when the first grating and the second grating have in-plane angular deviation and translation error at the same time, (b) is a schematic diagram of the state when the non-zero order interference fringes on both sides reach the same spatial width after the in-plane angular deviation correction is completed but there is still phase misalignment, and (c) is a schematic diagram of the state when the non-zero order interference fringes on both sides reach the phase complete alignment after the translation error correction is completed. Figure 9 This is a schematic diagram of the translation error calculation and directional compensation solidification process of the present invention; Figure 10 This is a schematic diagram of the nanoimprint replication of the splicing grating master of the present invention; Figure 11 This is a schematic diagram of the biaxial convergence state curve for error decoupling compensation according to the present invention. Detailed Implementation

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] See attached document Figure 1 The present invention provides a system for resolving translation errors of spliced ​​gratings. The system includes a reference module, a support module, an observation module, an adjustment module, a detection module, a solution module, and a solidification unit.

[0032] The reference module carries a first grating and a second grating prepared by dividing the same substrate. The reference module is provided with a first movable support stage and a second movable support stage. The first grating and the second grating are respectively placed on the upper surfaces of the first movable support stage and the second movable support stage. A bridging bonding groove is provided at the splicing junction of the upper surfaces of the first movable support stage and the second movable support stage. The bottom of the bridging bonding groove is connected to a microfluidic air pressure channel. The microfluidic air pressure channel is independently connected to an external air pressure control device. The microfluidic air pressure channel is used to squeeze the uncured bonding material upward after the non-zero order interference fringes reach a state of complete phase alignment.

[0033] The first and second movable support platforms are respectively equipped with microporous array negative pressure adsorption structures. The microporous array negative pressure adsorption structures attach the first and second gratings to the corresponding movable support surfaces by air pressure, thereby restricting the longitudinal translation of the first and second gratings relative to the corresponding movable support surfaces along the normal direction and the rotational displacement around the grating vector and scribe line directions, while retaining the in-plane translational degree of freedom and the rotational degree of freedom around the normal direction output by the movable support platform.

[0034] The carrier module is connected to the reference module, and the reference module is mounted on the top bearing surface of the carrier module. The carrier module drives the reference module to perform a rotation around the axis according to the set target incident angle, thereby adjusting the spatial attitude of the first grating and the second grating fixed on the reference module.

[0035] The observation module is located in the upper side region of the reference module to acquire surface and edge images of the first and second gratings. The observation module integrates a displacement sensing unit and includes an optical right-angle guide, an image acquisition component, and a displacement sensing unit. The image acquisition component extracts the relative position data between the edge of the first grating and the edge of the optical right-angle guide; at the same time, the image acquisition component extracts the preliminary spacing data and parallelism data between the second and first gratings. The displacement sensing unit is used to acquire the physical edge spacing between the first and second movable support platforms in real time. The main control computer calculates the real-time edge spacing between the first and second gratings based on the equivalent conversion of the physical edge spacing between the first and second movable support platforms.

[0036] The adjustment module is mounted on the base of the reference module, and the first movable support platform and the second movable support platform are respectively mounted on the adjustment module. The adjustment module receives the preliminary spacing data and parallelism data output by the observation module. Then, the adjustment module drives the first movable support platform to drive the first grating to complete the initial orientation calibration and keep it locked. Then, it drives the second movable support platform to drive the second grating to perform in-plane displacement. During the movement, the first grating and the second grating remain attached to the corresponding movable support surface to avoid friction and sliding of the bottom surface of the grating relative to the movable support surface, so as to change the in-plane position coordinates of the second grating relative to the first grating.

[0037] The detection module emits a single-wavelength detection beam toward the region containing the first and second gratings. The detection module integrates a beam splitter assembly and a reference mirror. The detection module uses the beam splitter assembly to receive the same non-zero diffraction order return light generated by the first and second gratings in the splicing region. The beam splitter assembly combines the same non-zero diffraction order return light with the reference beam formed by the reflection of the reference mirror. The detection module processes the beam after combining the same non-zero diffraction order return light and the reference beam to generate a non-zero order interference fringe image. Then, the detection module extracts the spatial periodic distribution data and phase distribution data from the non-zero order interference fringe image and simultaneously transmits the spatial periodic distribution data and phase distribution data to the solution module.

[0038] The calculation module is connected to both the detection module and the adjustment module. The calculation module receives the spatial periodic distribution data output by the detection module and calculates the in-plane angular deviation between the first and second gratings based on the spatial periodic distribution data. At the same time, the calculation module receives the phase distribution data output by the detection module and filters out the inherent phase and aberration interference terms from the phase distribution data. Then, when the adjustment module mechanically approximates the second grating to the first grating, and the mechanical positioning result and the edge phase reference calibration result together satisfy the half-period uniqueness constraint condition, the calculation module determines the integer period ambiguity term to zero and calculates the absolute translation error value of the first and second gratings in the vector direction based on the principal phase difference between the first and second grating regions.

[0039] Subsequently, the calculation module generates control feedback commands based on the in-plane angle deviation and absolute translation error values ​​and sends them to the adjustment module. Finally, the adjustment module receives the control feedback commands and drives the second grating to output the corresponding in-plane angle displacement and plane translation compensation actions until the non-zero order interference fringes corresponding to the first and second gratings present a state of equal spatial width and phase alignment.

[0040] In addition, the system is equipped with a curing unit that works in conjunction with the reference module. When the first grating and the second grating reach a phase alignment state, the curing unit emits a curing beam to irradiate the reference module, curing the adhesive material at the bottom splicing gap of the first grating and the second grating. This locks the in-plane two-dimensional coordinate position of the first grating and the second grating, and makes the first grating and the second grating physically bonded together into a complete master structure through the cured adhesive material.

