Active alignment and mounting method and apparatus for multi-channel optical assemblies

Abstract

The application discloses a kind of active alignment mounting method and device of multi-channel optical assembly, method includes: S1, between the mounting base surface of base body and multiple optical filters, respectively apply uncured optical glue layer;S2, obtain the first light spot center coordinates of each channel on near-field plane, second light spot center coordinates on far-field plane;S3, according to first, second light spot center coordinates, the angle compensation amount of each optical filter is calculated and angle adjustment is carried out, to change the thickness distribution of optical glue layer at corresponding position;S4, repeat execution S2 to S3, until the parallelism of outgoing light satisfies preset condition, then optical glue layer is cured.Through double-plane detection, the real direction angle of each channel outgoing light is obtained, the wedge-shaped cross section formed by uncured optical glue layer is used as active compensation element, and the outgoing direction of each channel is converged to parallel through multiple rounds of adjustment iteration, and finally the parallelism is determined by iterative convergence accuracy, to avoid the dependence on initial machining precision of base body.

Description

Technical Field

[0001] This invention relates to the field of optical communication technology, and in particular to an active alignment and mounting method and apparatus for multi-channel optical components. Background Technology

[0002] Multi-channel optical components (Z-Block) have wide applications in optical communication, optical sensing and other fields. Such components typically include a substrate and multiple filters that are sequentially mounted on the substrate along the optical path. Each filter corresponds to an independent optical channel. After the entire component is assembled, the outgoing light rays from each channel are required to maintain high parallelism to ensure the subsequent optical path coupling efficiency.

[0003] Currently, the industry typically uses a mounting process where filters are directly bonded to the mounting surface of a substrate using optical adhesive layers. This relies on the machining precision of the substrate mounting surface and the uniformity of the adhesive layer thickness to ensure the consistency of the filter orientation. However, surface shape errors exist in the substrate mounting surface during machining, and the thickness and wedge angle deviations of each filter also accumulate in the optical path, ultimately resulting in directional dispersion of the emitted light rays from each channel and low parallelism accuracy. Furthermore, the optical adhesive layer shrinks during curing, and the amount of shrinkage is affected by variations in the adhesive material and application rate. This causes a certain angular drift in the orientation of the cured filters compared to their initial adjusted state, further affecting the final parallelism of the emitted light rays from each channel. Summary of the Invention

[0004] The purpose of this invention is to provide an active alignment and mounting method and apparatus for multi-channel optical components. Under conditions where the incident light direction remains fixed and the initial positions of each filter are correctly placed, requiring only angle adjustment, the true direction angle of the emitted beam from each channel can be accurately obtained through dual-plane detection. Based on this, the angle compensation amount of each filter can be stably calculated. Before the optical adhesive layer cures, the filters are driven to tilt to form a reproducible wedge-shaped adhesive layer compensation, thereby achieving highly consistent parallelism of the multi-channel emitted light. The numerous technical effects of the preferred solutions among the many technical solutions provided by this invention are detailed below.

[0005] To achieve the above objectives, the present invention provides the following technical solution: This invention provides an active alignment and mounting method for a multi-channel optical component. The multi-channel optical component includes a substrate and multiple filters. The active alignment and mounting method includes the following steps: S1. Apply uncured optical adhesive layers between the mounting surface of the substrate and the multiple filters respectively; S2. Obtain the first spot center coordinates of each channel of the multi-channel optical component on the near-field plane, and obtain the second spot center coordinates of each channel on the far-field plane; wherein, the near-field plane and the far-field plane are two planes with different propagation distances arranged at intervals along the optical axis. S3. Based on the center coordinates of the first and second light spots of each channel, calculate the angle compensation amount for each filter, and drive the corresponding filter to adjust its angle according to the angle compensation amount, so as to change the thickness distribution of the optical adhesive layer at the corresponding position; wherein, the angle vector of each channel is calculated according to the following formula: theta x ,i=arctan((x far ,ix near ,i) / DeltaL); theta y i=arctan((y far ,iy near ,i) / DeltaL); Then, the angle compensation amount of each filter is determined according to the angle vector of each channel, wherein theta x ,i and theta y ,i represent the angular components of the i-th channel in the horizontal and vertical directions, respectively, and x represents the angular components of the i-th channel in the horizontal and vertical directions. near ,i and y near Let i be the coordinates of the center of the first spot of the i-th channel on the near-field plane, and x be the coordinates of the center of the first spot of the i-th channel. far ,i and y far ,i is the center coordinate of the second spot of the i-th channel on the far-field plane, and DeltaL is the axial distance between the near-field plane and the far-field plane; S4. Repeat steps S2 to S3 until the parallelism of the emitted light from the multi-channel optical component meets the preset conditions, then cure the optical adhesive layer.

[0006] Preferably, the axial distance between the near-field plane and the emission end face of the multi-channel optical component is 20mm, and the axial distance between the far-field plane and the emission end face is 80mm. Preferably, in step S3, the minimum single angular step size of the actuator used to drive the filter to adjust the angle is less than or equal to 0.001°.

[0007] Preferably, the outgoing light parallelism is the maximum pairwise angle between the outgoing light directions of each channel of the multi-channel optical component, and the objective function for calculating the parallelism is: J par =max {i<j}||Theta i -Theta j ||2; Among them, J par Theta is a parallelism index. i and Theta j Let i and j be the outgoing light direction vectors of the i-th and j-th channels, respectively, where i and j are channel indices and i < j.

[0008] Preferably, in step S3, during the angle adjustment process, each of the filters changes the local thickness of the uncured optical adhesive layer below it, so that the optical adhesive layer forms a wedge-shaped cross section that changes linearly along the X and / or Y directions before curing, so as to act as an active optical compensation element to offset the substrate surface error of the multi-channel optical assembly and / or the assembly and adjustment error of each filter.

[0009] Preferably, in step S3, determining the angle compensation amount of each filter based on the angle vector of each channel specifically includes: pre-calibrating or updating online the sensitivity matrix between the change in the angle of each axis of each filter at the current working point and the change in the error vector of the emitted beam; using the sensitivity matrix as the system matrix, and employing the damped least squares method or the weighted least squares method, solving for the angle increment vector that minimizes the directional angle error of each channel.

