An OCS multi-port optical path self-calibration packaging method

Abstract

The application provides an OCS multi-port optical path self-calibration packaging method, relates to the technical field of precision packaging and optical coupling calibration of optical communication devices, and comprises the following steps: firstly, input fiber collimator arrays, first micro-mirror arrays, second micro-mirror arrays and output fiber collimator arrays are installed in the same position of a dual-state adjustment frame, and a multi-port optical path error vector is constructed through multi-frequency calibration light and coded exchange states; then, reversible thermal elastic identification is performed on an elastic adjustment beam, and an influence matrix between optical path responses and adjustment units is established; finally, bidirectional laser stress writing areas, discrete writing positions and writing energy are determined according to the permanent trimming influence matrix, packaging deviations are continuously corrected after the clamp is released, temperature cycles and the cover plate is connected, and an electronic calibration file is generated. The application realizes online identification, permanent trimming and final state calibration of the multi-port optical path, and improves the packaging consistency, stability and batch manufacturing adaptability of the OCS.

Description

Technical Field

[0001] This invention relates to the field of precision packaging and optomechanical coupling calibration technology for optical communication devices, specifically an OCS multi-port optical path self-calibration packaging method. Background Technology

[0002] Free-space MEMS-type OCS (Optical Channel Switch) controls the reflection angle of a two-stage micromirror array to guide the collimated beam from any input fiber collimator to a designated output fiber collimator, thereby establishing a reconfigurable optical connection between the input and output ports. As the OCS port size increases, spatial position errors, angular errors, and base deformation between the input fiber collimator array, output fiber collimator array, first micromirror array, and second micromirror array simultaneously affect multiple optical paths, causing discrete port insertion loss, decreased coupling at edge ports, increased crosstalk at non-target ports, and mechanical assembly errors occupying the micromirror drive margin.

[0003] Current packaging processes typically employ an external six-degree-of-freedom (6DOF) orientation platform to clamp optical components. After the output power of the selected optical path reaches a set value, the optical component is secured using adhesives, laser welding, or mechanical fasteners. The external orientation platform and the final packaging support structure have different load transfer paths. The optical component is subjected to different constraints in the platform-held state, adhesive curing state, weld cooling state, and clamp release state. Therefore, the optimal position of the optical path obtained during the orientation adjustment phase will shift after packaging and fixation. Furthermore, maximizing the local power of a single port can cause the adjustment action of an optical component to simultaneously alter multiple related optical paths, resulting in the adjusted port deviating from the target state again.

[0004] Existing active alignment methods typically use an external displacement platform as a reversible actuator and an adhesive or welded structure as a fixed structure. These two are separated in terms of spatial position, stiffness boundaries, and force direction, meaning the displacement sensitivity measured by the attitude adjustment platform cannot be directly converted into a permanent pose change of the packaged structure. Retaining piezoelectric actuators, micromotors, or continuous heaters after packaging increases device size, drive circuitry, operating power consumption, and long-term drift sources, hindering the compactness and long-term stable operation of high-port OCSs.

[0005] Furthermore, high-port OCS performs power scans line by line on all input and output ports. The number of measurements increases rapidly with the number of ports, and it takes a considerable amount of time to regain the complete optical path state after a single attitude adjustment. During this period, changes in ambient temperature and light source power will affect the measurement results. Existing technologies lack a packaging method that can perform non-destructive response identification on the final support structure, implement permanent attitude writing based on the overall error of multiple ports, and update the physical model of the device after each writing operation. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an OCS multi-port optical path self-calibration packaging method, which solves the following problems: 1. This solves the problem that when existing OCS uses an external attitude adjustment platform for optical path alignment and then completes the encapsulation and fixation through adhesive bonding or welding, the attitude adjustment load path is inconsistent with the final load path. This causes changes in the optical component's pose after the fixture is released, the adhesive layer cures, the weld cools, and the encapsulation cover is connected. As a result, the optical path state during the attitude adjustment stage cannot be used as the final state of encapsulation.

[0007] 2. It solves the problem that in multi-port OCS, the same structural deviation of the input array, output array and micromirror array simultaneously affects multiple optical paths. Traditional port-by-port power maximization adjustment will change other paths while improving the current path. It lacks the overall error expression and collaborative adjustment basis for the insertion loss of the target path, port uniformity and crosstalk of non-target paths.

[0008] 3. This solves the problems that existing packaging structures only serve a fixing function, making it impossible to measure the actual direction of each bearing position on the multi-port optical path without changing the permanent state. Furthermore, the theoretical finite element model cannot fully include the processing error of the elastic beam, batch differences in materials, local differences in laser absorption, and cumulative residual stress changes, and the permanent adjustment amount depends on empirical parameters.

[0009] 4. It solves the problems of long time required for high-port OCS to scan all input-output combinations line by line, difficulty in completing the acquisition of multi-port responses after a single thermal disturbance or permanent adjustment within the same stable state, and lack of unified device-level correlation for optical path data, thermoelastic data, laser writing data and environmental verification data generated during the production stage.

[0010] To achieve the above objectives, the present invention provides the following technical solution: an OCS multi-port optical path self-calibration packaging method, comprising the following steps: Sp1. Install the input fiber collimator array, the first micro-mirror array, the second micro-mirror array and the output fiber collimator array on the optical component carrier ring of the corresponding co-position dual-state attitude adjustment frame, fix the base connecting ring of each co-position dual-state attitude adjustment frame to the packaging base, and complete the coarse positioning of each optical component through the external attitude adjustment mechanism. Sp2. Divide the input ports into groups of 8 each, configure different modulation frequencies for each input port in the same input port group, inject calibration light into each input port, control the first micromirror array and the second micromirror array to enter the encoding and switching state in sequence, collect the time-domain optical power signal of each output port, and perform synchronous demodulation according to each modulation frequency to obtain the transmission coefficient of the main connection path and the transmission coefficient of the non-main connection path. Sp3. Construct a multi-port optical path error vector based on the logarithmic residual between the transmission coefficient of the main connection path and the target transmission coefficient, the logarithmic residual between the transmission coefficient of the non-main connection path and the crosstalk limit value, and the weight of each calibration path. Sp4. Apply the first identification current and the second identification current to the resistance heating layer on each elastic attitude adjustment beam in each co-position dual-state attitude adjustment frame in sequence. Collect the multi-port optical path error vectors in the first stable temperature rise state, the second stable temperature rise state and the cooling recovery state respectively. Establish the reversible thermoelastic influence matrix based on the difference in optical path error vectors and temperature difference between the two stable temperature rise states. Sp5. Perform singular value decomposition on the reversible thermoelastic influence matrix, retain optical path error modes with singular values ​​not less than 5% of the maximum singular value, select the attitude adjustment unit and laser writing position according to the projection of the current multi-port optical path error vector on the retained optical path error mode, perform calibration laser writing in the first laser stress writing area or the second laser stress writing area of ​​the selected attitude adjustment unit, and establish a permanent adjustment influence matrix based on the difference of the multi-port optical path error vector before and after calibration laser writing under the same temperature state and the calibration laser energy. Sp6. Based on the current multi-port optical path error vector, permanent trimming influence matrix, single laser energy limit, cumulative laser energy limit, double-sided writing mutual exclusion condition of the same elastic attitude adjustment beam, adjacent writing bit interval condition, local temperature limit, residual stress limit, and the number of reserved final state writing bits, determine the laser stress writing area, laser writing bit, and laser writing energy of the current trimming cycle. Sp7. Perform laser writing according to the determined laser stress writing area, laser writing position and laser writing energy. After the in-situ dual-state attitude adjustment frame recovers to the reference temperature, re-acquire the multi-port optical path error vector. Update the permanent adjustment influence matrix according to the measured optical path error change before and after this laser writing. Sp8, repeat Sp6 and Sp7 until the absolute value of the weighted logarithmic error of each calibration path is no higher than 0.02. Release the holding load of the external attitude adjustment mechanism in three levels: 100%, 50% and 0%. After releasing the holding load at each level, re-execute the multi-port optical path measurement and laser writing. Sp9. After the external attitude adjustment mechanism is completely released, the optical path transmission matrix is ​​collected in the stable state of -20℃, 25℃ and 70℃ in sequence to complete one heating cycle and one cooling cycle. After returning to 25℃, the retained optical path error is laser-written. Then, the encapsulation cover is connected in segments according to the central symmetry order. After each group of connection areas is completed, the calibration path data is collected and laser writing is performed. Sp10, measure the optical path transmission matrix between all input ports and all output ports, and write the initial optical path transmission matrix, reversible thermoelastic effect matrix, permanent adjustment effect matrix, laser writing record, holding load release data, temperature cycling data, package cover connection data and the final optical path transmission matrix into the electronic package calibration file associated with the OCS device serial number.

[0011] Preferably, the co-position dual-state attitude adjustment frame includes a base connecting ring, an optical component support ring located inside the base connecting ring, and attitude adjustment units connected between the base connecting ring and the optical component support ring. Each attitude adjustment unit includes an outer fixed section continuously connected to the base connecting ring, an inner force transmission section continuously connected to the optical component support ring, and an elastic attitude adjustment beam located between the outer fixed section and the inner force transmission section. Thermal isolation grooves are respectively provided on both sides of the elastic attitude adjustment beam. A resistance heating layer, a temperature detection line, a first laser stress writing area, and a second laser stress writing area are provided on the surface of the elastic attitude adjustment beam. The first laser stress writing area and the second laser stress writing area are respectively located on opposite sides of the width direction of the elastic attitude adjustment beam.

[0012] Preferably, the first laser stress writing area and the second laser stress writing area are each provided with 12 discrete laser writing positions along the length direction of the elastic attitude adjustment beam. The center distance between adjacent laser writing positions is 180μm. An unwritten isolation segment is provided between adjacent laser writing positions. Each laser writing position is established with a coordinate index according to the attitude adjustment frame number, attitude adjustment unit number, laser stress writing area number and position number. In the same adjustment cycle, only one laser stress writing area is selected for the same elastic attitude adjustment beam.

[0013] Preferably, the OCS includes 64 input ports and 64 output ports. The 64 input ports are divided into 8 input port groups. The 8 input ports in the same input port group are configured with 8 modulation frequencies in the range of 5kHz to 40kHz. The first micromirror array and the second micromirror array enter 16 coding switching states in sequence. Each input port participates in the output path of the central region, the output path of the edge region, and the output path of the corner region.