[0041] See attached document Figure 2 This invention provides a method for resolving translation errors in spliced ​​gratings, the method comprising the following steps: S1, Obtain a first substrate and a second substrate prepared by dividing a substrate from the same source substrate, and process a first grating and a second grating with the target grating features on the surfaces of the first substrate and the second substrate, and control the total structural thickness of the first grating and the second grating to be consistent; Activate the micropore array negative pressure adsorption structure respectively set inside the first movable support stage and the second movable support stage, and attach the first grating and the second grating to the corresponding movable support surface by air pressure, thereby restricting the longitudinal translation of the first grating and the rotational displacement of the second grating relative to the corresponding movable support surface along the normal direction and around the grating vector and scribe line direction. The first movable support stage is used to drive the first grating to rotate around the normal direction, and the second movable support stage is used to drive the second grating to perform in-plane translation and rotation around the normal direction.

[0042] S2, acquire the relative position image of the edge of the first grating and the edge of the optical right-angle guide, extract the edge contour features in the relative position image and calculate the in-plane angle deviation; drive the first grating to perform planar rotation according to the in-plane angle deviation until the edge of the first grating is parallel to the edge of the optical right-angle guide; then acquire the relative position image between the second grating and the first grating, extract the spacing data and parallelism data in the relative position image; drive the second grating to move towards the first grating according to the spacing data and parallelism data, and maintain the non-contact state between the second grating and the first grating during the movement, so that the in-plane angle deviation and translation error between the second grating and the first grating enter the measurement capture range of the interference field of view.

[0043] S3, emit a single-wavelength detection beam toward the surface area where the first grating and the second grating are located, and drive the reference module carrying the first grating and the second grating to perform an axis rotation action according to the set target incident angle; receive the same non-zero diffraction order return light generated by the first grating and the second grating in the splicing area, process the same non-zero diffraction order return light, generate a non-zero order interference fringe image corresponding to the splicing area of ​​the first grating and the second grating; extract the spatial periodic distribution data and principal phase distribution data in the non-zero order interference fringe image.

[0044] S4: Receive the spatial periodic distribution data corresponding to the non-zero order interference fringe image, and calculate the in-plane angular deviation between the first and second gratings based on the spatial periodic distribution data; generate an angle correction command based on the in-plane angular deviation to drive the second grating to perform in-plane angular displacement compensation; cyclically execute the spatial periodic distribution data extraction and angular displacement compensation actions until the non-zero order interference fringes corresponding to the first and second gratings show a state of equal spatial width, and determine that the in-plane angular alignment is complete. At this time, the in-plane angular deviation has converged to within the allowable angular error range, providing a single-phase environment for translation error calculation.

[0045] S5: Receive principal phase distribution data corresponding to the non-zero order interference fringe image under equal spatial width conditions; perform mechanical positioning using a high-precision displacement stage and displacement sensor, and the displacement sensor measures the relative displacement between the first and second movable support stages. Then, the main control computer calculates the real-time edge spacing based on the relative displacement, and then physically approximates the first and second gratings to the set edge spacing range; read the edge phase reference offset corresponding to the first and second gratings, and algebraically synthesize the edge spacing, edge phase reference offset, and mechanical positioning error to obtain the initial translation error estimate in the vector direction; when the sum of the absolute value of the initial translation error estimate and the comprehensive measurement uncertainty is less than the grating constant divided by twice the absolute value of the diffraction order, the integer periodic ambiguity term is uniquely determined to be zero; With the periodic ambiguity term determined to be zero, the absolute translation error of the first and second gratings in the vector direction is calculated directly using the principal phase difference between the first and second grating regions. A translation correction command is generated based on the absolute translation error value to drive the second grating to perform planar translation orientation compensation until the non-zero order interference fringes corresponding to the spliced ​​gratings are in a state of complete phase alignment. The bonding material at the bottom of the first and second gratings is cured by ultraviolet light irradiation to lock the in-plane two-dimensional coordinate positions of the first and second gratings.

[0046] See attached document Figure 3 To be continued Figure 5 In specific implementations, step S1 provided by the present invention may include the following steps: S11, an optical flat with a preset surface flatness of less than 30 nanometers is obtained as the parent substrate, and the parent substrate is divided into homologous substrates along a preset dividing line with a width of 10 micrometers to 20 micrometers using a semiconductor cutting device, and a first substrate and a second substrate with matching dimensions are output; then, a nanoimprinting adhesive is coated on the surface of the first substrate and the second substrate, and then a hard limiting process is used to perform nanoimprinting operation on the first substrate and the second substrate, that is, in the closing stage of the imprinting action, the closing gap between the optical template with the target grating feature and the substrate is limited by the solid pad structure arranged around the imprinting head of the imprinting device.

[0047] Simultaneously, by combining the substrate thickness parameters of the first and second substrates given by the homogeneous splitting process and the fixed gap parameters defined by the solid pad structure, the thickness of the imprinted residual layer on the surface of the first and second substrates is kept consistent. The first grating and the second grating are generated by the consistency of the imprinted residual layer thickness. The first grating is G1 and the second grating is G2. The total structural thickness deviation of the first grating and the second grating is controlled within the subwavelength range of 10 nanometers to 100 nanometers of optical interference.

[0048] Phase marks are set on the optical template. The imprinting equipment uses the preset splicing edges of the first and second substrates and the phase marks of the optical template as a joint positioning reference to establish a definite positional relationship between the periodic phase starting point of the first and second gratings and the corresponding splicing edges. After imprinting, the grating constant, scribe line direction, effective diffraction surface height, and edge phase reference offset of the first and second gratings are measured using an atomic force microscope or a high-resolution optical microscopy measurement device. The edge phase reference offset is defined as the distance from the splicing edge to the phase reference line. Then, the measurement results obtained from the previous optical measurement steps are stored in the main control computer. When the measurement uncertainty of the grating constant difference, scribe line direction difference, effective diffraction surface height difference, and edge phase reference offset is less than the system set threshold of 0.1 nanometers to 1 nanometer, the first and second gratings enter the subsequent splicing process.