[0010] Preferably, the angle increment vector obtained by solving is subjected to quantization, rounding, and amplitude limiting processing before being output to the actuator. The processing formula is as follows: DeltaAlpha cmd =clip(round(DeltaAlpha / s) s,[-Alpha max Alpha max ]); Where s is the preset minimum angular step size, and Alpha max This is the preset maximum safe adjustment range for a single axis.

[0011] Preferably, the plurality of filters are arranged sequentially along the optical path direction; in step S3, when driving each of the filters to adjust its angle, the adjustment is performed in reverse order of the optical path, wherein the filter located at the end of the optical path is adjusted first, and the filter located at the beginning of the optical path is adjusted later.

[0012] Preferably, the preset conditions in step S4 include: if the parallelism index J of the emitted light of the multi-channel optical component is... par Less than or equal to the preset parallelism threshold J spec The mean square error J of the center position of the light spot on the near-field plane of each of the aforementioned channels ctr Less than or equal to the preset center error threshold Cspec After a preset number of consecutive attitude adjustments, if the angle change corresponding to a single attitude adjustment is less than a preset change threshold, then the uncured optical adhesive layer is cured. After curing, the coordinates of the emitted light spots of each channel on the near-field plane and the far-field plane are obtained again to verify the parallelism of the cured emitted light and the near-field center position.

[0013] An active alignment and mounting apparatus for implementing the above method includes: A three-axis angle adjustment mechanism is used to clamp multiple filters and drive the filters to adjust their angles. A dual-plane imaging component is used to acquire the coordinates of the emitted light spot of a multi-channel optical component on at least two planes with different propagation distances. A collimated light source is fixedly mounted at the incident end of the multi-channel optical assembly; A curing assembly for performing a curing process on an uncured optical adhesive layer applied between the substrate and the filter; The controller is communicatively connected to the three-axis angle adjustment mechanism, the dual-plane imaging component, and the solidification component.

[0014] Implementing one of the above-described technical solutions of this invention has the following advantages or beneficial effects: This application is implemented under the condition that the collimator incident light direction remains unchanged and the initial positions of each filter are correctly placed, requiring only angle adjustment; by obtaining the output light spot coordinates of each channel through the dual plane in step S2, the center position information of eight light spots in four channels can be obtained simultaneously, thereby inverting the required three-axis angle compensation amount for each filter; due to the fluidity of the liquid, the optical adhesive layer forms a wedge-shaped adhesive layer after angle adjustment in step S3, and this wedge-shaped adhesive layer is used as an active optical compensation element, so that it maintains the compensation shape after curing. Since the output light direction is continuously monitored and corrected in real time by repeated execution of S2 to S3 before curing, the compensation amount can gradually approach the optimal value with iteration, and the final parallelism is determined by the convergence accuracy after iteration, rather than by the initial processing accuracy of the substrate.

[0015] Furthermore, since the angular drift caused by curing shrinkage has been eliminated in advance during the iteration process after the parallelism of the optical adhesive layer meets the preset conditions, the parallelism after curing is no longer affected by the dispersion of the adhesive material, so that the parallelism of the multi-channel emitted light converges from the initial discrete distribution to a controllable accuracy range. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is a flowchart illustrating the active alignment and mounting method for multi-channel optical components provided by the present invention. Figure 2 This is a three-dimensional structural schematic diagram of the dual-plane imaging component of the active alignment and mounting device provided by the present invention; Figure 3 This is an exploded view of the dual-plane imaging component of the active alignment and mounting device provided by the present invention; Figure 4 This is a partial three-dimensional structural diagram of the dual-plane imaging component of the active alignment and mounting device provided by the present invention; Figure 5 This is a cross-sectional view of the dual-plane imaging component of the active alignment and mounting device provided by the present invention; Figure 6 This is a schematic diagram of the optical path structure of the multi-channel optical component of the present invention during active alignment and mounting; Figure 7 This is a schematic diagram of the center coordinates of 8 light spots acquired by the near-field and far-field dual-plane imaging component of the present invention; Figure 8 This is a schematic diagram comparing the parallelism distribution of the outgoing light before and after compensation in each channel of the multi-channel optical component of the present invention. Figure 9 This is a graph showing the effect of the minimum angular step size of the actuator of the present invention on the parallelism compensation effect.

[0017] In the figure: 1. Dual-plane imaging assembly; 11. Beam splitter; 12. Refraction assembly; 121. First refractor; 122. Second refractor; 123. Third refractor; 13. Imaging plane; 14. Housing; 15. Mounting base; 16. Sliding base; 2. Collimating light source; 3. First beam; 4. Second beam. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the present invention clearer, various exemplary embodiments described below will be referenced to the accompanying drawings, which form part of the exemplary embodiments, illustrating various exemplary embodiments that may be used to implement the present invention. Unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. It should be understood that they are merely examples of processes, methods, and apparatuses consistent with some aspects of the present invention disclosed as detailed in the appended claims, and other embodiments may be used, or structural and functional modifications may be made to the embodiments listed herein without departing from the scope and spirit of the present invention.

[0019] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," etc., indicate the orientation or positional relationship based on the accompanying drawings, and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the referred element must have a specific orientation, or be constructed and operated in a specific orientation. The terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. The term "multiple" means two or more. The terms "connected" and "linked" should be interpreted broadly, for example, they can be fixed connections, detachable connections, integral connections, mechanical connections, electrical connections, communication connections, direct connections, indirect connections through an intermediate medium, and can be the internal connection of two elements or the interaction relationship between two elements. The term "and / or" includes any and all combinations of one or more of the related listed items. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0020] To illustrate the technical solution described in this invention, specific embodiments are described below, showing only the parts related to the embodiments of this invention.

[0021] like Figure 1 As shown, this invention provides an active alignment and mounting method for a multi-channel optical component. The multi-channel optical component includes a substrate and multiple filters. The active alignment and mounting method includes the following steps: S1. Apply uncured optical adhesive layers between the mounting surface of the substrate and multiple filters respectively; the optical adhesive layers can be UV-cured epoxy resin adhesive or thermosetting optical adhesive. The liquid flow of the optical adhesive layer provides space for thickness adjustment for the angle adjustment of the filters.