[0014] Preferably, the calibration path in Sp3 covers the central, edge, and corner regions of the input fiber collimator array, covers the central, edge, and corner regions of the output fiber collimator array, and passes through the central, edge, and corner regions of the first and second micromirror arrays, respectively. The weights of the edge region path, corner region path, and long path across the mirror array are higher than the weights of the short path in the central region.

[0015] Preferably, the first stable temperature rise in Sp4 is 5°C higher than the reference temperature, and the second stable temperature rise is 10°C higher than the reference temperature. Each attitude adjustment unit performs two repeated identifications. When the directional angle between the unit temperature rise optical path influence vectors obtained from the two identifications is not higher than 5° and the amplitude difference is not higher than 10%, the corresponding unit temperature rise optical path influence vector is written into the reversible thermoelastic influence matrix.

[0016] Preferably, in Sp5, an attitude adjustment unit is selected based on the right singular vector component corresponding to the retained optical path error mode. Calibration laser writing is performed at the unused laser writing position of the selected attitude adjustment unit. After the calibration laser writing is completed, when the difference between the temperature of the attitude adjustment beam and the reference temperature before the calibration laser writing is not higher than 0.5℃, the calibration multi-port optical path error vector is collected, and the vector obtained by dividing the difference between the multi-port optical path error vector before and after calibration by the effective calibration laser energy is written into the permanent adjustment influence matrix.

[0017] Preferably, in Sp6, the first laser stress writing area and the second laser stress writing area of ​​each attitude adjustment unit are set as laser writing variables with opposite directions. The laser writing variable that minimizes the weighted sum of the remaining multi-port optical path error, laser writing energy and writing difference between adjacent attitude adjustment units is solved. The actual execution amount of the initial adjustment stage is set to 50% of the solution result, and the actual execution amount of the fine adjustment stage is set to 20% of the solution result.

[0018] Preferably, in Sp7, the prediction residual between the actual multi-port optical path error change and the predicted change obtained by multiplying the permanent adjustment influence matrix by the actual laser writing vector is calculated. The prediction residual is then distributed to the corresponding column of the permanent adjustment influence matrix along the direction of the actual laser writing vector. When the directional angle between the actual multi-port optical path error change and the predicted change is higher than 30°, the laser writing bit involved in this laser writing is stopped, and the calibration laser writing is re-executed at the adjacent unused laser writing bit.

[0019] Preferably, the electronic packaging calibration file in Sp10 establishes a data hierarchy according to the OCS device serial number, optical component number, co-position dual-state attitude adjustment frame number, attitude adjustment unit number, laser stress writing area number, and laser writing position number, and records design prior data, incoming material inspection data, input calibration light monitoring data, output time-domain optical power data, heating current data, attitude adjustment beam temperature data, laser pulse data, holding load release data, temperature cycle data, packaging cover plate connection data, and final full-port acceptance data.

[0020] This invention provides an OCS multi-port optical path self-calibration packaging method. It has the following beneficial effects: 1. The present invention sets the resistance heating layer and the bidirectional laser stress writing area on the same elastic attitude adjustment beam, so that the reversible thermoelastic identification and permanent residual stress writing are transmitted along the same outer fixed section, elastic attitude adjustment section, inner force transmission section and optical component bearing ring, reducing the difference between the external attitude adjustment load path and the final packaging bearing path, and transforming the packaging bearing structure into a device-level physical identification and permanent adjustment structure.

[0021] 2. This invention obtains the transmission coefficients of the main connection path and the non-main connection path between multiple input ports and multiple output ports by multi-frequency calibration light, encoding and switching status and synchronous demodulation. It incorporates the target path insertion loss, port consistency and non-target path crosstalk into the same multi-port optical path error vector, reducing the time occupied by each scan to environmental stabilization.

[0022] 3. The present invention performs optical path error mode screening based on the reversible thermoelastic influence matrix measured on the current device, and establishes a permanent adjustment influence matrix by writing with low-energy calibration laser. After each permanent writing, the permanent adjustment influence matrix is ​​updated according to the actual optical path changes, so that the subsequent writing parameters include the processing error of the current attitude frame, the material state, the laser absorption state and the cumulative residual stress state.

[0023] 4. After the external attitude adjustment mechanism is disengaged in stages, temperature is cycled, and the package cover is connected, the present invention continues to perform multi-port optical path measurement and permanent laser adjustment. After the adjustment is completed, the precision displacement actuator with continuous power supply is not retained in the OCS. The structural data, optical path data, thermoelectric data, laser writing data and environmental verification data of the production stage are associated with the same device serial number through the electronic package calibration file. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the overall structure of the OCS multi-port optical path self-calibration packaging system of the present invention; Figure 2 This is a schematic diagram of the overall structure of the co-position dual-state attitude adjustment frame of the present invention; Figure 3 This is a schematic diagram of the co-position dual-state attitude adjustment unit structure of the present invention; Figure 4 This is a schematic diagram of the bidirectional laser stress writing region and discrete laser writing bit arrangement of the present invention; Figure 5 This is a schematic diagram of the multi-frequency calibration optical and code switching state measurement process of the present invention; Figure 6 This is a schematic diagram of the reversible thermoelasticity influence matrix identification process of the present invention; Figure 7 This is a schematic diagram of the permanent adjustment closed-loop process of the present invention; Figure 8 This is a schematic diagram of the complete packaging method of the present invention; Figure 9 This is a system interface diagram of the present invention; Figure 10 This is a system diagram illustrating the core principle of the present invention; Figure 11 This is a process flow diagram of the present invention; Figure 12 This is a diagram of the key structural system of the present invention; Figure 13 This is a diagram of the final state verification system of the present invention; Figure 14 This is a data asset system diagram of the present invention. Detailed Implementation

[0025] The technical solutions of 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. Specific Implementation Example 1: like Figures 1 to 14 As shown, an OCS multi-port optical path self-calibration packaging method involves mounting an input fiber collimator array, a first micromirror array, a second micromirror array, and an output fiber collimator array onto the optical component carrier rings of corresponding co-located dual-state attitude adjustment frames, and fixing the base connecting rings of each co-located dual-state attitude adjustment frame to the packaging base. Each co-located dual-state attitude adjustment frame includes a base connecting ring, an optical component carrier ring, and multiple attitude adjustment units connected between the two. Each attitude adjustment unit includes an outer fixed section, an elastic attitude adjustment beam, and an inner force transmission section. The elastic attitude adjustment beam is provided with a resistance heating layer, a temperature detection circuit, a first laser stress writing area, and a second laser stress writing area. The method utilizes electrical... The heat-resistant layer generates reversible thermoelastic deformation and collects changes in the multi-port optical path, establishing a reversible thermoelastic influence matrix corresponding to each attitude adjustment unit; calibrated laser writing is performed through the first or second laser stress writing area, establishing a permanent adjustment influence matrix corresponding to each laser writing direction; the laser stress writing area, laser writing position, and laser writing energy are determined based on the current multi-port optical path error vector and the permanent adjustment influence matrix; after each laser writing, the multi-port optical path is remeasured and the permanent adjustment influence matrix is ​​updated until the external attitude adjustment mechanism is completely released, the temperature cycle is completed, and the state of the multi-port optical path after the encapsulation cover is connected meets the set acceptance conditions.

[0027] The base connecting ring, the optical component carrier ring, and multiple attitude adjustment units are integrally machined from low thermal expansion metal. The base connecting ring is provided with positioning surfaces and positioning holes that mate with the packaging base. The optical component carrier ring is provided with mounting grooves and end face limiting platforms that mate with the corresponding optical components. The elastic attitude adjustment beam spans the thermal isolation gap between the base connecting ring and the optical component carrier ring. Thermal isolation grooves extending along the beam length are provided on both sides of the elastic attitude adjustment beam.

[0028] The resistance heating layer includes an insulating layer, a serpentine metal resistance layer, and a protective layer. The serpentine metal resistance layer is located in the middle of the elastic attitude adjustment beam and connected to a constant current drive circuit. The temperature detection line is located adjacent to the serpentine metal resistance layer or is composed of the real-time resistance value of the serpentine metal resistance layer. The controller obtains the real-time temperature of the elastic attitude adjustment beam based on the heating current, terminal voltage, reference resistance, and resistance temperature coefficient.

[0029] The first laser stress writing area and the second laser stress writing area are located on opposite sides of the width direction or opposite surfaces of the thickness direction of the elastic attitude adjustment beam, respectively. The first laser stress writing area and the second laser stress writing area are divided into multiple discrete laser writing positions along the length direction of the elastic attitude adjustment beam, and an unwritten isolation segment is set between adjacent laser writing positions. Each laser writing position is indexed according to the attitude adjustment frame number, attitude adjustment unit number, laser stress writing area number and position number.

[0030] The input port is divided into multiple input port groups. Each input port in the same input port group is injected with calibration light with different modulation frequencies. The first micromirror array and the second micromirror array are controlled to enter multiple encoding switching states in sequence. The output photoelectric detection array synchronously collects the time-domain optical power signal of each output port and performs synchronous demodulation according to each modulation frequency to obtain the main connection path transmission coefficient and the non-main connection path transmission coefficient.

[0031] A multi-port optical path error vector is constructed based on the logarithmic residual between the transmission coefficient of the main connection path and the corresponding target transmission coefficient, the logarithmic residual between the transmission coefficient of the non-main connection path and the corresponding crosstalk limit value, and the weight of each calibration path. The calibration path covers the central region, edge region and corner region of the input fiber collimator array, covers the central region, edge region and corner region of the output fiber collimator array, and passes through different spatial regions of the first micromirror array and the second micromirror array.

[0032] Two identification currents corresponding to stable temperature rises are applied to the resistive heating layer of each attitude adjustment unit to obtain the multi-port optical path error vectors under the two stable temperature rise states. The unit temperature rise optical path influence vector is determined based on the difference between the two multi-port optical path error vectors and the difference between the two stable temperature rises. The unit temperature rise optical path influence vectors of all attitude adjustment units are arranged according to the attitude adjustment unit number to form a reversible thermoelastic influence matrix.