[0049] S12 provides a reference module with an internal micro-hole array. The reference module includes a first movable support stage and a second movable support stage. The first movable support stage is used to output in-plane angular displacement, and the second movable support stage is used to output in-plane angular displacement and planar translation. Furthermore, the bottom of the first movable support stage and the second movable support stage are equipped with a multi-axis high-precision piezoelectric displacement stage that combines large-stroke mechanical positioning and micro-nano-level piezoelectric positioning. The upper surfaces of the first movable support stage and the second movable support stage are coplanarly assembled and together form a physically constrained reference support surface.

[0050] The first movable support stage and the second movable support stage are respectively equipped with independent micropore arrays and negative pressure gas paths. The prepared first grating and second grating are placed on the upper surfaces of the first movable support stage and the second movable support stage, respectively. Then, the corresponding negative pressure adsorption structure is activated, and negative pressure gas cavities are constructed between the bottom surfaces of the first grating and the second grating and the corresponding movable support surfaces by gas path pumping. When the pressure sensor installed inside the negative pressure adsorption structure detects that the air pressure in the negative pressure chamber has dropped and entered the first negative pressure range of -80kPa to -50kPa, the external atmospheric pressure applies a sufficient vertically downward uniformly distributed load through the top surfaces of the first and second gratings. Under the action of the uniformly distributed load, the first and second gratings undergo elastic micro-deformation and attach themselves to the corresponding movable bearing surfaces. At the same time, the clamping gap between the bottom surfaces of the first and second gratings and the corresponding movable bearing surfaces converges as the pressure inside the negative pressure chamber decreases.

[0051] Negative pressure adsorption is used to restrict the normal and out-of-plane rotational degrees of freedom of the first and second gratings relative to their respective movable bearing surfaces. The effective diffraction plane height difference between the first and second gratings is also affected by the substrate thickness, the thickness of the imprinted residual layer, and the grating structure height. Based on the effective diffraction plane height measurement results output by the measuring equipment, the main control computer confirms that the effective diffraction plane height difference is less than the height threshold of 10 to 50 nanometers set by the system. The numerical convergence state of the corresponding three-dimensional plane error vector satisfies the following mathematical relationship: ; In the formula, This is the real-time negative pressure scalar value generated by the air circuit pumping in the negative pressure air chamber; To define the target vacuum threshold of -85 kPa to -95 kPa for achieving a rigid mechanical fit inside the gas cavity; The height difference parameter between the effective diffraction surface of the second grating and the effective diffraction surface of the first grating along the normal direction; The rotational displacement error parameter of the second grating relative to the first grating about the grating vector direction; This is the rotational displacement error parameter of the second grating relative to the first grating around the grating scribe line direction.

[0052] Based on the same substrate segmentation preparation steps and negative pressure adsorption structure attachment steps, the normal degree of freedom of the grating substrate is restricted. Ultimately, the longitudinal translation of the first grating and the second grating relative to the corresponding movable bearing surface along the normal direction and the rotational displacement around the grating vector and scribe line direction are restricted. The degree of freedom that needs to be further adjusted between the first grating and the second grating is limited to in-plane translation and rotation around the normal direction.

[0053] A bridging and bonding groove is provided at the junction of the upper surfaces of the first movable support platform and the second movable support platform. The inner surface of the bridging and bonding groove is coated with an anti-adhesion and demolding coating. At the same time, the bridging and bonding groove is located outside the micropore array and is isolated from the negative pressure air path. Before the first grating and the second grating are placed, UV-curable bonding material is applied in the bridging and bonding groove.

[0054] The upper surface of the uncured adhesive material is lower than the upper surfaces of the first and second movable support platforms, and the first and second gratings do not come into contact with the uncured adhesive material during in-plane positioning. This avoids the liquid viscous resistance generated by the uncured adhesive material from interfering with the high-precision in-plane positioning action of the multi-axis high-precision piezoelectric displacement stage, thereby ensuring the nanometer-level execution accuracy of the planar translational orientation compensation action. At the same time, the reference module uses a material that can transmit the curing wavelength, or an ultraviolet curing light channel is set on the side of the bridging adhesive groove.

[0055] In specific implementations, step S2 provided by this invention may include the following steps: S21, set an optical right-angle guide at the edge of the reference module, and start the image acquisition component arranged above the reference module to acquire the relative position image of the edge of the first grating and the edge of the optical right-angle guide; perform denoising and edge operator convolution processing on the relative position image according to the existing Canny edge detection algorithm, extract the edge contour features in the relative position image and calculate the in-plane angle deviation; the calculation process combines the Hough transform line fitting algorithm to extract the spatial direction of the straight line of the first grating edge and the straight line of the optical right-angle guide respectively, and then directly obtain the scalar value of the in-plane angle deviation by analyzing the angle between the normal vectors of the two edge lines.

[0056] S22, the in-plane angle deviation is input into the motion controller. The motion controller drives the first movable support stage to rotate the first grating around the normal direction according to the in-plane angle deviation until the edge of the first grating is parallel to the edge of the optical right angle guide. After the edge is parallel, the motion controller locks the in-plane angle position of the first movable support stage and keeps the in-plane orientation of the first grating unchanged during subsequent detection and compensation.