[0022] S2. Obtain the center coordinates of the first spot of each channel of the multi-channel optical component on the near-field plane, and obtain the center coordinates of the second spot of each channel on the far-field plane; wherein, the near-field plane and the far-field plane are two planes with different propagation distances arranged at intervals along the optical axis. S3. Based on the center coordinates of the first and second light spots of each channel, calculate the angle compensation amount for each filter, and drive the corresponding filter to adjust its angle according to the angle compensation amount, thereby changing the thickness distribution of the optical adhesive layer at the corresponding position; for example, pressing down on one side of the filter thins the adhesive layer, while lifting up on the other side thickens the adhesive layer, forming a wedge-shaped adhesive layer structure; wherein, the angle vector of each channel is calculated according to the following formula: theta x ,i=arctan((x far ,ix near ,i) / DeltaL); theta y i=arctan((y far ,iy near ,i) / DeltaL); Then, the angle compensation amount of each filter is determined according to the angle vector of each channel, wherein theta x ,i and theta y ,i represent the angular components of the i-th channel in the horizontal and vertical directions, respectively, and x represents the angular components of the i-th channel in the horizontal and vertical directions. near ,i and y near Let i be the coordinates of the center of the first spot of the i-th channel on the near-field plane, and x be the coordinates of the center of the first spot of the i-th channel. far ,i and y far ,i is the center coordinate of the second spot of the i-th channel on the far-field plane, and DeltaL is the axial distance between the near-field plane and the far-field plane; S4. Repeat steps S2 to S3 to gradually bring the outgoing light from each channel closer to the optimal parallelism state until the parallelism of the outgoing light from the multi-channel optical components meets the preset conditions. Then, cure the optical adhesive layer to fix each filter. At the same time, the angular drift caused by curing shrinkage is added to the compensation solution model and deducted, so that the parallelism after final curing and locking is not affected by the material dispersion of the adhesive itself.

[0023] This application is implemented under the condition that the collimator incident light direction remains unchanged and the initial positions of each filter are correctly placed, requiring only angle adjustment. Using the coordinates of the emitted light spots from each channel obtained through the dual planes in step S2, the center position information of eight light spots across four channels can be obtained simultaneously, thereby retrieving the required triaxial angle compensation for each filter. Due to its liquid fluidity, the optical adhesive layer forms a wedge-shaped adhesive layer after angle adjustment in step S3. This wedge-shaped adhesive layer serves as an active optical compensation element, maintaining its compensation shape after curing. Because the emitted light direction is continuously monitored and corrected in real time through repeated execution of S2 to S3 before curing, the compensation amount gradually approaches the optimal value with iteration. The final parallelism is determined by the convergence accuracy after iteration, rather than by the initial processing accuracy of the substrate. Simultaneously, because the angular drift caused by curing shrinkage is eliminated before the parallelism meets the preset conditions during the curing process, the parallelism after curing is no longer affected by the dispersion of the adhesive material, allowing the parallelism of the multi-channel emitted light to converge from the initial discrete distribution to a controllable accuracy range.

[0024] Specifically, the near-field plane and the far-field plane are arranged alternately along the propagation direction of the emitted light, with an axial distance of DeltaL between them. For any channel in the multi-channel optical assembly, the coordinates (x, y, y) of its first spot center on the near-field plane are obtained. near ,i,y near (i) and the coordinates of the center of the second spot on the far-field plane (x,i) far ,i,y far To obtain two observation points at different distances from the outgoing rays of each channel, i), the lateral position of the outgoing ray on a single plane includes both translation and tilt. If the beam is translated as a whole but its direction remains unchanged, the spot coordinates on the two planes will have the same displacement, and the coordinate difference will be zero. If the beam has a tilt angle, the spot coordinates on the far-field plane will shift relative to the spot coordinates on the near-field plane in a manner proportional to the tilt angle. Thus, the coordinate difference formed by the same beam on two planes separated by DeltaL is only caused by the tilt angle of the beam, and the translation component is naturally eliminated when subtracted. Using the arctangent of the ratio of the above coordinate difference to DeltaL, i.e., theta... x ,i=arctan((x far ,i–x near ,i) / DeltaL) and theta y i=arctan((y far ,i–y near By calculating ,i) / DeltaL), the angle vectors theta of the channel in the horizontal and vertical directions can be obtained respectively. x i、theta yThen, based on the difference in angle vectors of all channels, the required angle compensation for each filter is calculated.

[0025] It should be noted that in existing filter mounting processes, the mounting surface of the substrate has surface shape errors, and the thickness and wedge angle of each filter inevitably have manufacturing deviations. This makes it difficult for the emission directions of multiple channels to be parallel. If the position of the light spot on a single plane is used for judgment, it is impossible to distinguish between the translation and tilt of the emitted light, and the accuracy of the obtained angle deviation is low. This application introduces axially spaced near-field and far-field planes, using DeltaL as the longitudinal scale and the difference between the coordinates of the centers of the two sets of light spots as the lateral input, to convert the position measurement into quantitative feedback of the directional angle, so that the actual emission direction of the emitted light of each channel can be accurately obtained. The angle compensation amount calculated in this way can drive the corresponding filter to tilt in a targeted manner to change the thickness distribution of the optical adhesive layer. Before the adhesive layer cures, a wedge-shaped cross section is formed for active compensation. This method makes the emission directions of the emitted light of each channel gradually converge to parallelism. The convergence accuracy is determined by the accuracy of the directional angle detection, rather than by the initial processing accuracy of the substrate and the filter assembly and adjustment error, so as to achieve high parallelism of the emitted optical fibers of multiple channels.

[0026] Furthermore, after step S4 determines that the parallelism meets the standard, based on the measured refractive index of the current optical adhesive layer and the thickness of the adhesive layer beneath each filter, the expected angular drift is read from the calibration data, inverted, and superimposed into the angular increment vector. Before complete curing, each filter undergoes a round of bias adjustment, allowing the curing shrinkage to naturally pull the filter back to the target orientation. In another embodiment, instead of relying on the refractive index calibration model, a short-term, low-dose pre-curing is performed before complete curing. The angular drift of each channel direction caused by the pre-curing is measured, inverted, and used to drive the filter for compensation adjustment before complete curing. Both methods ensure that the angular drift caused by the curing shrinkage of the optical adhesive layer is quantitatively predicted or measured and pre-compensated before curing, so that the parallelism of the emitted light after final curing is not affected by the shrinkage and dispersion of the adhesive material.