[0033] Singular value decomposition is performed on the reversible thermoelastic influence matrix. Controllable optical path error modes are selected according to the ratio between the singular values ​​and the maximum singular value. The current multi-port optical path error vector is projected onto the controllable optical path error mode. Based on the projection results, the corresponding attitude adjustment unit and the unused laser writing bit are selected to perform low-energy calibration laser writing. The permanent optical path influence vector per unit laser energy is determined based on the difference between the multi-port optical path error vectors before and after calibration laser writing under the condition of consistent temperature and the calibration laser energy.

[0034] The permanent optical path influence vectors of each unit laser energy are arranged to form a permanent adjustment influence matrix. The first laser stress writing area and the second laser stress writing area of ​​each attitude adjustment unit are set as writing variables with opposite directions. The laser writing vector of the current adjustment cycle is solved under the constraints of single laser writing energy, cumulative laser writing energy, bilateral writing mutual exclusion, adjacent writing bit interval, local temperature, residual stress of elastic attitude adjustment beam and remaining number of writing bits.

[0035] After laser writing is completed, wait for the co-position dual-state attitude adjustment frame to return to the reference temperature, then collect the multi-port optical path error vector after laser writing, update the permanent adjustment influence matrix according to the actual optical path error change, the actual laser writing vector and the permanent adjustment influence matrix before writing, and use the updated permanent adjustment influence matrix in subsequent adjustment cycles.

[0036] After the multi-port optical path error enters the initial acceptance range, the holding load of the external attitude adjustment mechanism is released step by step according to multiple load levels. After each level of holding load is released, the multi-port optical path error vector is re-acquired and laser writing is performed. After the external attitude adjustment mechanism is completely released, the OCS is subjected to low temperature, room temperature and high temperature cycles, and the encapsulation cover is connected in a centrally symmetrical order. Laser writing is performed on the optical path error retained after returning to room temperature and the optical path error retained after the encapsulation cover is connected.

[0037] After completing the packaging and adjustment, the optical path transmission matrix between all input ports and all output ports is measured, and the design prior data, incoming material inspection data, initial optical path transmission matrix, reversible thermoelastic effect matrix, permanent adjustment effect matrix, cumulative laser energy of each laser writing bit, fixture load release data, temperature cycle data, package cover connection data, and final optical path transmission matrix are written into the electronic packaging calibration file associated with the OCS device serial number. Specific Implementation Example 2: like Figures 1 to 14 As shown, based on the content of the above specific embodiments, the following content is further disclosed: Overall implementation architecture: This embodiment employs a free-space MEMS-type OCS with 64 input ports and 64 output ports. The OCS includes an input fiber optic collimator array, a first micromirror array, a second micromirror array, an output fiber optic collimator array, a packaging base, a mirror array driving circuit, and a packaging cover. The input fiber optic collimator array, the output fiber optic collimator array, the first micromirror array, and the second micromirror array are each mounted on the packaging base using independent, co-located dual-state attitude adjustment frames.

[0039] The input and output fiber collimator arrays each have twelve corresponding dual-state attitude adjustment frames, with four attitude adjustment units having elastic attitude adjustment beams arranged along the lateral action direction, four having elastic attitude adjustment beams arranged along the vertical action direction, and two having elastic attitude adjustment beams arranged at an angle relative to the packaging base to form the optical axis action component. These two elastic attitude adjustment beams are also arranged tangentially along the optical component's support ring. The lateral and vertical attitude adjustment units are positioned at different heights on the optical component's support ring. Pitch and yaw corrections are generated by writing differential residual stress between attitude adjustment units in the same direction.

[0040] The corresponding co-located dual-state attitude adjustment frames of the first and second micromirror arrays are each equipped with eight attitude adjustment units, which are distributed along the four sides and four corners of the rectangular optical component carrier frame. Attitude adjustment units located on opposite sides form a translation action combination, attitude adjustment units located at diagonal positions form a tilt action combination, and attitude adjustment units located on adjacent sides form a planar shear action combination.

[0041] In-situ dual-state pose adjustment frame structure parameters: The co-located dual-state attitude adjustment frame is integrally machined from 4J29 low-expansion alloy. The main body thickness of the attitude adjustment frame at the input and output ends is set to 0.6 mm, and the main body thickness of the attitude adjustment frame for the first and second micromirror arrays is set to 0.4 mm. A thermal isolation gap with a width of 0.30 mm is set between the base connecting ring and the optical component bearing ring, and the elastic attitude adjustment beam spans the thermal isolation gap.

[0042] The elastic attitude adjustment beams of the input and output attitude adjustment frames are set to a length of 5.0 mm, a width at the narrowest point of 0.50 mm, a width of 1.50 mm for the outer fixed section, and a width of 1.20 mm for the inner force transmission section. Thermal isolation grooves with a width of 0.20 mm are provided on both sides of the elastic attitude adjustment beams, extending from the inner side of the outer fixed section to the outer side of the inner force transmission section. The elastic attitude adjustment beams of the first and second micromirror array attitude adjustment frames are set to a length of 3.5 mm and a width at the narrowest point of 0.35 mm.

[0043] The attitude adjustment frame is formed by wire cutting, including the shape, thermal isolation gap, and thermal isolation groove. The base connection surface and the optical component mounting surface are then precision ground. After machining, the overall flatness, the width of the elastic attitude adjustment beam, the width of the thermal isolation groove, and the position of the mounting surface are inspected. The inspection results are stored in association with the attitude adjustment frame's serial number. The maximum equivalent stress of the elastic attitude adjustment beam under normal working load is limited to below 30% of the material's yield strength, and the local residual stress after laser adjustment is limited to below 70% of the material's yield strength.

[0044] Resistance heating layer and temperature detection circuit: An insulating layer, a platinum serpentine resistor layer, and a protective layer are sequentially formed on the surface of the elastic attitude adjustment beam. A constant current drive board is connected to both ends of the platinum serpentine resistor layer. The cold resistance of a single heating layer in the input and output attitude adjustment frames is controlled within the range of 150Ω to 250Ω, while the cold resistance of a single heating layer in the micromirror array attitude adjustment frame is controlled within the range of 100Ω to 200Ω.

[0045] The attitude adjustment frame was placed on a temperature control platform, and the resistance values ​​of each resistive heating layer were measured at 20℃, 30℃, 40℃, and 50℃. Linear fitting was performed on each measurement point to obtain the reference resistance and resistance temperature coefficient of each attitude adjustment unit. During formal identification, the constant current drive board collected the current and terminal voltage in real time and calculated the resistance value. Then, the temperature of the attitude adjustment beam was obtained based on the reference resistance and resistance temperature coefficient.

[0046] For each attitude adjustment unit, a first identification temperature rise and a second identification temperature rise are set. In this embodiment, the first identification temperature rise is set to 5°C, and the second identification temperature rise is set to 10°C. Each heating process includes six stages: reference sampling, temperature rise, stabilization and holding, optical path sampling, stopping heating, and cooling and retesting. The stabilization and holding condition is set to the attitude adjustment beam temperature change rate being lower than a predetermined threshold during continuous sampling time, and the cooling and retesting condition is set to the temperature difference between the attitude adjustment beam and the reference temperature being lower than 0.5°C.

[0047] Bidirectional laser stress writing structure: Each elastic attitude adjustment beam has a first laser stress writing area and a second laser stress writing area set on both sides of its width direction. The two laser stress writing areas are each set with twelve discrete laser writing positions along the length direction of the elastic attitude adjustment beam. The center distance between adjacent laser writing positions is set to 180μm, and an unwritten isolation segment is reserved between adjacent laser writing positions.

[0048] Three non-collinear machine vision reference marks are set on the surface of the base connecting ring. The controller establishes the current attitude adjustment frame coordinate system based on the three machine vision reference marks. The coordinates of each laser writing position are jointly determined by the attitude adjustment frame serial number, attitude adjustment unit number, first laser stress writing area or second laser stress writing area number, and position number along the length direction of the elastic attitude adjustment beam.

[0049] This embodiment uses a 1064nm pulsed fiber laser with a spot diameter of 60μm. Before formal packaging, laser energy calibration is performed using elastic beam test pieces of the same batch of 4J29 alloy, with the same thickness and surface treatment. The energy ranges for no permanent displacement, stable permanent displacement, and surface damage are measured respectively. The formal laser writing energy is limited to the stable permanent displacement energy range. Each laser writing position has a single-use energy limit and a cumulative energy limit. Laser writing positions that reach the cumulative energy limit are no longer used.

[0050] Multi-frequency calibration optical data acquisition: The 64 input ports were divided into 8 input port groups, each containing 8 input ports. Each of the 8 input ports within the same group was configured with a different modulation frequency, ranging from 5 kHz to 40 kHz, with the interval between adjacent modulation frequencies greater than the synchronous demodulation frequency resolution. The modulation depth of the input calibration light was set within the range of 2% to 8%, and the average calibration light power of each input port was simultaneously measured using a monitoring photodetector at the input end.

[0051] No. Each input port in time The instantaneous calibration optical power is expressed by the following formula: ; In the formula, Indicates the first Each input port in time Instantaneous calibration optical power at that time; This indicates the port number of the current input port; This indicates the time calculated from the start of the current sampling period; Indicates the first The average optical power of the calibration light at each input port; Indicates the first Each input port calibrates the modulation depth of the light; Indicates assignment to the first The modulation frequency of each input port; Indicates the first The initial phase of the modulated signal at each input port; Represents pi; The symbol represents the cosine function; the square brackets indicate that the terms within the function are multiplied as a whole.

[0052] In the In the first encoding exchange state, the first... The detection signal of each output port is represented by the following formula: ; In the formula, Indicates the first The first encoding exchange state Each output port at time The detection signal at that time; Indicates the sequence number of the encoding exchange state; Indicates the output port number; This indicates the time calculated from the start of the current sampling period; This indicates the number of input ports participating in the measurement within the current input port group; Indicates the input port number in the summation operation; Indicates the first The first encoding exchange state The input port to the first Optical power transmission coefficient of each output port; Indicates the first Each input port in time Instantaneous calibration optical power at that time; Indicates the first The first encoding exchange state Each output detection channel in time Detection noise and non-calibration light interference during operation; Indicates from the first input port to the... The summation is performed on each input port.