[0057] The edge parallel state is defined as a machine vision resolution angle threshold that is updated in real time and whose in-plane angle deviation is less than a preset range of 0.5 arcseconds to 2.0 arcseconds. In addition, the value of the machine vision resolution angle threshold is determined by the arctangent of the single pixel size of the image acquisition component and the field of view width. The physical orientation calibration reference of the first grating in the two-dimensional plane is directly established. The main control computer reads the angle data between the measured first grating scribe line direction and the physical edge of the first grating, and performs coordinate system rotation compensation on the physical orientation calibration reference according to the angle data to ensure that the rotation axis direction of the bearing module is strictly parallel to the first grating scribe line direction.

[0058] S23, control the center of the field of view of the image acquisition component to translate to the junction area of ​​the first grating and the second grating, and then acquire the relative position image between the second grating and the first grating; call the existing edge extraction and line fitting algorithm to extract the line equation parameters of the edge of the second grating and the edge of the first grating contained in the relative position image, and output the physical distance information by calculating the normal pixel distance of the two line equations in the image coordinate system and combining the optical magnification, and output the rotation deviation information by calculating the relative slope difference of the two line equations, thereby extracting the spacing data and parallelism data in the relative position image.

[0059] S24, the motion controller drives the second movable support platform to move the second grating towards the first grating based on the spacing and parallelism data. During the movement, the second grating remains attached to the upper surface of the second movable support platform, and there is no relative sliding between the bottom surface of the second grating and the second movable support platform. A collision safety distance threshold of 10 to 50 micrometers corresponding to the closed-loop positioning accuracy is preset at the system's bottom layer, and the effective observation field of view of the image acquisition component is preset to an upper limit greater than the collision safety distance threshold. When the spacing data decreases to equal the collision safety distance threshold, the motion controller outputs a braking command, ensuring that the second grating and the first grating are physically suspended and isolated throughout the process. This allows the in-plane angular deviation and translation error between the second and first gratings to smoothly enter the measurement and capture range of the interferometric field of view. The error convergence state satisfies the following mathematical relationship: ; In the formula, This represents the in-plane angular deviation of the second grating relative to the first grating; To detect the wavelength of the light beam; This represents the total number of pixels captured in the image. For pixel physical size; This refers to the in-plane translation error; To effectively observe the physical width of the field of view.

[0060] The controlled physical deviation within the measurement capture range reflects that the initial installation deviation is within the critical range where the spatial frequency of the interference fringes does not exceed the resolution limit of the detector. The effective observation field of view physical width is calculated based on the total number of image acquisition pixels, pixel physical size, and optical magnification of the imaging relay component. The minimum spatial period of the interference fringes satisfies that the number of sampling pixels corresponding to each fringe period is not less than the minimum number of sampling pixels of 3 to 5 pixels set by the system. Furthermore, the splicing area of ​​the first grating and the second grating is completely located within the effective observation field of view, thereby establishing the interferometric detection conditions for single-wavelength interferometric detection.

[0061] See attached document Figure 6 In specific implementations, step S3 provided by the present invention may include the following steps: S31, activate the single-wavelength interferometer, such as a Zygo interferometer, and emit a single-wavelength detection beam toward the surface region where the first and second gratings are located; use a beam expanding and collimating optical lens group to expand and collimate the output signal of the light source so that the physical spot of the detection beam completely covers the splicing boundary region of the first and second gratings; combine the spatial grating equation with the numerical calculation of the detection wavelength in the range of 500 nm to 700 nm to realize the Littrow diffraction target incident angle of the non-zero order diffraction light returning along the original path. In this calculation, the sine value of the target incident angle is equal to the product of the diffraction order and the detection wavelength divided by twice the grating constant.

[0062] The rotation axis of the carrier module is parallel to the grating line direction data after the azimuth calibration is completed and stored. The detection beam changes the incident angle in the main section perpendicular to the grating line direction. According to the set target incident angle of 30 degrees to 60 degrees, the reference module is driven to rotate around the rotation axis, so that the non-zero order diffracted light generated by the first grating and the second grating enters the receiving aperture of the interferometer. The rotation actuator of the reference module is controlled by a direct drive rotary motor. The physical positioning accuracy of the target incident angle meets the geometric optical matching condition that the non-zero order diffracted light path strictly returns along the original optical axis.

[0063] S32, the beam splitter assembly arranged at the front end of the optical path receives the same non-zero diffraction order return light generated by the first grating and the second grating in the splicing area, and combines the non-zero diffraction order return light with the reference beam inside the interferometer; the wavefront of the non-zero diffraction order return light after beam combining carries the in-plane angle information and vector direction translation information of the first grating and the second grating respectively.

[0064] The imaging relay component maps the returned wavefronts corresponding to the first grating and the second grating to different sub-regions of the photodetector. The reference beam inside the measurement system is emitted by the same single-wavelength laser source component and formed by directional reflection through an internal reference mirror. The returned wavefront corresponding to the first grating interferes with the reference beam in the first sub-region, and the returned wavefront corresponding to the second grating interferes with the reference beam in the second sub-region, forming the first interferometric sub-image and the second interferometric sub-image located in the same detector coordinate system. The main control computer performs spatial coordinate registration on the first interferometric sub-image and the second interferometric sub-image, and extracts the spatial frequency, phase gradient, and phase intercept of the two sub-regions.

[0065] S33, the intensity distribution of the interference fringes on the target surface of the photodetector satisfies the cosine function modulation law, and the intensity distribution of the interference light corresponding to a single wavelength satisfies the following mathematical relationship: ; In the formula, Scalar for the real-time light intensity distribution of the interference fringe image in the two-dimensional spatial coordinate system of the photosensitive target surface; The horizontal pixel coordinate parameters of the photosensitive target surface of the photodetector in a two-dimensional Cartesian coordinate system; The vertical pixel coordinate parameters of the photosensitive target surface of the photodetector in a two-dimensional Cartesian coordinate system; The background bias light intensity parameter caused by backscattering from the optical path system; The fringe modulation contrast amplitude parameter is determined by the amplitude of the interferometric light. This is interference phase distribution data that includes physical displacement information.