[0027] As an optional implementation, the axial distance between the near-field plane and the exit face of the multi-channel optical component is 20 mm, and the axial distance between the far-field plane and the exit face is 80 mm. The minimum single angular step size of the actuator used to drive the filter for angle adjustment is less than or equal to 0.001°. Specifically, DeltaL = 60 mm. When there is a small deviation in the actual exit direction of a certain channel in the multi-channel optical component, the coordinate offset caused by this deviation in the far-field plane is greater than its offset in the near-field plane. The difference between the two is amplified by DeltaL, so that even a very small angular deviation can be presented as a significant change in the far-field spot coordinates. At the same time, the minimum single angular step size of the actuator driving the filter for angle adjustment is set to no more than 0.001°. After the actuator acts on the filter at this step size, it will cause a corresponding micron-level difference in the thickness of the uncured optical adhesive layer at the edge of the filter, thereby forming a small wedge angle along the X or Y direction. The larger DeltaL is, the more obvious the change in far-field coordinates is for the same angle change, and the higher the accuracy of the direction angle calculation; the smaller the actuator step size is, the more precise the controllable increment of the filter tilt angle is, and the smaller the compensation step size of the optical adhesive wedge angle is. The synergistic effect of the two can enable high-resolution detection and correction of the multi-channel emission direction.

[0028] This invention sets the distance between the far-field plane and the exit end face at 80mm, resulting in a DeltaL of 60mm. The actuator adjusts the angle step by no more than 0.001° per operation. After acquiring the near-field and far-field spot coordinates each time, the true direction angle of each channel can be accurately calculated by the ratio of the coordinate difference to DeltaL. Based on this, the corresponding filter angle adjustment is calculated. Finally, before the optical adhesive layer cures, the maximum pairwise included angles of the exit directions of each channel are iteratively converged to the preset conditions. This ensures that the parallelism of the emitted light from the multi-channel optical component is unaffected by substrate surface errors and filter assembly / adjustment errors.

[0029] As an optional implementation, the outgoing light parallelism is the maximum pairwise angle between the outgoing light directions of each channel of the multi-channel optical component, and the objective function for calculating the parallelism is: J par =max {i<j} ||Theta i -Theta j ||2; Among them, J par Theta is a parallelism index. i and Theta j Let i and j be the outgoing light direction vectors of the i-th and j-th channels, respectively, where i and j are channel indices and i < j.

[0030] Specifically, the parallelism objective function J parThe value of Theta depends on the channel direction vectors Theta obtained in step S2. i Step S2: After each acquisition of near-field and far-field spot coordinates, the direction vectors of all channels are calculated. The direction vectors of all channels are then paired, and the L2 norm of the direction vector difference between each pair is calculated to obtain the spatial angle between the output light directions of that pair of channels. After all angles have been calculated, the maximum value is taken as the parallelism standard J for this round of adjustment. par J par This is then substituted into the judgment condition of step S4 and compared with the preset parallelism threshold: if J par If the deviation exceeds the threshold, it indicates that at least one pair of channels has an directional deviation that exceeds the allowable range. In step S3, a new angle compensation amount is calculated based on the relative deviation of the directional vectors of each channel, and the corresponding filter is tilted to change the thickness distribution of the uncured adhesive layer and correct the direction of the emitted light. After adjustment, step S2 is executed again to obtain a new directional vector and recalculate J. par This is to gradually converge the emission directions of each channel.

[0031] By using the largest pairwise included angle as the parallelism standard, that is, the included angle of the pair with the largest directional difference among all channels is used as the constraint. When J par When the angle drops below a preset threshold, the directional angle between any two channels is less than or equal to that threshold, eliminating the risk of local channels being significantly deviated and obscured by the averaging effect. This ensures that the parallelism of the multi-channel output light after final curing is not affected by the initial accuracy of the substrate and filter, but rather by J. par The convergence accuracy is determined during the iterative process of repeatedly executing steps S2 and S3.

[0032] As an optional implementation, in step S3, during the angle adjustment process, each filter changes the local thickness of the uncured optical adhesive layer beneath it, causing the optical adhesive layer to form a wedge-shaped cross-section that linearly varies along the X and / or Y directions before curing. This serves as an active optical compensation element to offset the substrate surface shape error of the multi-channel optical assembly and / or the assembly and adjustment error of each filter. Specifically, when a small rotation angle is applied to the filter around the X and / or Y axes, the lower surface of the filter tilts locally relative to the mounting surface of the substrate. One edge of the filter moves closer to the substrate and presses against the uncured optical adhesive layer below, reducing the thickness of the adhesive layer on that side. Meanwhile, the opposite edge lifts away from the substrate, increasing the thickness of the adhesive layer. This forms a wedge-shaped cross-section with a linear thickness variation along the X and / or Y directions before curing. The angle compensation amount determines the tilt direction and tilt amount of the filter, which in turn determines the change in the thickness distribution of the optical adhesive layer. This invention utilizes the liquid fluidity of the optical adhesive layer before curing to transform the angle adjustment action of the filter into the state of the wedge-shaped cross section of the optical adhesive layer. That is, a compensation matching the error is formed under the filter to actively offset the deviation caused by the combined error of the substrate surface shape and the filter assembly and adjustment at the corresponding position, so as to obtain a high consistency of multi-channel parallelism.

[0033] It should be noted that the optical adhesive layer, as an active optical compensation element, specifically involves adjusting the angle of the optical adhesive layer in step S3 to form a wedge-shaped cross-section and then curing it in step S4. This solid wedge-shaped adhesive layer is equivalent to an optical compensation sheet with a preset wedge angle in the optical path. The resulting beam deflection is opposite in direction and equivalent in magnitude to the pointing deviation introduced by the substrate surface error and filter assembly error, thus actively canceling out the aforementioned errors. The refractive index of the optical adhesive layer ranges from 1.50 to 1.58, and angular drift caused by curing shrinkage is compensated through model parameters or online calibration.

[0034] As an optional implementation, in step S3, the angle compensation amount of each filter is determined according to the angle vector of each channel. Specifically, this includes: pre-calibrating or updating online the sensitivity matrix between the change of each axis angle of each filter at the current working point and the change of the output beam error vector; using the sensitivity matrix as the system matrix, and employing the damped least squares method or the weighted least squares method, solving for the angle increment vector that minimizes the directional angle error of each channel.