[0053] The controller performs synchronous demodulation according to the modulation frequency of each input port, and the estimated value of the optical power transmission coefficient is obtained according to the following formula: ; In the formula, This indicates the first time obtained through synchronous demodulation. The first encoding exchange state The input port to the first Estimated optical power transmission coefficient for each output port; symbol " "" indicates that the value was estimated from the sampled data; Indicates the output port number; Indicates the input port number; Indicates the sequence number of the encoding exchange state; Indicates the first Each input port calibrates the modulation depth of the light; Indicates the first The average optical power of the calibration light at each input port; This indicates the measurement time length corresponding to one synchronous demodulation. Indicates the start time of the current synchronous demodulation time interval; Indicates the end time of the current synchronization and demodulation time interval; Indicates the first The first encoding exchange state Each output port at time The detection signal at that time; Indicates the first The modulation frequency of each input port; Indicates the first The initial phase of the modulated signal at each input port; Represents pi; Represents the cosine function; Indicates the starting time Until the end time Integrate within the time frame; Indicates that the integration variable is time. Coefficient 2 is used to convert the cosine correlation component obtained by synchronous demodulation into the corresponding modulation amplitude.

[0054] This implementation sets up sixteen coding exchange states. Each input port establishes a main connection calibration path with at least one central region output port, one edge region output port, and one corner region output port. At the same time, it collects the modulation frequency components formed by the non-main connection input ports at each output port and converts them into non-main connection path transmission coefficients.

[0055] Construction of multi-port optical path error vector: For the first A calibration path is used to convert the measured optical power transmission coefficient and the target optical power transmission coefficient into a weighted logarithmic error: ; In the formula, Indicates the first Weighted logarithmic error of each calibration path; This indicates the path number determined according to the preset calibration path sequence; Indicates the first The weights corresponding to each calibration path; Represent the natural logarithm function; Indicates the first Measured optical power transmission coefficient of the calibration path; symbol " This indicates that the transmission coefficient is obtained by synchronous demodulation of the output time-domain optical power signal; This represents a positive stable term that prevents logarithmic operations from failing when the measured or target optical power transmission coefficients approach zero. Indicates the first The target optical power transmission coefficient of the calibration path; superscript " "" indicates that the corresponding value is a predetermined target value; the fractional term represents the normalized ratio between the measured transmission coefficient and the target transmission coefficient.

[0056] The weighted logarithmic errors of all calibration paths are arranged in a fixed path order to form a multiport optical path error vector: ; In the formula, This represents the current multi-port optical path error vector; bold text indicates that the corresponding variable is a vector. This represents the weighted logarithmic error of the first calibration path; This represents the weighted logarithmic error of the second calibration path; Indicates the first Weighted logarithmic error of each calibration path; Indicates the total number of paths participating in the current fast calibration; symbol " "Indicates from the 3rd path to the ..." The corresponding weighted logarithmic errors of the paths are arranged in the same order; the superscript " "" indicates matrix transpose, converting horizontally arranged data into a column vector; square brackets indicate combining the errors of each path into the same vector.

[0057] The target optical power transmission factor of the main connection path is determined by bare-core test data, optical design data, and product insertion loss requirements. The target optical power transmission factor of the non-main connection path is determined by conversion from the crosstalk limit. Edge paths, corner paths, and long paths across mirror arrays are assigned higher weights than short paths in the center, while non-main connection paths approaching the crosstalk limit have their corresponding weights increased.

[0058] Identification of the reversible thermoelasticity influence matrix: For the first Each attitude adjustment unit applies a first identification current and a second identification current to obtain multi-port optical path error vectors under two stable temperature rise states. The optical path influence vector per unit temperature rise of each attitude adjustment unit is determined according to the following formula: ; In the formula, Indicates the first The unit temperature rise optical path influence vector of each attitude adjustment unit; Indicates the orientation adjustment unit number; Indicates the first The multi-port optical path error vector measured when each attitude adjustment unit is in the second stable temperature rise state; superscript " "Indicates that the corresponding state is the higher of the two identified temperature rise states; Indicates the first The multi-port optical path error vector measured when each attitude adjustment unit is in the first stable temperature rise state; superscript " "Indicates that the corresponding state is the lower of the two identified temperature rise states; Indicates the first Temperature of the attitude adjustment beam when the attitude adjustment unit is in the second stable temperature rise state; Indicates the first The attitude adjustment unit is in the first stable temperature rise state when the attitude adjustment beam temperature is; the numerator represents the multi-port optical path error change vector between two stable temperature rise states; the denominator represents the temperature difference between two stable temperature rise states; the fractional result represents the multi-port optical path error change caused by a unit temperature change.

[0059] The optical path influence vectors of all attitude adjustment units per unit temperature rise are arranged according to the attitude adjustment unit number to form a reversible thermoelastic influence matrix: ; In the formula, This represents the reversible thermoelastic effect matrix; This represents the optical path influence vector per unit temperature rise of the first attitude adjustment unit; This represents the optical path influence vector per unit temperature rise of the second attitude adjustment unit; Indicates the first The unit temperature rise optical path influence vector of each attitude adjustment unit; Indicates the total number of pose adjustment units participating in the current identification; symbol " "Indicates the 3rd attitude adjustment unit to the 1st attitude adjustment unit" The unit temperature rise optical path influence vectors of each attitude adjustment unit are arranged in numerical order; square brackets indicate that the unit temperature rise optical path influence vectors are combined as matrix columns.

[0060] Each attitude adjustment unit performs two repeated identifications, calculating the directional angle and amplitude difference between the two unit temperature rise optical path influence vectors. If the two identification results exceed the preset repeatability range, the corresponding attitude adjustment unit is marked as an unstable attitude adjustment unit and will not enter the initialization process of the permanent adjustment influence matrix. Heating is stopped after each identification, and the attitude adjustment beam is allowed to return to the reference temperature before confirming that the multi-port optical path error vector has returned to the measurement range before heating.

[0061] Controllable optical path error mode extraction: Singular value decomposition of the reversible thermoelastic influence matrix: ; In the formula, This represents the reversible thermoelastic effect matrix; The left singular vector matrix represents the optical path error space, and the left singular vectors describe the orthogonal transformation directions of the multi-port optical path error. Represents a singular value diagonal matrix, where the diagonal elements are arranged in descending order; Greek letters Represents a diagonal matrix composed of sets of singular values; This represents the right singular vector matrix in the action space of the attitude adjustment unit, where each right singular vector describes an orthogonal cooperative action combination of multiple attitude adjustment units; the superscript " "" indicates the transpose of the right singular vector matrix; matrix product indicates that the reversible thermoelastic influence matrix is ​​composed of the optical path error direction, singular value amplitude, and attitude adjustment unit combination.

[0062] The optical path error modes to be permanently adjusted are selected according to the following formula: ; In the formula, Represents the singular value diagonal matrix with the first... One singular value; Indicates the singular value index; This indicates the threshold for retaining singular values. Represents the largest singular value in a singular value diagonal matrix; the symbol " " indicates the first Each singular value is not less than the product of the singular value retention ratio threshold and the largest singular value; the optical path error mode corresponding to the singular value satisfying this inequality is included in the permanent adjustment calculation. In this embodiment... Set to 0.05.

[0063] The current multi-port optical path error vector is projected onto the retained left singular vector to obtain the amplitude of the current error in each controllable optical path error mode. Based on the components of each attitude adjustment unit in the right singular vector, the attitude adjustment unit that has a large contribution to the main error mode and whose repeated identification results are stable is selected for permanent response calibration.

[0064] Establishment of Permanent Response Calibration and Permanent Adjustment Influence Matrix Low-energy calibration laser writing is performed on the unused laser writing position of the selected attitude adjustment unit. After the attitude adjustment frame returns to the reference temperature, the multi-port optical path error vector after calibration laser writing is acquired. The permanent optical path influence vector per unit laser energy is calculated according to the following formula: ; In the formula, Indicates the first The permanent optical path influence vector per unit laser energy corresponding to each laser writing direction; Indicates the laser writing direction number that participated in the permanent response calibration; Indicates the first The multi-port optical path error vector measured after calibration laser writing is completed in each laser writing direction and the temperature is restored to the reference temperature; subscript " "This indicates that the corresponding data was obtained after the calibration laser was written; Indicates the first The multi-port optical path error vector measured before laser writing is used to calibrate the laser writing direction; subscript " "This indicates that the corresponding data was obtained before the calibration laser was written; Indicates the first The effective calibration laser energy corresponds to each laser writing direction; the numerator represents the vector of permanent multi-port optical path error change caused by calibration laser writing; the fractional result represents the permanent optical path change corresponding to a unit laser energy.

[0065] Arrange all permanent optical path influence vectors per unit laser energy to form a permanent adjustment influence matrix: ; In the formula, This represents the permanent adjustment effect matrix; This represents the permanent optical path influence vector per unit laser energy in the first laser writing direction; This represents the permanent optical path influence vector per unit laser energy in the second laser writing direction; Indicates the first Permanent optical path influence vector per unit laser energy in each laser writing direction; Indicates the total number of laser writing directions involved in permanent tuning; symbol " "Indicates the direction of the third laser writing to the first The permanent optical path influence vectors per unit laser energy in each laser writing direction are arranged in numerical order; square brackets indicate that the permanent optical path influence vectors per unit laser energy are combined as matrix columns.

[0066] For attitude adjustment units that have not yet undergone calibration laser writing, an initial permanent optical path influence vector is established based on the ratio between the permanent response and the reversible thermoelastic response of attitude adjustment units with the same structure. After the corresponding attitude adjustment unit performs its first formal laser writing, its permanent optical path influence vector is updated based on the actual optical path changes.

[0067] Solving for the laser writing vector: The first and second laser stress writing regions of each attitude adjustment unit are set as equivalent laser writing variables with opposite directions. The optimal laser writing vector for the current adjustment cycle is solved according to the following formula: ; In the formula, This represents the optimal equivalent laser writing vector obtained by the current tuning loop; the superscript " " indicates that the corresponding vector is the target solution obtained from the optimization calculation; This represents a candidate write vector composed of the laser energies to be determined corresponding to each laser write direction; This means selecting the write vector that minimizes the objective function within the curly braces from all candidate write vectors that satisfy the constraints. This represents the multi-port optical path error weighting matrix; This represents the current multi-port optical path error vector; This represents the permanent adjustment effect matrix; This represents the predicted multiport optical path error change vector corresponding to the candidate laser writing vector; This represents the predicted remaining optical path error vector after performing candidate laser writing; This represents the 2-norm of a vector; the square of the upper right corner of the 2-norm indicates squaring the 2-norm result. This represents the total laser writing energy confinement coefficient; This represents the laser write cost matrix set based on the remaining number of laser write bits, cumulative laser energy, and permanent response stability. This indicates that adjacent attitude adjustment units write the difference constraint coefficient; This represents a difference matrix that describes the spatial relationship between adjacent attitude control units or adjacent action regions. This represents the laser writing difference vector between adjacent attitude adjustment units; the first term in the curly braces represents the cost of predicting the remaining multi-port optical path error, the second term represents the laser writing energy cost, and the third term represents the spatial non-uniform writing cost.