[0066] Based on existing Fourier transform fringe processing algorithms, frequency domain filtering and inverse transform demodulation operations are performed on the light intensity distribution data output by the photodetector. The enclosed phase containing physical displacement information is separated through demodulation operations, and the spatial periodic distribution data and principal phase distribution data in the non-zero order interference fringe image are extracted. The values ​​of the extracted principal phase distribution data are enclosed in a mathematical truncation interval consisting of negative pi to positive pi.

[0067] See attached document Figure 7 and attached Figure 8 In specific implementations, step S4 provided by the present invention may include the following steps: S41, the main control computer receives the spatial periodic distribution data corresponding to the non-zero order interference fringe image; when there is an in-plane angular deviation between the first grating and the second grating, the returning wavefronts corresponding to the first grating and the second grating have different propagation directions relative to the common reference wavefront, causing the first and second interferometric sub-images to exhibit different fringe spatial frequencies, fringe directions, or phase gradients. A definite mapping relationship is established between the spatial frequency difference or phase gradient difference between the two sub-regions and the in-plane angular deviation; the main control computer establishes a mathematical calculation model between the fringe spatial period and the in-plane angular deviation based on the grating diffraction interference beat frequency principle, and calculates the in-plane angular deviation between the first grating and the second grating based on the spatial periodic distribution data. The corresponding calculation model satisfies the following formula: ; In the formula, This represents the in-plane angular deviation of the second grating relative to the first grating; To detect the wavelength of the light beam; The object-side equivalent fringe spacing value is calculated based on the spatial frequency difference or phase gradient difference between the first and second interferometric images. Furthermore, the object-side equivalent fringe spacing value is obtained by converting the extracted spatial periodic distribution data into physical scale values ​​using the optical magnification of the system imaging relay component.

[0068] The main control computer calculates and outputs the absolute value of the inward angle deviation under the current state based on the set wavelength value and the extracted object equivalent fringe spacing value.

[0069] S42, the main control computer generates an angle correction command based on the in-plane angle deviation and sends the angle correction command to the motion controller; the motion controller has a built-in digital-to-analog converter and voltage amplifier. The digital-to-analog converter receives the angle correction command in digital form and converts it into an analog voltage signal; the voltage amplifier amplifies the analog voltage signal to the working voltage range of 0 volts to 150 volts of the piezoelectric ceramic material; The voltage amplifier outputs an amplified control voltage, which drives a multi-axis high-precision piezoelectric displacement stage connected to the second grating to generate a microscopic mechanical angular displacement, thereby driving the second grating to perform in-plane angular displacement compensation. The physical rotation direction of the in-plane angular displacement compensation is opposite to the mathematical sign of the in-plane angular deviation, causing the second grating to continuously approach the absolute zero position with decreasing deflection angle while maintaining physical non-contact.

[0070] S43, during the process of the multi-axis high-precision piezoelectric displacement stage driving the second grating to rotate, the measurement system cyclically performs spatial periodic distribution data extraction and angular displacement compensation actions at a sampling frequency of 10 Hz to 100 Hz, and synchronously updates the in-plane angular deviation between the first grating and the second grating; the main control computer continuously monitors the physical distribution characteristics of the interference fringe image until the non-zero order interference fringes corresponding to the first grating and the second grating present a spatially equal width state. The spatially equal width state is defined as the fringe spatial frequency difference, fringe direction difference, and phase gradient difference between the first interferometric sub-image and the second interferometric sub-image being less than the corresponding system-set dimensionless threshold of 0.01 to 0.05 after normalization. The normalization process involves dividing the calculated fringe spatial frequency difference, fringe direction difference, and phase gradient difference by the system's preset non-zero full-scale reference constants, specifically the detector's maximum measurable spatial frequency, the set field-of-view angle reference range, and the maximum phase gradient corresponding to the Nyquist sampling limit. This avoids the risk of division by zero caused by local reference values ​​approaching zero, while eliminating their dimensional differences and converting them into relative error proportionality coefficients. Furthermore, the real-time updated absolute value of the in-plane angle deviation is less than the system's set limit angle threshold.

[0071] Meanwhile, the value of the limit angle threshold is determined by the closed-loop angular resolution of the multi-axis high-precision piezoelectric displacement stage and is specifically limited to the range of 0.1 microradians to 1.0 microradians during the system initialization phase; when the space equal width state is reached, the motion controller outputs a voltage holding command to determine that the in-plane angle alignment is completed, providing a single-phase environment for translation error calculation.

[0072] See attached document Figure 8 To be continued Figure 10 In specific implementations, step S5 provided by the present invention may include the following steps: S51, the main control computer receives the principal phase distribution data corresponding to the non-zero order interference fringe image under the same spatial width state; in order to address the integer periodic ambiguity problem caused by analog-digital operations in the interference phase measurement, that is, the periodic ambiguity problem of translation error, the main control computer controls the motion controller to drive the multi-axis high-precision piezoelectric displacement stage equipped with the second grating to perform initial mechanical positioning.

[0073] Capacitive displacement sensors are installed on the opposite sides of the first and second movable support platforms. These sensors measure the relative displacement between the two platforms in real time and feed the data back to the main control computer. The main control computer then calculates the real-time edge spacing between the first and second gratings based on the initial installation position of the gratings on the corresponding movable support surfaces. Next, the main control computer combines the micro-nano-level positioning function of the multi-axis high-precision piezoelectric displacement stage to form a closed-loop control, further approximating the real-time edge spacing from the anti-collision safety distance range to the set target physical spacing of 100 to 500 nanometers.