[0035] Specifically, we define channel indices i = 1, 2, 3, 4, and denote the near-field plane working distance as L. N =20mm, far-field plane working distance is L F =80mm, axial spacing DeltaL=L F -L N =60mm. The coordinates of the near-field spot center measured for each channel are p. N,i=(x N ,i,y N The coordinates of the far-field spot center are p,i). F ,i=(x F ,i,y F The control quantity for each filter is defined as Alpha. i =[phi x ,i,phi y ,i,phi z ,i] T The unit is degrees, representing the angles around the X-axis, Y-axis, and Z-axis, respectively.

[0036] The orientation angle of each channel is calculated using the near-field and far-field spot center coordinates obtained in step S2, according to the following formula: theta x ,i=arctan((x F ,ix N ,i) / DeltaL); theta y i=arctan((y F ,iy N ,i) / DeltaL); Obtain the direction vector Theta of the i-th channel. i =[theta x ,i,theta y ,i] T .

[0037] Based on the direction vectors of each channel, a target direction is set (using a self-consistent method, i.e., the weighted average of the direction vectors of all channels, such as Theta*=(sum(w i Theta i )) / (sum(w i Alternatively, use the nominal value given by the optical design, such as Theta*=Theta nom Theta* represents the preset target direction vector for all channels, and the parallelism index is calculated: J par =max {i<j} ||Theta i -Theta j ||2.

[0038] Define the error vector of the i-th filter as: r i =[theta x ,i-theta x *,theta y ,i-thetay *,x N ,ix N i ref ,y N ,iy N i ref ] T ; Among them, (theta) x theta y (x) represents the target direction angle of the channel. N i ref, y N i ref () represents the coordinates of the target spot center in the near-field plane, theta x *This refers to the angular component of the target direction vector, theta, in the horizontal X direction.

[0039] Pre-calibrate or update the sensitivity matrix of each filter at the current operating point online. For the i-th filter, at the current operating point Alpha... i A small perturbation ±delta (preferably delta = 0.01°) is applied at the point, the change in the error vector is recorded, and the sensitivity matrix H is obtained using the central difference method. i (Dimensions are 4×3): in, Represents the sensitivity matrix of the i-th filter. The kth column; Let e ​​be the initial attitude angle vector of the i-th filter (including the angles around the X, Y, and Z axes). k Let be the unit direction vector along the k-th axis, and delta be the perturbation step size.

[0040] With sensitivity matrix H i As the system matrix, a damped least squares optimization problem is constructed to constrain the near-field spot center drift while reducing the directional angular error of each channel: Its closed-form solution is: DeltaAlpha i Let W be the angle increment vector of the i-th filter to be solved. i This is a weighted diagonal matrix, where lambda is the damping factor. It is the identity matrix; H is the current error vector. i Let T be the sensitivity matrix, and T be the transpose of the matrix.

[0041] The initial value of the damping factor lambda is selected based on the condition number of the sensitivity matrix, ranging from 0.001 to 10: when the condition number of the sensitivity matrix is ​​large and the matrix is ​​ill-conditioned, a larger damping factor is used to enhance solution stability; when the condition number is small and the matrix behavior is good, a smaller damping factor is used to ensure the response speed of error correction. During the iteration process, the damping factor is dynamically adjusted according to the convergence situation: if the parallelism index decreases significantly after the current iteration, the damping factor is reduced in the next iteration to accelerate convergence; if the parallelism index oscillates or increases instead of decreasing, the damping factor is increased to suppress overcorrection.

[0042] In weighted least squares, the weighting coefficient W i The allocation principle is to assign weights to the error components of each channel according to their current deviation magnitude. Channels with larger deviations are assigned larger weight coefficients, so that the angle increment obtained from the solution first compresses the angle between the channel pairs with the largest deviations; channels with smaller deviations are assigned smaller weight coefficients to avoid unnecessary disturbances to their already approaching target state. Each weight coefficient is updated before each iteration based on the latest detected channel direction vector error in step S2 to reflect the convergence progress of each channel in real time.

[0043] Furthermore, the obtained angle increment vector undergoes quantization, rounding, and amplitude limiting before being output to the actuator. The processing formula is as follows: DeltaAlpha cmd =clip(round(DeltaAlpha / s) s,[-Alpha max Alpha max ]); Among them, DeltaAlpha cmd The final angle increment command output to the actuator is s, which is the preset minimum angle step size, preferably 0.001°. Alpha max This is the preset maximum safe adjustment range for a single axis. `round` is the rounding function, and `clip` is the amplitude limiting function.

[0044] When multiple filters are arranged sequentially along the optical path, the angle adjustment of each filter is performed in reverse order of the optical path. That is, the filter at the end of the optical path is adjusted first, and the filter at the beginning of the optical path is adjusted later. If the crosstalk between channels is significant, all angle variables of all filters are merged into a unified vector, and the angle increment of all filters is solved at once using the global damped least squares method.

[0045] Specifically, the components of the original angle increment vector DeltaAlpha obtained by numerical solution are usually continuous real numbers, while the actuator driving the filter to adjust the angle is limited by its own motion resolution and can only respond to displacements that are integer multiples of the minimum angular step size s. Therefore, the numerical precision of the two is not directly matched. The processing formula is DeltaAlpha. cmd =clip(round(DeltaAlpha / s) s,[-Alpha max Alpha max First, each component of the original increment is divided by s, then rounded to the nearest integer, and then multiplied by s again to normalize the increment value to the nearest executable step size node, aligning the instruction with the executor; subsequently, the normalized value is forcibly constrained to [-Alpha] using the clip function. max Alpha max Within the single-axis symmetric interval of ], Alpha max This is the preset maximum safe adjustment range for a single axis. Its value is determined based on the elastic deformation limit of the optical adhesive layer in its uncured state. Adjustments exceeding this range may cause the adhesive layer to be excessively squeezed, overflowing, torn, or become trapped with air bubbles due to partial detachment of the filter from the adhesive layer. DeltaAlpha is then processed through quantization, rounding, and amplitude limiting. cmd The output is then sent to the actuator to drive the filter to complete the angle adjustment for this round. The thickness distribution of the optical adhesive layer changes accordingly after adjustment, thereby changing the emission direction of the light emitted from each channel.