[0068] Each equivalent laser write variable satisfies the following single write range constraint: ; In the formula, Indicates the first The equivalent laser write amount of each attitude adjustment unit in the current adjustment cycle; Indicates the orientation adjustment unit number; Indicates the first The maximum laser writing amount allowed in the current adjustment cycle for the first laser stress writing area of ​​each attitude adjustment unit; superscript " "Indicates that the corresponding constraint belongs to the positive direction of the first laser stress writing region; Indicates the first The second laser stress writing area of ​​each attitude adjustment unit is the maximum laser writing amount allowed in the current adjustment cycle; superscript " "" indicates that the corresponding constraint belongs to the reverse direction of the second laser stress writing region; the symbol " "This indicates that the equivalent laser writing amount must not be lower than the upper limit in the negative direction and must not be higher than the upper limit in the positive direction; when..." When the value is positive, the first laser stress writing region is selected. When the value is negative, the second laser stress writing region is selected. When the value is zero, the current attitude adjustment unit does not perform laser writing.

[0069] The optimization calculation simultaneously performs cumulative energy constraints, mutual exclusion constraints for writing on both sides of the same elastic attitude adjustment beam, spacing constraints between adjacent laser writing positions, local temperature constraints, residual stress constraints of the elastic attitude adjustment beam, and reserved constraints for final-state trimming writing positions. The actual execution amount in the initial trimming stage is set to 50% of the optimal laser writing vector, and the actual execution amount in the fine trimming stage is set to 20% of the optimal laser writing vector.

[0070] Permanent adjustment influence matrix online update: No. After the adjustment cycle is completed and the temperature returns to the reference temperature, the permanent adjustment effect matrix is ​​updated according to the actual changes in the multi-port optical path: ; In the formula, Indicates the first The permanent adjustment effect matrix obtained after the completion of the next adjustment cycle; subscript " " indicates that the corresponding matrix is ​​used in the next adjustment cycle; Indicates the first The permanent adjustment effect matrix used before the next adjustment cycle is executed; Indicates the permanent adjustment cycle number; Indicates the first The change in the measured multi-port optical path error vector before and after the laser writing; Greek letters This represents the difference between the previous and current states; Indicates the first The laser writing vector actually executed in the next adjustment cycle; Indicates the first The optical path error change vector is obtained by predicting the permanent adjustment influence matrix before the next adjustment cycle is executed. This represents the prediction residual vector between the actual optical path change and the predicted optical path change. Indicates the first The transpose of the actual laser writing vector; Indicates the first Sum of squares of each element in the actual laser writing vector; The positive stable term is used to prevent the denominator from being zero; the fractional term represents the incremental correction of the permanent adjustment influence matrix along the actual laser writing direction.

[0071] When the directional angle between the actual multi-port optical path change and the predicted multi-port optical path change exceeds a preset threshold, the above matrix update is not performed, the corresponding attitude adjustment unit is paused, and the laser focus height, machine vision coordinates, attitude adjustment beam temperature, write bit surface state, and optical power repeatability are checked. After eliminating measurement anomalies, low-energy recalibration is performed on adjacent unused laser write bits.

[0072] Data source and usage process: The design prior data originates from the OCS optical design file, the input fiber collimator array coordinate file, the output fiber collimator array coordinate file, the first micromirror array coordinate file, the second micromirror array coordinate file, and the co-located dual-state attitude adjustment frame structure design file. This prior data is used to establish the target optical power transmission coefficient, encoding exchange state, calibration path set, initial action direction of the attitude adjustment unit, upper limit of heating temperature rise, and laser writing safety window; it does not directly replace the physical identification results of individual devices.

[0073] Incoming material inspection data comes from attitude adjustment frame size inspection equipment, flatness inspection equipment, resistance inspection equipment, insulation resistance inspection equipment, collimator array end face inspection equipment, and micromirror array flatness inspection equipment. This data is used to determine whether optical components and attitude adjustment frames have entered the self-calibration station, and to determine the allowable identification current, cumulative laser writing energy, and remaining number of write bits for a single attitude adjustment unit.

[0074] Real-time optical data originates from the input-side monitoring photodetector and the output-side photodetector array. This data includes the average power of the input calibration light, the output time-domain optical power signal, the transmission coefficients of the main connected path after synchronous demodulation, the transmission coefficients of the non-main connected paths, the repeatability of measurements, path insertion loss, and crosstalk at non-target ports. This real-time optical data is used to construct the multi-port optical path error vector, identify the reversible thermoelastic effect matrix, calibrate the permanent adjustment effect matrix, and verify the results of each laser writing operation.

[0075] Thermoelectric data is derived from the constant current drive board and temperature acquisition board, including heating current, heating end voltage, real-time resistance of the heating layer, real-time temperature of the attitude adjustment beam, heating time, stabilization time, and cooling time. This thermoelectric data is used to calculate the temperature difference between two identified temperature rise states and to determine whether the co-position dual-state attitude adjustment frame has returned to the reference temperature after laser writing.

[0076] The laser process data originates from the laser controller and machine vision positioning system, and includes laser number, laser wavelength, pulse width, single pulse energy, number of pulses, spot diameter, focal height, attitude adjustment frame coordinates, attitude adjustment unit number, laser stress writing area number, and laser writing bit number. This laser process data is used to form the actual laser writing vector for each instance and to cumulatively record the laser energy received at each laser writing bit.

[0077] Environmental validation data is derived from temperature control testing equipment, package cover connection equipment, and optical path measurement system. This includes the optical path transmission matrix at low temperature stable state, room temperature stable state, high temperature stable state, after temperature recovery, and before and after package cover connection. The environmental validation data is used to distinguish between temperature-dependent optical path errors that change with temperature and disappear after recovery to room temperature, and permanent package optical path errors that remain after recovery to room temperature.

[0078] Complete application packaging process: Sp1. Process the input end co-position dual-state attitude adjustment frame, the output end co-position dual-state attitude adjustment frame, the first micro-mirror array co-position dual-state attitude adjustment frame and the second micro-mirror array co-position dual-state attitude adjustment frame according to the OCS optical component size, and process the elastic beam test piece using the same batch of materials. Obtain the temperature rise window of the resistance heating layer, the laser stable permanent displacement energy range and the surface damage energy range through the test piece.

[0079] Sp2. Inspect the external dimensions, flatness, elastic attitude adjustment beam width, thermal isolation groove width, cold resistance of heating layer and insulation resistance of each co-position dual-state attitude adjustment frame. Associate the test data with the attitude adjustment frame serial number and stop the subsequent assembly of attitude adjustment frames that exceed the specified test range.

[0080] Sp3. Fix the input fiber collimator array, the first micromirror array, the second micromirror array, and the output fiber collimator array to the corresponding optical component carrier rings respectively. Perform coarse positioning through an external six-degree-of-freedom attitude adjustment mechanism to ensure that the key calibration path has stable output optical power. Then fix each base connection ring to the packaging base in a centrally symmetrical order.

[0081] Sp4. Gradually reduce the clamping stiffness of the external six-degree-of-freedom attitude adjustment mechanism so that the main support load of each optical component is borne by the co-position dual-state attitude adjustment frame. Inject multi-frequency calibration light into the input port group, control the two-stage micro-mirror array to enter the encoding exchange state, collect the initial optical path transmission coefficient and construct the initial multi-port optical path error vector.

[0082] Sp5. The first and second temperature rise identifications are performed on each attitude adjustment unit in sequence. The multi-port optical path error vectors under the two temperature rise states and the cooling recovery state are collected to establish a reversible thermoelastic influence matrix. Unstable attitude adjustment units are eliminated by repeating the identification results.

[0083] Sp6. Perform singular value decomposition on the reversible thermoelastic influence matrix to extract the controllable optical path error mode. Project the current multi-port optical path error onto the controllable optical path error mode. Select the attitude adjustment unit and the unused laser writing bit for low-energy permanent response calibration based on the projection amplitude and the right singular vector component, and establish the initial permanent adjustment influence matrix.

[0084] Sp7. Solve the current laser writing vector based on the current multi-port optical path error vector, permanent adjustment influence matrix and laser writing constraint, perform partial proportional laser writing, and re-acquire the multi-port optical path error vector after the attitude adjustment frame recovers to the reference temperature, and update the permanent adjustment influence matrix according to the actual optical path changes.

[0085] Sp8 repeatedly executes laser writing vector solution, laser stress writing, reference temperature recovery, multi-port optical path retesting, and permanent adjustment influence matrix update. In the initial stage, priority is given to reducing the overall translation and tilt errors of the input and output ends. In the fine stage, the weights of edge paths, corner paths, cross-mirror array long paths, and non-main connection paths are increased.

[0086] Sp9. When the multi-port optical path error enters the initial acceptance range, the holding load of the external six-degree-of-freedom attitude adjustment mechanism is divided into multiple levels and released step by step. After each level of holding load is released, wait for the structure to stabilize, re-acquire the multi-port optical path error vector, and use the remaining laser to write the bit to correct the permanent optical path error retained after the load is released.

[0087] After the Sp10 and external six-degree-of-freedom attitude adjustment mechanism are completely released, the OCS is placed in low temperature, room temperature and high temperature environments to complete the heating and cooling cycle. The optical path transmission matrix of each stable temperature state is recorded. The optical path error that is still retained after returning to room temperature is included in the permanent laser adjustment. The optical path error that changes with temperature and disappears after returning to room temperature is written into the micro-mirror drive temperature compensation data.

[0088] Sp11. Connect the encapsulation cover plates in segments according to the central symmetry sequence. After each set of encapsulation cover plate connection areas is completed, remeasure the critical calibration path, match the permanent optical path error retained after the encapsulation cover plates are connected with the permanent adjustment influence matrix, and correct it using the final state adjustment reserved laser writing bit.