[0074] Subsequently, the main control computer reads the stored first grating edge phase reference offset and second grating edge phase reference offset; then, according to mutually perpendicular vector coordinate directions, it algebraically synthesizes the real-time edge spacing with the two edge phase reference offsets to obtain the initial translation error estimate of the first grating and the second grating in the vector direction, and the initial translation error estimate is equal to the algebraic sum of the real-time edge spacing, the first grating edge phase reference offset, and the second grating edge phase reference offset in the corresponding vector coordinate directions; Specifically, if a one-dimensional coordinate axis is defined along the grating vector direction, with the direction from the physical edge of the first grating to the physical edge of the second grating as the positive coordinate direction, then the algebraic expression for the initial translation error estimate of the vector direction satisfies: ; In the formula, This is the initial translation error estimate; This represents the absolute positive value of the real-time edge spacing obtained through equivalent conversion; and These are the edge phase reference offsets of the first grating and the second grating, respectively.

[0075] In addition, the offset is positive when the pre-calibrated phase reference line is located outside the physical edge of the grating; and negative when the phase reference line is located inside the physical edge.

[0076] Simultaneously, the main control computer evaluates the comprehensive measurement uncertainty, which is composed of positioning errors of the multi-axis high-precision piezoelectric displacement stage, measurement errors of the capacitive displacement sensor, calibration errors of the edge phase reference offset, and dimensional change errors caused by temperature variations. To ensure the absolute reliability of the error defuzzification calculation, the comprehensive measurement uncertainty is calculated using the worst-case synthesis principle of the upper limit of the absolute values ​​of each component error. The specific expression satisfies: ; In the formula, To comprehensively measure uncertainty; The maximum positioning error limit for closed-loop control of a multi-axis high-precision piezoelectric displacement stage; This represents the maximum measurement error limit for the capacitive displacement sensor. This is the maximum calibration error limit for the edge phase reference offset in the early stage; This represents the extreme value of the error caused by changes in the mechanical structure dimensions due to variations in the system's ambient temperature.

[0077] When the sum of the absolute value of the initial translation error estimate and the overall measurement uncertainty is less than the grating constant divided by twice the absolute value of the diffraction order, the integer periodic ambiguity term corresponding to the interference phase is uniquely determined to be zero.

[0078] S52, the main control computer establishes a unified physical reference position based on the calibration mapping relationship between the pixel coordinate system of the photodetector and the physical plane of the grating; then, based on the unique determination condition that the periodic ambiguity term is zero, the main control computer performs phase demodulation on the first interferometric sub-image and the second interferometric sub-image respectively, and subtracts the system baseline phase, common tilt term and fixed optical path aberration term from the phase distribution of the two sub-regions.

[0079] The principal phase difference between the first and second grating regions is calculated at a unified physical reference position. When the integer periodic ambiguity term is zero, a unique correspondence is established between the principal phase difference and the absolute translation error in the vector direction. The main control computer uses the principal phase difference to calculate the absolute translation error value between the first and second gratings in the vector direction. The system baseline phase is pre-measured using a standard complete grating or a stitched grating at the calibration zero position and stored in the main control computer. The corresponding translation error calculation formula satisfies the following formula: ; In the formula, This represents the absolute translation error between the first and second gratings in the vector direction. The grating constant is the number of the spliced ​​gratings; The diffraction order corresponds to the non-zero order interference fringes; To obtain the principal phase difference by extracting phase values ​​from the first and second interferometric sub-images at a unified physical reference position and subtracting the system baseline phase, common tilt term, and fixed optical path aberration term.

[0080] The main control computer outputs the absolute translation error value, thereby achieving the purpose of eliminating periodic ambiguity and obtaining the true translation error through high-precision mechanical positioning dimensionality reduction.

[0081] S53, the main control computer generates a translation correction command based on the absolute translation error value and sends the translation correction command to the motion controller; the motion controller converts the translation correction command into a drive voltage and outputs the drive voltage to drive the multi-axis high-precision piezoelectric displacement stage, which in turn drives the second movable support stage to drive the second grating to perform planar translation orientation compensation.

[0082] After each translation compensation is completed, the main control computer re-extracts the spatial periodic distribution data and principal phase distribution data of the first and second interferometric images, and simultaneously calculates the in-plane angular deviation and absolute translation error. When the in-plane angular deviation exceeds the limit angle threshold, the motion controller re-executes the in-plane angular displacement compensation; when the absolute translation error exceeds the positioning accuracy threshold of 1 nanometer to 5 nanometers, the motion controller continues to execute the planar translation compensation.

[0083] When the in-plane angle deviation is less than the limit angle threshold and the absolute translation error is less than the system's preset positioning accuracy threshold of 1 to 5 nanometers, it is determined that the non-zero order interference fringes have reached the state of equal spatial width and complete phase alignment. After reaching the state of equal spatial width and complete phase alignment, the motion controller maintains the driving position of the first movable support platform and the second movable support platform, and maintains the pressure stability in the two negative pressure air chambers. Next, the uncured adhesive material is squeezed upward through a microfluidic pneumatic channel located at the bottom of the bonding groove. At the same time, the upward pressure difference generated by the squeezing air in the microfluidic pneumatic channel is controlled to be strictly less than the absolute value of the downward adsorption pressure difference generated by the vacuum adsorption pressure in the negative pressure chamber. This ensures that the top surface of the uncured adhesive material is in full contact with the bottom surfaces of the first and second gratings, thereby preventing the upward mechanical thrust generated by the contact of the adhesive material from destroying the established phase alignment state. Subsequently, the curing equipment is started, and ultraviolet light is irradiated onto the adhesive material through the light-transmitting area of ​​the reference module or the curing light channel on the side of the bonding groove.