[0046] As an optional implementation, multiple filters are arranged sequentially along the optical path. In step S3, when driving each filter to adjust its angle, the adjustment is performed in reverse order of the optical path, with the filter at the end of the optical path being adjusted first, and the filter at the beginning of the optical path being adjusted later. Specifically, in step S3, after calculating the angle increment vector of each filter in this round based on the sensitivity matrix, the actuator is first driven to adjust the last filter closest to the exit end. After its attitude is adjusted, the filter immediately in front of it is adjusted, and so on, until finally the first filter closest to the incident end is adjusted. Because the angle change of the end filter only affects the output direction of its own channel, and there is no longer a downstream filter to deflect the beam a second time, its adjustment effect can be fully observed in the next round of detection in step S2; when the filter immediately in front of it is adjusted, its output beam needs to pass through the adjusted end filter again, but since the attitude of the end filter is fixed, the direction change introduced by the adjustment of the filter immediately in front of it only adds a known fixed deflection amount when passing through the end filter to reduce crosstalk.

[0047] It should be noted that the sensitivity matrix was calibrated under the condition that each filter was subjected to an independent angular perturbation at the current operating point, while the other filters maintained their attitudes. The reverse adjustment method ensures that the attitudes of each downstream filter of the adjusted filter are fixed when each adjustment action is performed, which is consistent with the boundary conditions of the independent perturbation of each filter during calibration. This ensures that the input-output relationship described by the sensitivity matrix is ​​accurately maintained during execution, avoiding the introduction of additional model errors due to changes in boundary conditions.

[0048] As an optional implementation, the preset conditions in step S4 include: if the parallelism index J of the emitted light of the multi-channel optical component is... par Less than or equal to the preset parallelism threshold J spec The mean square error J of the center position of the light spot on the near-field plane for each channel ctr Less than or equal to the preset center error threshold C spec Furthermore, after a preset number of attitude adjustments, if the angle change corresponding to a single attitude adjustment is less than a preset change threshold, then the uncured optical adhesive layer is cured. After curing, the coordinates of the emitted light spot of each channel on the near-field plane and the far-field plane are obtained again to verify the parallelism of the cured emitted light and the near-field center position.

[0049] Specifically, after each round of step S3 completes the filter angle adjustment, step S2 is executed to reacquire the output light spot coordinates of each channel on the near-field plane and far-field plane, and the orientation angle of each channel is calculated according to the following formula: theta x ,i=arctan((x F ,ix N ,i) / DeltaL); theta y i=arctan((y F ,iy N ,i) / DeltaL); Obtain the direction vector Theta of the i-th channel. i =[theta x ,i,theta y ,i] T Therefore, the parallelism index for this round is obtained: J par =max {i<j} ||Theta i -Theta j ||2.

[0050] Meanwhile, based on the center coordinates p of the light spots of each channel on the near-field plane... N ,i=(x N ,i,y Ni) and the target center position p N i ref =(x_N,i ref ,y N i ref ), calculate the mean square error J of the near-field spot center position. ctr .

[0051] In addition, the maximum amplitude of each component in the current round angle increment vector is recorded as the single-round angle change amount ||DeltaAlpha|| inf The value is then compared with a preset change threshold, epsilon. When all the preset conditions are met, it is determined that the emitted light from each channel has reached the corresponding standard, and the uncured optical adhesive layer is then cured.

[0052] It should be noted that the parallelism index J par Meeting the standard indicates that the emission direction of each channel is within the required accuracy range; the near-field center mean square error J ctr The achievement of the target indicates that the spatial position of each channel beam in the near field plane has converged to near the target position, avoiding coupling loss caused by overall translation; the small change in angle indicates that the filter attitude has entered a steady state and the optical adhesive layer thickness distribution no longer changes significantly; the combination of the three ensures that the component state meets the requirements of directional consistency and positioning accuracy at the time of curing trigger.

[0053] After the curing process is performed, the coordinates of the emitted light spots of each channel on the near-field and far-field planes are acquired again. The parallelism index and the mean square error of the near-field center position after curing are recalculated to verify whether the parallelism and near-field center position of the emitted light after curing are still within the specified range. If the parallelism or near-field center position after curing exceeds the preset verification range, the corresponding filter is driven to perform micro-angle compensation adjustment and the curing process is performed again. Through retesting and micro-compensation settings after curing, the influence of factors such as optical adhesive layer curing shrinkage and adhesive material dispersion on the final component performance is eliminated.

[0054] As an optional implementation, the multi-channel optical component is not limited to a four-channel structure. In one alternative embodiment, the multi-channel optical component is a two-channel structure, with two filters sequentially arranged on the substrate along the optical path direction, each corresponding to one of the two independent optical channels. In step S2, the dual-plane imaging component 1 acquires the spot coordinates of the two channels on the near-field plane and the far-field plane, totaling four spot center position information. Step S3 calculates the angle compensation amount of the two filters based on this information, and through closed-loop iteration from steps S2 to S4, the output light directions of the two channels are converged to parallelism. In another alternative embodiment, the multi-channel optical component is an eight-channel structure, with eight filters sequentially arranged on the substrate along the optical path direction. Step S2 simultaneously acquires the center coordinates of sixteen spots on the near-field plane and the far-field plane of the eight channels. Step S3 constructs a sensitivity matrix based on the eight-channel direction vectors, uniformly solves the angle increment vectors of the eight filters, and achieves high parallelism of the eight-channel output light after iterative convergence in step S4. The above implementations with different channel numbers all employ the same active alignment and mounting process in steps S1 to S4. The only differences lie in the number of spot coordinates acquired in step S2, the dimension of the sensitivity matrix in step S3, and the parallelism objective function J. par The number of pairwise combinations changes accordingly.

[0055] like Figure 1 , Figure 2 , Figure 3 , Figure 4 and Figure 5 As shown, an active alignment and mounting device for implementing the above method includes: A three-axis angle adjustment mechanism is used to clamp multiple filters and drive the filters to adjust their angles. Dual-plane imaging component 1 is used to acquire the coordinates of the emitted light spot of the multi-channel optical component on at least two planes with different propagation distances; Collimating light source 2 is fixedly mounted at the incident end of the multi-channel optical assembly; A curing assembly for performing a curing process on an uncured optical adhesive layer applied between a substrate and a filter; The controller is communicatively connected to the three-axis angle adjustment mechanism, the dual-plane imaging component 1, and the solidification component.