[0089] Sp12. Complete the optical path transmission matrix measurement between all input ports and all output ports, obtain the insertion loss and corresponding non-target port crosstalk of each target connection state, and write the initial optical path data, reversible thermoelastic effect matrix, permanent adjustment effect matrix, laser writing record, fixture load release data, temperature cycle data, package cover connection data and the final full-port optical path transmission matrix into the electronic package calibration file.

[0090] Exception handling and process safety: When the input optical power fluctuation exceeds the specified range, the current sampled data is not included in the multi-port optical path error vector; when the baseline change of the output photoelectric detection channel exceeds the specified range, the output detection channel gain calibration is re-executed; when the multi-port optical path error does not recover to the measurement range before heating after the attitude adjustment unit cools down, the corresponding attitude adjustment unit exits the permanent adjustment; when the effective rank of the permanent adjustment influence matrix is ​​lower than the number of current main optical path error modes, the coarse positioning process is returned; when the remaining laser writing bits of a certain optical component are lower than the final state adjustment reserved number, the current stage of writing is stopped.

[0091] The laser is hardware-interlocked with the machine vision positioning system, temperature acquisition system, packaging base positioning detection system, and packaging cover position detection system. The laser remains in a disabled triggering state if the machine vision has not confirmed the laser writing coordinates, the attitude adjustment beam temperature exceeds the upper limit of the writing temperature, the packaging base has not completed positioning, or the packaging cover blocks the laser path. After each laser writing operation, a fixed cooling and waiting process is performed. The permanent adjustment influence matrix is ​​not updated using the current optical path data until the attitude adjustment beam temperature returns to the reference temperature range.

[0092] Electronic Packaging Calibration Files and Service Applications: The electronic packaging calibration file is established according to the OCS device serial number. The file records the input fiber collimator array number, output fiber collimator array number, first micromirror array number, second micromirror array number, each co-position dual-state attitude adjustment frame number, each attitude adjustment unit number, and each laser writing position number. It also records the timestamp and station number corresponding to each thermoelastic identification, permanent response calibration, formal laser writing, fixture load release, temperature cycling, and package cover connection.

[0093] During the OCS service phase, laser writing is no longer performed on the in-situ dual-state attitude adjustment frame. The device controller adjusts the micromirror drive bias based on temperature-related optical path error data in the electronic packaging calibration file, and periodically selects critical paths covering the main spatial regions of the input, output, and two micromirror arrays for status detection. When the current optical path change direction matches the historical influence direction of the in-situ dual-state attitude adjustment frame at the input end, the abnormal position is recorded as the input end packaging area; when the current optical path change direction matches the historical influence direction of the in-situ dual-state attitude adjustment frame at the output end, the abnormal position is recorded as the output end packaging area; when the current optical path change direction matches the historical influence direction of the first or second micromirror array, the abnormal position is recorded as the corresponding micromirror array area.

[0094] 128-port extension implementation method: For an OCS with 128 input ports and 128 output ports, the input ports are divided into 16 input port groups, each with 8 distinct modulation frequencies. Multiple coded switching states are used to ensure that each input port participates in at least the center path, edge path, corner path, and long cross-region path. Sixteen attitude adjustment units are set in the co-located dual-state attitude adjustment frames at the input and output ends, and twelve attitude adjustment units are set in the two groups of micromirror array attitude adjustment frames. The additional attitude adjustment units are located at the middle of the long side of the optical component carrier ring and at the boundary of the mirror array carrier frame area.

[0095] The 128-port OCS does not measure all 16,384 input-output combinations during the rapid calibration phase. Instead, it progressively increases the calibration path based on the effective rank of the reversible thermoelastic influence matrix until adding a new path no longer increases the number of controllable optical path error modes. After completing rapid closed-loop tuning, the optical path transmission matrix corresponding to all input-output combinations is measured during the final inspection phase. This extended approach retains the core technologies of in-situ reversible thermoelastic identification, bidirectional permanent residual stress writing, online updating of the device-level permanent tuning influence matrix, and fixture-free final-state calibration. Specific Implementation Example 3: like Figures 1 to 14 As shown, based on the content of the above specific embodiments, the following content is further disclosed: The OCS multi-port optical path self-calibration packaging method of the present invention uses an input fiber collimator array, a first micromirror array, a second micromirror array, an output fiber collimator array, a packaging base, and a corresponding co-located dual-state attitude adjustment frame as the packaging object. It uses the multi-port optical power transmission state as the direct evaluation basis for the packaging attitude, the reversible thermoelastic deformation of the elastic attitude adjustment beam in the co-located dual-state attitude adjustment frame as the structural response identification means, and the residual stress formed by the laser stress writing area on the elastic attitude adjustment beam as the permanent adjustment means. This allows optical path measurement, structural response identification, permanent attitude correction, fixture unloading compensation, temperature cycle compensation, and packaging cover plate connection compensation to be continuously executed according to a unified data link.

[0097] I. Preparation of encapsulated objects and establishment of unified coordinates: Before self-calibration packaging, incoming material inspections were performed on the input-side in-situ dual-state attitude adjustment frame, the first micromirror array in-situ dual-state attitude adjustment frame, the second micromirror array in-situ dual-state attitude adjustment frame, and the output-side in-situ dual-state attitude adjustment frame. The inspections included the flatness of the base connecting ring, the position of the positioning holes, the dimensions of the symmetrical welding area, the roundness of the optical component bearing ring, the width of the optical component mounting slot, the height of the end-face limiting stage, the width of the elastic attitude adjustment beam, the width of the thermal isolation groove, the resistance of the resistance heating layer, and the continuity of the temperature detection circuit. The inspection results were recorded in the process record associated with the corresponding attitude adjustment frame serial number. Attitude adjustment frames whose inspection results exceeded the design tolerances were not included in the subsequent assembly process.

[0098] The packaging base is fixed to the device clamping stage by three non-collinear mechanical reference points. The machine vision positioning device identifies the reference marks on the packaging base and establishes the packaging base coordinate system. A longitudinal coordinate axis is established along the initial propagation direction of the input beam, a transverse coordinate axis is established along the port arrangement direction of the input fiber collimator array, and a vertical coordinate axis is established along the normal of the packaging base. The coordinates of the positioning holes, symmetrical welding areas, elastic attitude adjustment beams, first laser stress writing areas, second laser stress writing areas, and discrete laser writing positions of each co-located dual-state attitude adjustment frame are all transformed to the packaging base coordinate system, so that the machine vision positioning device, pulsed laser adjustment device, and external attitude adjustment mechanism use the same coordinate reference.

[0099] II. Optical component installation and coarse positioning: The input fiber collimator array is installed in the optical component mounting slot of the input end co-position dual-state attitude adjustment frame, with its end face abutting against the end face limiting stage; the first micromirror array and the second micromirror array are respectively installed in their corresponding co-position dual-state attitude adjustment frames; the output fiber collimator array is installed in the output end co-position dual-state attitude adjustment frame. Each optical component is connected to the optical component carrier ring using a locally fixed structure, with the connection points avoiding the elastic attitude adjustment beam, thermal isolation groove, resistance heating layer, temperature detection circuitry, and laser stress writing area.

[0100] External attitude adjustment mechanisms are connected to the carrier rings of each optical component and apply lateral, vertical, axial, pitch, and yaw displacements to the input fiber collimator array, the first micromirror array, the second micromirror array, and the output fiber collimator array. During the coarse positioning stage, the center input port, edge input port, and corner input port are sequentially activated, and the two sets of micromirror arrays are controlled to establish corresponding output paths. When continuous and stable output optical power is detected on the calibration paths in the center, edge, and corner regions, the current position of each external attitude adjustment mechanism is recorded.

[0101] After coarse positioning, the base connecting rings of each co-positioned dual-state attitude adjustment frame are installed onto the packaging base through the positioning surface and positioning holes, and the symmetrical welding areas are fixed in a diagonally alternating sequence. During the fixing process, the external attitude adjustment mechanism is kept in a low-stiffness holding state to prevent welding heat input and local shrinkage from directly causing excessive abrupt changes in the optical component's pose. After the base connecting rings are fixed, the external attitude adjustment mechanism retains only the holding load used to limit large-range displacement, and the main support load of the optical component is transferred to the corresponding co-positioned dual-state attitude adjustment frame.

[0102] III. Multi-frequency calibration optical injection and code switching status settings: The multi-channel calibration light source is connected to multiple input ports, and each input port participating in the same measurement cycle is assigned an independent modulation frequency. The modulation frequency avoids the mains frequency, the micromirror drive frequency, the constant current drive board switching frequency, and their main harmonics. For a 64-input port OCS, all input ports are divided into eight groups, each group containing eight input ports, and each group uses eight different modulation frequencies.

[0103] The average calibration power of each input port is collected by the input monitoring channel before measurement. If the input power fluctuation exceeds the set range, the data will not be included in the optical path error calculation. The first and second micromirror arrays operate sequentially according to the pre-established coded exchange state table. Each coded exchange state includes at least the central region path, the edge region path, the corner region path, and the long propagation path that crosses different micromirror regions.

[0104] The output photoelectric detection array synchronously acquires the time-domain optical power signals of all output ports. The industrial control computer performs synchronous demodulation according to the modulation frequency corresponding to each input port, obtaining the main connection path transmission coefficient formed by the target input port at the target output port, and the non-main connection path transmission coefficients formed by the other input ports at the output port. The main connection path transmission coefficient is used to calculate the current optical path coupling state, and the non-main connection path transmission coefficients are used to calculate the current crosstalk state.

[0105] Three sets of data are continuously acquired for each coding exchange state, with each set including a complete modulation period. When the dispersion of the three sets of data is lower than the measurement stability threshold, the average of the three sets of data is used as the current transmission coefficient; when the dispersion is higher than the measurement stability threshold, the input optical power, micromirror drive status, and output detection channel gain are rechecked.

[0106] IV. Construction of Multi-Port Optical Path Error Vector: The industrial control computer reads the target optical power transmission coefficient corresponding to the current device. The target optical power transmission coefficient is determined by optical design data, input fiber collimator array bare component test data, first micromirror array bare component test data, second micromirror array bare component test data, and output fiber collimator array bare component test data.