[0084] After the bonding material reaches 85% to 95% of the set curing degree, the negative pressure adsorption state of the first and second gratings and the driving holding state of the movable support stage are released in sequence. After curing, the in-plane angle deviation and absolute translation error are detected again. When the absolute value of the detected in-plane angle deviation is less than the limit angle threshold set by the system and the value of the detected absolute translation error is less than the set translation tolerance limit, the in-plane two-dimensional coordinate position locking of the first and second gratings is completed. At the same time, the cured bonding material forms a rigid connecting bridge at the bottom of the first and second gratings. Then, the negative pressure adsorption state of the first and second movable support stages is released, and the splicing grating master plate, which is physically integrated by the rigid connecting bridge, is demolded from the cross bonding groove. The grating is then mass-produced and replicated using nanoimprint replication technology to achieve efficient preparation of large-area splicing gratings.

[0085] See attached document Figure 11 To aid in understanding the technical solution of this invention, the following is an application embodiment of the method for defuzzifying the translation error of spliced ​​gratings.

[0086] The application example receives a first grating and a second grating, which are prepared by dividing a substrate from the same source and have the characteristics of the target grating, and pre-sets a high-precision single-wavelength interferometer and a machine vision component as a hardware reference for error detection and state observation.

[0087] The reference module acquires the first and second gratings and places them on the movable support surface inside the reference module. Then, the internal micro-pore array negative pressure adsorption structure is activated, and the set negative pressure air path is used to draw pressure, so that the first and second gratings are tightly attached to the corresponding movable support surface by elastic micro-deformation. This process restricts the longitudinal translation of the first and second gratings along the normal direction, as well as the rotational displacement around the grating vector and scribe line direction, thereby reducing the three-dimensional spatial adjustment requirements of the grating to pure in-plane translation and rotation around the normal direction.

[0088] The observation module activates the image acquisition component to obtain the relative position image of the edge of the first grating and the preset optical right-angle guide, and extracts the edge contour features to calculate the in-plane angle deviation. Then, the linkage adjustment module drives the first grating to rotate to a parallel state with the guide, completes the absolute calibration of the physical orientation and locks it. Subsequently, the observation module shifts the field of view to the junction area of ​​the first and second gratings to obtain the relative position image and extracts the spacing data and parallelism data. Based on the spacing data and parallelism data, the adjustment module drives the second grating to perform non-contact mechanical approach to the first grating, and triggers braking when the preset anti-collision safety distance threshold is reached, so that the first and second gratings can safely enter the effective measurement and acquisition range of the interferometric field of view while maintaining physical suspension and isolation.

[0089] The detection module emits a single-wavelength detection beam toward the region containing the first and second gratings. Simultaneously, the carrier module drives the reference module to rotate according to the target incident angle, causing the non-zero order diffracted light of a specific order to return along its original path and combine with the reference beam inside the interferometer, thereby generating a non-zero order interference fringe image containing displacement information. The calculation module extracts the spatial periodic distribution data from the interference fringe image and calculates the in-plane angle deviation scalar based on the mathematical mapping model of the fringe spatial period and the in-plane angle deviation, while generating an angle correction command. The adjustment module receives the angle correction command and drives the second grating to perform in-plane angle displacement compensation. The system performs a closed-loop operation of detection, calculation, and compensation at high frequency until the interference fringes corresponding to the stitched area show a spatially equal width state. At this point, it is determined that the in-plane angle deviation has been completely eliminated.

[0090] The solution module receives the principal phase distribution data under the condition of equal width in the fringe space, and performs algebraic synthesis by combining the real-time physical edge spacing measured by the displacement sensing unit in the observation module with the pre-calibrated edge phase reference offset to obtain the initial translation error estimate of the vector direction. Under the condition that the comprehensive measurement error meets the uniquely determined reference constraint set by the system, the solution module determines the integer periodic fuzzy term of the interference phase to be zero, and directly uses the principal phase difference between the first grating and the second grating regions to accurately calculate the absolute translation error value.

[0091] The adjustment module drives the second grating to perform planar translational orientation fine-tuning based on the absolute translation error until the interference fringes of the first and second gratings are in a state of complete phase alignment. Finally, the microfluidic pneumatic channel squeezes the uncured adhesive material in the bonding groove upward, so that the uncured adhesive material contacts the bottom surfaces of the first and second gratings. Then, the curing unit emits an ultraviolet beam to complete the bonding and curing, realizing the permanent locking of the two-dimensional coordinates within the splicing grating master plate.

[0092] While specific embodiments of the present invention have been described above, those skilled in the art should understand that these specific embodiments are merely illustrative. Those skilled in the art can omit, substitute, and modify the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, combining the above method steps to perform substantially the same function and achieve substantially the same result using substantially the same method falls within the scope of the present invention.

Claims

1. A method for resolving translation errors in spliced ​​gratings, characterized in that, include: The first grating and the second grating are attached to the corresponding movable bearing surfaces of the first movable bearing platform and the second movable bearing platform by air pressure. The first movable bearing platform outputs the rotation of the first grating around the normal direction, and the second movable bearing platform outputs the in-plane translation and rotation around the normal direction of the second grating. Drive the second grating to move in the direction of the first grating, so that the in-plane angle deviation and translation error are brought into the measurement capture range of the interferometric field of view; Generate a non-zero order interference fringe image corresponding to the splicing region of the first grating and the second grating. The non-zero order interference fringe image includes a first interferometric sub-image and a second interferometric sub-image. Extract the spatial periodic distribution data and principal phase distribution data from the non-zero order interference fringe image. Based on the spatial periodic distribution data, the in-plane angular deviation is calculated, and the second grating is driven to perform in-plane angular displacement compensation until the non-zero order interference fringes present a state of equal spatial width. The real-time edge spacing and the offset of the two edge phase references are algebraically synthesized to obtain the initial translation error estimate in the vector direction. Based on the real-time edge spacing, the second grating is driven to approach the first grating, so that the sum of the absolute value of the initial translation error estimate and the preset comprehensive measurement uncertainty is less than the grating constant of the spliced ​​grating divided by twice the absolute value of the diffraction order, so that the integer periodic ambiguity term corresponding to the interference phase is uniquely determined to be zero. The absolute translation error value in the vector direction is calculated, and the second grating is driven to perform planar translation orientation compensation to lock the in-plane two-dimensional coordinate positions of the first grating and the second grating.