[0056] Specifically, the collimating light source 2 is a collimator, fixedly mounted at the incident end of the multi-channel optical assembly, emitting incident beams to multiple filters arranged sequentially along the optical path. A three-axis angle adjustment mechanism clamps each filter. While the uncured optical adhesive layer remains in a liquid, flowable state, it receives angle increment commands from the controller, driving the filters to deflect around the X and Y axes. This causes a local tilt of the lower surface of the filter relative to the substrate mounting surface, thereby changing the thickness distribution of the optical adhesive layer at various locations below the filter. A dual-plane imaging assembly 1 is arranged on one side of the exit end of the multi-channel optical assembly. Inside, it acquires images of the emitted light spots from each channel on two imaging planes spaced apart along the optical axis. After image processing, the center coordinates of the light spots on the near-field and far-field planes of each channel are extracted, and the coordinate data is transmitted to the controller in real time. After receiving the two sets of coordinates uploaded by the dual-plane imaging component 1, the controller repeats steps S2 to S4. First, it calculates the direction vector and parallelism value of each channel. If the preset conditions are not met, it constructs a sensitivity matrix and solves the angle increment vector. After quantization, rounding, and amplitude limiting, it generates an angle adjustment command and sends it to the three-axis angle adjustment mechanism to drive the filter to complete one round of attitude adjustment. Then, it obtains the dual-plane coordinates again to enter the next round of iteration. When the parallelism index and angle change continuously meet the preset conditions, the controller sends a curing command to the curing component. The curing component performs ultraviolet irradiation or thermal curing treatment on the optical adhesive layer to fix each filter in the current position.

[0057] Furthermore, the dual-plane imaging assembly 1 includes a beam splitter 11, a refraction assembly 12, and an imaging plane 13. The refraction assembly 12 works in conjunction with the beam splitter 11. After the outgoing light from the multi-channel optical assembly is incident on the beam splitter 11, a portion is directly transmitted to form the first beam 3, while the other portion is reflected into the refraction assembly to form the second beam 4. The refraction assembly 12 reflects the second beam 4 multiple times, increasing its propagation distance to the imaging plane 13. Ultimately, the second beam 4 and the first beam 3 converge together on the imaging plane 13. It should be noted that the region on the imaging plane 13 corresponding to the landing point of the first beam 3 is the near-field plane, and the region corresponding to the landing point of the second beam 4 is the far-field plane. This allows for the simultaneous acquisition of spot images at two different propagation distances on a single imaging plane 13, enabling the controller to extract the center coordinates of the spot for each channel under near-field and far-field conditions.

[0058] Furthermore, the dual-plane imaging assembly 1 also includes a housing 14, a mounting base 15 fixed within the housing 14, and a sliding base 16 slidably mounted on the housing 14. The beam splitter 11 is fixed on the mounting base 15, and the sliding base 16 can slide relative to the mounting base 15 along the optical axis to adjust the distance between it and the mounting base 15; the imaging plane 13 is disposed above the mounting base 15 and its position remains fixed. The refraction assembly 12 includes a first refractor 121, a second refractor 122, and a third refractor 123, wherein the first refractor 121 and the second refractor 122 are symmetrically arranged vertically on the sliding base 16, and the third refractor 123 is fixed on the mounting base 15 and cooperates with the second refractor 122; after the second beam 4 is reflected from the beam splitter 11, it is refracted by the first refractor 121 and then refracted again by the second refractor 122, and finally reflected by the third refractor 123 to the far-field region on the imaging plane 13. By adjusting the position of the sliding seat 16, the positions of the first refracting mirror 121 and the second refracting mirror 122 in the optical path can be changed, thereby flexibly setting the total optical path distance of the second beam 4 from the exit end face of the multi-channel optical component to the far-field plane to adapt to the needs of different focal lengths or parallelism detection sensitivities.

[0059] It should be noted that, as Figure 6 As shown, collimated light source 2 enters from the lower right side of the multi-channel optical assembly at an incident angle of 8°. It passes sequentially through multiple filters within the multi-channel optical assembly along the optical path. Each filter reflects the light of its corresponding wavelength, while the remaining wavelengths continue to be transmitted to the next filter, ultimately forming four-way beam splitting. The incident light propagates in a Z-shape within the multi-channel optical assembly. The outgoing light from each channel exits from the upper left side of the multi-channel optical assembly and is captured by a 20mm near-field camera and an 80mm far-field camera, respectively, to obtain light spot images. In the figure, the light spots of each channel correspond to near-field (Near) and far-field (Far) spots on the near-field and far-field planes, respectively. The coordinate difference between these two points is calculated by the controller and used to invert the required angle compensation for each filter.

[0060] like Figure 7 As shown, the dual-plane imaging component 1 acquires the emitted light spot images of each channel on the near-field plane and the far-field plane, respectively. After image processing, the center coordinates of the light spots for each channel are extracted. The upper half of the figure shows the four light spot centers on the near-field plane, corresponding to channels 1 to 4; the lower half shows the four light spot centers on the far-field plane, also corresponding to channels 1 to 4. The difference between the coordinates of the light spot centers of the same channel on the near-field and far-field planes is used to calculate the actual direction angle vector of the emitted light from that channel. By comparing the direction angle vectors of all channels, the parallelism index J between the emitted light from each channel can be obtained. par . Figure 7The labels CH1 to CH4 indicate the channel number to which each spot belongs, which facilitates the controller to perform paired calculations of the coordinates of the same channel on two planes.

[0061] like Figure 8 As shown, the horizontal axis represents the parallelism index (the maximum pairwise angle between the output light directions of each channel, in degrees), and the vertical axis represents the frequency. The figure includes two distribution curves: the parallelism distribution before compensation and the parallelism distribution after compensation (step size 0.001°). Two specification lines are also marked: an upper specification limit of 0.10° and a relaxed specification of 0.20°. As can be seen from the figure, the distribution curve on the right represents the parallelism distribution before compensation, with a mean of approximately 0.3349°, and a large number of samples falling outside the 0.10° specification line. The distribution curve on the left represents the parallelism distribution after compensation, significantly narrowing and shifting to the left, with the mean dropping to approximately 0.05°, and the vast majority of samples concentrated within the 0.10° specification line. This figure visually reflects the improvement effect of the proposed method on parallelism through closed-loop iterative compensation, achieving a yield of approximately 99.53% at a 0.10° specification. The labels for CH1 to CH4 in the figure have been omitted; this distribution plot represents the statistical results of the overall parallelism of the four channels.