[0107] For the main connection path, the measured transmission coefficient and the target transmission coefficient are logarithmically normalized to form the main connection path error. For non-main connection paths, the measured transmission coefficient is compared with the transmission coefficient corresponding to the specified crosstalk upper limit to form the non-main connection path error. Corresponding weights are assigned to the center path, edge path, corner path, and long cross-regional path, with higher weights assigned to edge paths, corner paths, and paths with higher current losses.

[0108] All path errors are arranged in a fixed order to form a multi-port optical path error vector. This fixed order is associated with the input port number, output port number, encoding exchange status number, and path type number, ensuring that the multi-port optical path error vectors obtained in different measurement cycles have the same dimension and the same element meaning. After the multi-port optical path error vector is established, the average path error, maximum path error, dispersion of the main connecting path, and maximum value of the non-main connecting path are calculated simultaneously and used as state variables in subsequent adjustment stages.

[0109] V. Identification of Reversible Thermoelastic Response: Before the formal laser stress writing, the industrial control computer sequentially drives the resistive heating layers in each co-located dual-state attitude adjustment frame. The identification process for each attitude adjustment unit includes baseline state measurement, first temperature rise state measurement, second temperature rise state measurement, and cooling recovery measurement.

[0110] The baseline state measurement is performed when the temperature difference between the elastic attitude adjustment beam and the packaging base is below a specified value. In the first temperature rise state, a first identification current is applied to the corresponding resistive heating layer via a constant current drive board, causing the elastic attitude adjustment beam to reach a first stable temperature rise. In the second temperature rise state, a second identification current is applied, causing the elastic attitude adjustment beam to reach a second stable temperature rise. In each temperature rise state, the coded switching state measurement is re-executed, and the corresponding multi-port optical path error vector is obtained.

[0111] Based on the relationship between the first temperature rise state, the second temperature rise state, and the corresponding multi-port optical path error vector, the unit temperature rise optical path influence vector of the attitude adjustment unit is determined. The unit temperature rise optical path influence vectors of all attitude adjustment units are arranged according to the attitude adjustment unit number to form a reversible thermoelastic influence matrix.

[0112] After the second temperature rise measurement is completed, heating is stopped, and the elastic attitude adjustment beam is allowed to recover to the reference temperature. When the difference between the multi-port optical path error vector after cooling and the reference error vector before heating is lower than the recovery threshold, the attitude adjustment unit is confirmed to be in the reversible thermoelastic working range; when the difference is higher than the recovery threshold, the attitude adjustment unit is stopped, and the optical component fixation status, plastic deformation of the attitude adjustment beam, and abnormality of the resistance heating layer are checked.

[0113] VI. Optical path error modes and attitude adjustment unit selection: Singular value decomposition is performed on the reversible thermoelastic influence matrix to convert multiple path errors into multiple optical path error modes that can be stably controlled by the attitude adjustment unit. Modes with singular values ​​below the noise threshold are not included in the permanent adjustment calculation. The retained optical path error modes correspond to the overall lateral offset, overall vertical offset, axial offset, pitch offset, yaw offset, and local area offset that can be actually changed by the current attitude adjustment frame.

[0114] The current multi-port optical path error vector is projected onto each optical path error mode to obtain the error amplitude of each mode. The mode with the largest error amplitude is determined as the current priority adjustment mode, and the attitude adjustment units participating in the adjustment of this mode are selected according to the contribution ratio of each attitude adjustment unit in the right singular vector.

[0115] When the current multi-port optical path error mainly falls within the uncontrollable space of the reversible thermoelastic influence matrix, permanent laser calibration is not performed. Instead, the process returns to the coarse positioning step, where the corresponding optical components are readjusted via an external attitude adjustment mechanism. Only when the current main error is within the controllable modal space of the attitude adjustment frame will the permanent response calibration step be initiated.

[0116] VII. Permanent Response Calibration and Write Direction Determination: A stable attitude adjustment unit is selected in each primary action direction, and an unused discrete laser writing bit is selected in either the first or second laser stress writing area of ​​that attitude adjustment unit. The machine vision positioning device identifies the machine vision reference mark and calculates the laser focus positioning coordinates based on the transformation relationship between the attitude adjustment frame coordinates and the pulsed laser trimming device coordinates.

[0117] The pulsed laser trimming device moves to the selected discrete laser writing position and performs low-energy calibration writing. After calibration writing, optical path measurement is stopped, and the elastic attitude adjustment beam and the packaging base are allowed to return to the same reference temperature before the multi-port optical path error vector is re-acquired. Based on the change in multi-port optical path error before and after calibration writing and the calibration writing energy, the permanent optical path influence vector corresponding to the laser stress writing direction is determined.

[0118] The first and second laser stress writing regions respectively create permanent pose effects in opposite directions. For the same elastic attitude adjustment beam, the controller selects one of the laser stress writing regions based on the current error direction, and does not simultaneously write to both laser stress writing regions of the same elastic attitude adjustment beam in the same adjustment cycle.

[0119] All effective permanent optical path influence vectors are arranged according to the laser stress writing area number to form a permanent adjustment influence matrix. For attitude adjustment units of the same structure that have not undergone independent calibration writing, the initial permanent optical path influence vector is established using the calibration coefficients of the attitude adjustment frames in the same batch and the reversible thermoelastic influence vector of the attitude adjustment unit, and subsequently updated based on the actual writing results.

[0120] VIII. Calculation of Laser Writing Position and Writing Energy: The industrial control computer calculates the write vector for the current adjustment cycle based on the current multi-port optical path error vector and the permanent adjustment influence matrix. During the calculation process, the single write energy, cumulative write energy, mutual exclusion conditions on both sides of the same elastic attitude adjustment beam, usage order of adjacent write bits, local temperature of the elastic attitude adjustment beam, and residual stress are all constrained.

[0121] In the coarse adjustment phase, priority is given to correcting the overall lateral offset, overall vertical offset, pitch offset, and yaw offset, and only a portion of the write energy is calculated per operation. In the fine adjustment phase, the weights of edge paths, corner paths, long cross-region paths, and non-primary connection paths are increased, and the write energy per operation is reduced.

[0122] Each discrete laser write bit is recorded sequentially from the first write bit to the twelfth write bit. Once the current write bit reaches its cumulative energy limit, the controller switches to the next unused write bit. An unwritten isolation segment is maintained between adjacent write bits to prevent overlapping of local heat-affected zones caused by continuous writing.

[0123] IX. Laser writing and permanent adjustment affect matrix updates: The pulsed laser trimming device performs the writing process according to the laser stress writing area, discrete laser writing position, focal coordinates, number of pulses, and single pulse energy determined in the current trimming cycle. During the writing process, the laser number, writing area number, writing position number, focal height, pulse energy, number of pulses, and machine vision positioning results are recorded in real time.

[0124] After laser writing is completed, wait for the temperature difference between the elastic attitude adjustment beam and the packaging base to return to the specified range before re-performing the multi-port optical path measurement. Based on the difference between the actual optical path error change and the predicted change in the permanent adjustment influence matrix, update the corresponding columns in the permanent adjustment influence matrix that participated in this writing.

[0125] When the actual optical path change direction is consistent with the predicted direction and the change amplitude is within the set range, the updated permanent adjustment influence matrix enters the next adjustment cycle; when the actual optical path change direction is opposite to the predicted direction, the corresponding writing area is stopped from continuing to write, and the laser focus offset, writing bit surface state, local damage of the elastic attitude adjustment beam, optical path measurement fluctuation and temperature recovery status are checked.

[0126] 10. Deactivation of external attitude adjustment mechanisms in stages: Once the multi-port optical path error enters the initial calibration range, the holding load applied to each optical component by the external attitude adjustment mechanism is gradually reduced according to multiple load levels. After each level of reduction in holding load, the elastic state of the co-position dual-state attitude adjustment frame is allowed to stabilize before the multi-port optical path measurement is re-executed.

[0127] The optical path error generated after load removal is reprojected onto the controllable optical path error mode and permanently adjusted using the remaining discrete laser-written bits. The graded release process is executed sequentially until the external attitude adjustment mechanism is completely detached from the optical component support ring.

[0128] After the external attitude adjustment mechanism is completely detached, three consecutive multi-port optical path measurements are performed. When the path error dispersion of the three measurements is below the specified range, and all critical calibration paths are within the acceptance range, the temperature cycling step begins.

[0129] XI. Temperature Cycling and Temperature Error Separation: The OCS, having completed fixture-free calibration, was placed in a temperature-controlled environment, and the following processes were sequentially executed: low-temperature stabilization, room-temperature stabilization, high-temperature stabilization, and room-temperature recovery. For each temperature stabilization state, the complete calibration path transfer matrix was acquired, and the package base temperature, the temperature of each attitude adjustment frame, and the micromirror drive bias were recorded.

[0130] Optical path errors that change with temperature and disappear after returning to room temperature are recorded as temperature-dependent optical path errors; optical path errors that persist after temperature cycling are recorded as permanent encapsulation optical path errors. Temperature-dependent optical path errors are converted into temperature-driven compensation data for the first and second micromirror arrays, while permanent encapsulation optical path errors are corrected using the remaining discrete laser-written bits.

[0131] After temperature cycling, laser writing uses a lower single-write energy than the initial trimming stage, and prioritizes the use of reserved discrete laser write bits to avoid consuming all final state compensation space.

[0132] 12. Encapsulation cover connection and final adjustment: Before installing the encapsulation cover, multi-port optical path reference data is collected when the cover is not connected. The encapsulation cover is connected in segments according to a centrally symmetrical sequence. After each set of symmetrical connection areas is completed, the temperature of the encapsulation base is allowed to recover and key calibration path data is collected.

[0133] The optical path changes caused by the cover plate connection are matched according to the historical influence vectors corresponding to the input end, the first micromirror array, the second micromirror array, and the output end to determine the main affected co-position dual-state attitude adjustment frames. The cover plate connection error that is still retained after room temperature recovery is corrected by the reserved discrete laser writing bits of the corresponding attitude adjustment frames.

[0134] After all cover plates are connected, perform a full rapid calibration path measurement again. When all critical paths meet the specified range, three consecutive measurement results are stable, and the remaining drive range of the micromirrors meets the design requirements, proceed to the full-port acceptance step.

[0135] Thirteen, Full-port acceptance and electronic packaging calibration file generation: After rapid calibration, the first and second micromirror arrays are sequentially controlled to establish target connection states between all input ports and all output ports. For the OCS with 64 input ports and 64 output ports, the transmission coefficients of 4,096 target connection states are measured, and the transmission coefficients of non-target output ports under each target connection state are recorded simultaneously.