2. The method for resolving translation errors of spliced ​​gratings according to claim 1, characterized in that, The step of driving the second grating to move towards the first grating so that the in-plane angular deviation and translation error enter the measurement capture range of the interferometric field of view includes: Obtain the relative position image of the edge of the first grating and the edge of the optical right-angle guide, extract the edge contour features in the relative position image and calculate the in-plane angle deviation; The first grating is driven to perform planar rotation based on the in-plane angle deviation until the edge of the first grating is parallel to the edge of the optical right angle guide. Acquire the relative position image between the second grating and the first grating, and extract the spacing data and parallelism data from the relative position image; The second grating is driven to move towards the direction of the first grating based on the spacing and parallelism data.

3. The method for resolving translation errors of spliced ​​gratings according to claim 2, characterized in that, The step of driving the second grating to move towards the first grating based on the spacing data and parallelism data includes: The system's underlying layer presets a collision safety distance threshold in the range of 10 to 50 micrometers, corresponding to the closed-loop positioning accuracy. When the spacing data decreases to the level of the collision avoidance safety distance threshold, the motion controller outputs a braking command to ensure that the second grating and the first grating are physically suspended and isolated throughout the process.

4. The method for resolving translation errors of spliced ​​gratings according to claim 1, characterized in that, The step of generating a non-zero order interference fringe image corresponding to the stitching region of the first grating and the second grating includes: The reference module carrying the first and second gratings is driven to perform a rotation around the axis according to the set target incident angle. Receive the return light of the same non-zero diffraction order generated by the first grating and the second grating in the splicing area; The beam splitter assembly combines the return light of the same non-zero diffraction order with the reference beam formed by the reflection of the reference mirror. The detection module processes the beam after the return light of the same non-zero diffraction order is combined with the reference beam to generate a non-zero order interference fringe image.

5. The method for resolving translation errors of spliced ​​gratings according to claim 1, characterized in that, The steps for calculating the in-plane angular deviation based on spatial periodic distribution data include: Based on the spatial frequency difference or phase gradient difference between the first and second interferometric images, and combined with the optical magnification of the system imaging relay component, physical scale conversion is performed to obtain the object-side equivalent fringe spacing value. The main control computer calculates and outputs the absolute value of the inward angle deviation under the current state based on the set wavelength value and the extracted object equivalent fringe spacing value.

6. The method for resolving translation errors of spliced ​​gratings according to claim 1, characterized in that, The preset comprehensive measurement uncertainty is composed of the positioning error of the multi-axis high-precision piezoelectric displacement stage, the measurement error of the capacitive displacement sensor, the calibration error of the edge phase reference offset, and the dimensional change error caused by temperature changes.

7. The method for resolving translation errors of spliced ​​gratings according to claim 1, characterized in that, The steps for calculating the absolute translation error value of the vector direction include: Based on the unique determining condition that the periodic ambiguity term is zero, the main control computer performs phase demodulation on the first and second interferometric sub-images respectively; The absolute translation error of the first grating and the second grating in the vector direction is calculated directly using the principal phase difference between the first grating region corresponding to the first interferometer image and the second grating region corresponding to the second interferometer image.

8. The method for resolving translation errors of spliced ​​gratings according to claim 1, characterized in that, The step of attaching the first grating and the second grating to the corresponding movable bearing surfaces of the first movable bearing platform and the second movable bearing platform by air pressure includes: The microporous array negative pressure adsorption structure, which is respectively set inside the first movable support platform and the second movable support platform, is activated. Negative pressure air chambers are constructed between the bottom surfaces of the first and second gratings and their corresponding movable bearing surfaces by air pressure extraction. When the air pressure in the negative pressure chamber drops and enters the first negative pressure range of -80kPa to -50kPa, the external atmospheric pressure applies a sufficient vertically downward uniformly distributed load through the top surfaces of the first and second gratings. Under uniformly distributed load, the first grating and the second grating undergo elastic micro-deformation and attach the first grating and the second grating to the corresponding movable bearing surface.

9. The method for resolving translation errors of spliced ​​gratings according to claim 8, characterized in that, The step of locking the in-plane two-dimensional coordinate positions of the first grating and the second grating includes: The uncured adhesive material is extruded upward through microfluidic air pressure channels arranged at the bottom of the bridging adhesive groove of the corresponding movable bearing surface; The upward pressure difference generated by the extrusion pressure in the microfluidic gas channel is strictly less than the absolute value of the downward adsorption pressure difference generated by the vacuum adsorption pressure in the negative pressure chamber, so that the top surface of the uncured adhesive material is in full contact with the bottom surfaces of the first grating and the second grating. The bonding material at the bottom of the first and second gratings is cured by ultraviolet light irradiation, thereby locking the in-plane two-dimensional coordinate positions of the first and second gratings.

10. The method for resolving translation errors of spliced ​​gratings according to claim 1, characterized in that, The steps until the non-zero order interference fringes exhibit a spatially equal width include: The spatial equal width state is defined as follows: after normalization processing based on the system preset non-zero constant, the fringe spatial frequency difference, fringe direction difference, and phase gradient difference between the first interferometric sub-image and the second interferometric sub-image are all less than the corresponding system-set dimensionless threshold of 0.01 to 0.05, and the absolute value of the in-plane angle deviation updated in real time is less than the system-set limit angle threshold. When the space is equal in width, the motion controller outputs a voltage holding command to determine that the in-plane angle alignment is complete.