[0062] like Figure 9 As shown, the horizontal axis represents the minimum step size of the actuator, in meters (1 meter = 0.001°), ranging from 1 meter to 10 meters. The left vertical axis represents the P95 quantile of the parallelism after compensation (solid blue line), and the right vertical axis represents the yield at a 0.10° specification (dashed green line). The graph shows that when the step size is 1 meter (0.001°), the compensated P95 is approximately 0.0770°, and the yield is approximately 99.84%, both optimal values ​​for each step size. As the step size increases from 1 meter to 10 meters, the P95 quantile shows a slow upward trend, while the yield shows a slow downward trend, indicating that a smaller step size results in higher resolution for angle adjustment and correspondingly improved compensation accuracy. This graph demonstrates that setting the minimum step size of the actuator to 0.001° achieves the best overall performance in terms of both P95 quantile and yield.

[0063] The above description is merely a preferred embodiment of the present invention. Those skilled in the art will understand that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the present invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. An active alignment and mounting method for a multi-channel optical component, characterized in that, Multichannel optical components include a substrate and multiple filters. The active alignment and mounting method includes the following steps: S1. Apply uncured optical adhesive layers between the mounting surface of the substrate and the multiple filters respectively; S2. Obtain the first spot center coordinates of each channel of the multi-channel optical component on the near-field plane, and obtain the second spot center coordinates of each channel on the far-field plane; wherein, the near-field plane and the far-field plane are two planes with different propagation distances arranged at intervals along the optical axis. S3. Based on the center coordinates of the first and second light spots of each channel, calculate the angle compensation amount for each filter, and drive the corresponding filter to adjust its angle according to the angle compensation amount, so as to change the thickness distribution of the optical adhesive layer at the corresponding position; wherein, the angle vector of each channel is calculated according to the following formula: theta x ,i=arctan((x far ,i-x near ,i) / DeltaL); theta y ,i=arctan((y far ,iy near ,i) / DeltaL); Then, the angle compensation amount of each filter is determined according to the angle vector of each channel, wherein theta x ,i and theta y ,i represent the angular components of the i-th channel in the horizontal and vertical directions, respectively, and x represents the angular components of the i-th channel in the horizontal and vertical directions. near ,i and y near Let i be the coordinates of the center of the first spot of the i-th channel on the near-field plane, and x be the coordinates of the center of the first spot of the i-th channel. far ,i and y far ,i is the center coordinate of the second spot of the i-th channel on the far-field plane, and DeltaL is the axial distance between the near-field plane and the far-field plane; S4. Repeat steps S2 to S3 until the parallelism of the emitted light from the multi-channel optical component meets the preset conditions, then cure the optical adhesive layer.

2. The active alignment and mounting method for multi-channel optical components according to claim 1, characterized in that, The axial distance between the near-field plane and the emission end face of the multi-channel optical component is 20 mm, and the axial distance between the far-field plane and the emission end face is 80 mm.

3. The active alignment and mounting method for multi-channel optical components according to claim 1, characterized in that, In step S3, the minimum single angular step of the actuator used to drive the filter to adjust the angle is less than or equal to 0.001°.

4. The active alignment and mounting method for multi-channel optical components according to claim 1, characterized in that, The outgoing light parallelism is the maximum pairwise angle between the outgoing light directions of each channel of the multi-channel optical component, and the objective function for calculating the parallelism is: J par =max {i<j} ||Theta i -Theta j ||2; Among them, J par Theta is a parallelism index. i and Theta j Let i and j be the outgoing light direction vectors of the i-th and j-th channels, respectively, where i and j are channel indices and i < j.

5. The active alignment and mounting method for multi-channel optical components according to claim 1, characterized in that, In step S3, during the angle adjustment process, each filter changes the local thickness of the uncured optical adhesive layer below it, so that the optical adhesive layer forms a wedge-shaped cross section that changes linearly along the X and / or Y directions before curing, so as to act as an active optical compensation element to offset the substrate surface error of the multi-channel optical assembly and / or the assembly and adjustment error of each filter.

6. The active alignment and mounting method for multi-channel optical components according to claim 1, characterized in that, In step S3, the angle compensation amount of each filter is determined according to the angle vector of each channel. Specifically, this includes: pre-calibrating or updating online the sensitivity matrix between the change of each axis angle of each filter at the current working point and the change of the output beam error vector; using the sensitivity matrix as the system matrix, the damped least squares method or weighted least squares method is used to solve for the angle increment vector that minimizes the directional angle error of each channel.

7. The active alignment and mounting method for multi-channel optical components according to claim 6, characterized in that, The angle increment vector obtained by solving is quantized, rounded, and amplitude-limited before being output to the actuator. The processing formula is: DeltaAlpha cmd =clip(round(DeltaAlpha / s) s,[-Alpha max Alpha max ]); Where s is the preset minimum angular step size, and Alpha max This is the preset maximum safe adjustment range for a single axis.

8. The active alignment and mounting method for multi-channel optical components according to claim 1, characterized in that, The plurality of filters are arranged sequentially along the optical path direction; in step S3, when driving each of the filters to adjust its angle, the adjustment is performed in reverse order of the optical path, wherein the filter located at the end of the optical path is adjusted first, and the filter located at the beginning of the optical path is adjusted later.

9. The active alignment and mounting method for a multi-channel optical component according to claim 4, characterized in that, The preset conditions in step S4 include: if the parallelism index J of the emitted light of the multi-channel optical component is... par Less than or equal to the preset parallelism threshold J spec The mean square error J of the center position of the light spot on the near-field plane of each of the aforementioned channels ctr Less than or equal to the preset center error threshold C spec After a preset number of consecutive attitude adjustments, if the angle change corresponding to a single attitude adjustment is less than a preset change threshold, then the uncured optical adhesive layer is cured. After curing, the coordinates of the emitted light spots of each channel on the near-field plane and the far-field plane are obtained again to verify the parallelism of the cured emitted light and the near-field center position.

10. An active alignment and mounting apparatus for implementing the active alignment and mounting method according to any one of claims 1 to 9, characterized in that, include: A three-axis angle adjustment mechanism is used to clamp multiple filters and drive the filters to adjust their angles. A dual-plane imaging component is used to acquire the coordinates of the emitted light spot of a multi-channel optical component on at least two planes with different propagation distances. A collimated light source is fixedly mounted at the incident end of the multi-channel optical assembly; A curing component is configured to perform a curing process on an uncured optical adhesive layer applied between the substrate and the filter; The controller is communicatively connected to the three-axis angle adjustment mechanism, the dual-plane imaging component, and the solidification component.

Images