[0136] The final acceptance data includes the insertion loss of each target path, the dispersion of insertion loss between ports, the crosstalk of each non-target path, the path change under different temperature conditions, the path change before and after the package cover is connected, and the remaining driving range of the micromirror.

[0137] The electronic packaging calibration file is associated with the OCS device serial number. The file content includes the number of each optical component, the number of each co-position dual-state attitude adjustment frame, the initial transmission matrix, the multi-port optical path error vector, the reversible thermoelastic effect matrix, the permanent adjustment effect matrix, the laser writing record for each time, the cumulative energy of each discrete laser writing bit, the graded release data of the external attitude adjustment mechanism, the temperature cycle data, the cover plate connection data, and the final full-port transmission matrix.

[0138] XIV. Abnormal Status Handling: When the input calibration optical power fluctuation exceeds the set range, stop the current measurement and re-stabilize the multi-channel calibration light source; when the output photoelectric detection array channel gain deviates from the calibration value, re-perform the detection channel gain calibration; when the optical path state does not recover after the resistance heating layer cools, stop the corresponding posture adjustment unit from participating in permanent adjustment; when the effective rank of the permanent adjustment influence matrix is ​​lower than the number of current main optical path error modes, return to the coarse positioning step; when the direction of optical path change after laser writing is opposite to the predicted direction, stop writing in the corresponding laser stress writing area; when the remaining number of discrete laser writing bits is lower than the final state compensation reserved number, stop the current adjustment stage and perform process verification.

[0139] Through the above implementation process, each method step in the claims has a clearly defined execution object, input data, processing procedure, judgment condition, output result, and subsequent connection relationship.

[0140] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising a reference structure" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0141] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A self-calibrating packaging method for multi-port optical paths in an OCS (Optical Channel Switching) system, characterized in that, Includes the following steps: Sp1. Install the input fiber collimator array, the first micro-mirror array, the second micro-mirror array and the output fiber collimator array on the optical component carrier ring of the corresponding co-position dual-state attitude adjustment frame, fix the base connecting ring of each co-position dual-state attitude adjustment frame to the packaging base, and complete the coarse positioning of each optical component through the external attitude adjustment mechanism. Sp2. Divide the input ports into groups of 8 each, configure different modulation frequencies for each input port in the same input port group, inject calibration light into each input port, control the first micromirror array and the second micromirror array to enter the encoding and switching state in sequence, collect the time-domain optical power signal of each output port, and perform synchronous demodulation according to each modulation frequency to obtain the transmission coefficient of the main connection path and the transmission coefficient of the non-main connection path. Sp3. Construct a multi-port optical path error vector based on the logarithmic residual between the transmission coefficient of the main connection path and the target transmission coefficient, the logarithmic residual between the transmission coefficient of the non-main connection path and the crosstalk limit value, and the weight of each calibration path. Sp4. Apply the first identification current and the second identification current to the resistance heating layer on each elastic attitude adjustment beam in each co-position dual-state attitude adjustment frame in sequence. Collect the multi-port optical path error vectors in the first stable temperature rise state, the second stable temperature rise state and the cooling recovery state respectively. Establish the reversible thermoelastic influence matrix based on the difference in optical path error vectors and temperature difference between the two stable temperature rise states. Sp5. Perform singular value decomposition on the reversible thermoelastic influence matrix, retain optical path error modes with singular values ​​not less than 5% of the maximum singular value, select the attitude adjustment unit and laser writing position according to the projection of the current multi-port optical path error vector on the retained optical path error mode, perform calibration laser writing in the first or second laser stress writing area of ​​the selected attitude adjustment unit, and establish a permanent adjustment influence matrix based on the difference of the multi-port optical path error vector before and after calibration laser writing under the same temperature state and the calibration laser energy. Sp6. Based on the current multi-port optical path error vector, permanent trimming influence matrix, single laser energy limit, cumulative laser energy limit, double-sided writing mutual exclusion condition of the same elastic attitude adjustment beam, adjacent writing bit interval condition, local temperature limit, residual stress limit, and the number of reserved final state writing bits, determine the laser stress writing area, laser writing bit, and laser writing energy of the current trimming cycle. Sp7. Perform laser writing according to the determined laser stress writing area, laser writing position and laser writing energy. After the in-situ dual-state attitude adjustment frame recovers to the reference temperature, re-acquire the multi-port optical path error vector. Update the permanent adjustment influence matrix according to the measured optical path error change before and after this laser writing. Sp8, repeat Sp6 and Sp7 until the absolute value of the weighted logarithmic error of each calibration path is not higher than the preset value. Release the holding load of the external attitude adjustment mechanism in three levels: 100%, 50% and 0%. After each level of holding load is released, re-execute the multi-port optical path measurement and laser writing. Sp9. After the external attitude adjustment mechanism is completely released, the optical path transmission matrix is ​​collected in the stable state of -20℃, 25℃ and 70℃ in sequence to complete one heating cycle and one cooling cycle. After returning to 25℃, the retained optical path error is laser-written. Then, the encapsulation cover is connected in segments according to the central symmetry order. After each group of connection areas is completed, the calibration path data is collected and laser writing is performed. Sp10, measure the optical path transmission matrix between all input ports and all output ports, and write the initial optical path transmission matrix, reversible thermoelastic effect matrix, permanent adjustment effect matrix, laser writing record, holding load release data, temperature cycling data, package cover connection data and the final optical path transmission matrix into the electronic package calibration file associated with the OCS device serial number.

2. The OCS multi-port optical path self-calibration packaging method according to claim 1, characterized in that: The co-position dual-state attitude adjustment frame includes a base connecting ring, an optical component support ring located inside the base connecting ring, and attitude adjustment units connected between the base connecting ring and the optical component support ring. Each attitude adjustment unit includes an outer fixed section continuously connected to the base connecting ring, an inner force transmission section continuously connected to the optical component support ring, and an elastic attitude adjustment beam located between the outer fixed section and the inner force transmission section. Thermal isolation grooves are respectively provided on both sides of the elastic attitude adjustment beam. A resistance heating layer, a temperature detection line, a first laser stress writing area, and a second laser stress writing area are provided on the surface of the elastic attitude adjustment beam. The first laser stress writing area and the second laser stress writing area are respectively located on opposite sides of the width direction of the elastic attitude adjustment beam.

3. The OCS multi-port optical path self-calibration packaging method according to claim 2, characterized in that: The first laser stress writing area and the second laser stress writing area are each set with 12 discrete laser writing positions along the length of the elastic attitude adjustment beam. The center distance between adjacent laser writing positions is 180μm. An unwritten isolation section is set between adjacent laser writing positions. Each laser writing position is indexed according to the attitude adjustment frame number, attitude adjustment unit number, laser stress writing area number and position number. In the same adjustment cycle, only one laser stress writing area is selected for the same elastic attitude adjustment beam.

4. The OCS multi-port optical path self-calibration packaging method according to claim 1, characterized in that: The OCS includes 64 input ports and 64 output ports. The 64 input ports are divided into 8 input port groups. The 8 input ports in the same input port group are configured with 8 modulation frequencies in the range of 5kHz to 40kHz. The first micromirror array and the second micromirror array enter 16 coding switching states in sequence. Each input port participates in the output path of the central region, the output path of the edge region, and the output path of the corner region.

5. The OCS multi-port optical path self-calibration packaging method according to claim 1, characterized in that: The calibration path in Sp3 covers the central, edge, and corner regions of the input fiber collimator array, and the central, edge, and corner regions of the output fiber collimator array. It also passes through the central, edge, and corner regions of the first and second micromirror arrays, respectively. The edge region path, corner region path, and long path across the mirror array have higher weights than the short path in the central region.

6. The OCS multi-port optical path self-calibration packaging method according to claim 1, characterized in that: The first stable temperature rise in Sp4 is 5°C higher than the reference temperature, and the second stable temperature rise is 10°C higher than the reference temperature. Each attitude adjustment unit performs two repeated identifications. When the directional angle between the unit temperature rise optical path influence vectors obtained from the two identifications is not higher than 5° and the amplitude difference is not higher than 10%, the corresponding unit temperature rise optical path influence vector is written into the reversible thermoelastic influence matrix.

7. The OCS multi-port optical path self-calibration packaging method according to claim 1, characterized in that: In Sp5, the attitude adjustment unit is selected according to the right singular vector component corresponding to the retained optical path error mode. Calibration laser writing is performed in the unused laser writing position of the selected attitude adjustment unit. After the calibration laser writing is completed, when the difference between the temperature of the attitude adjustment beam and the reference temperature before the calibration laser writing is not higher than 0.5℃, the calibration multi-port optical path error vector is collected, and the vector obtained by dividing the difference between the multi-port optical path error vector before and after calibration by the effective calibration laser energy is written into the permanent adjustment influence matrix.

8. The OCS multi-port optical path self-calibration packaging method according to claim 1, characterized in that: In Sp6, the first laser stress writing area and the second laser stress writing area of ​​each attitude adjustment unit are set as laser writing variables with opposite directions. The solution is to find the laser writing variable that minimizes the weighted sum of the remaining multi-port optical path error, laser writing energy and writing difference between adjacent attitude adjustment units. The actual execution amount of the initial adjustment stage is set to 50% of the solution result, and the actual execution amount of the fine adjustment stage is set to 20% of the solution result.

9. The OCS multi-port optical path self-calibration packaging method according to claim 1, characterized in that: In Sp7, the prediction residual between the actual multi-port optical path error change and the predicted change obtained by multiplying the permanent adjustment influence matrix by the actual laser writing vector is calculated. The prediction residual is then distributed to the corresponding column of the permanent adjustment influence matrix along the direction of the actual laser writing vector. When the directional angle between the actual multi-port optical path error change and the predicted change is higher than 30°, the laser writing bit involved in this laser writing is stopped, and the calibration laser writing is re-executed at the adjacent unused laser writing bit.

10. The OCS multi-port optical path self-calibration packaging method according to claim 1, characterized in that: The electronic packaging calibration file in Sp10 establishes a data hierarchy according to the OCS device serial number, optical component number, co-position dual-state attitude adjustment frame number, attitude adjustment unit number, laser stress writing area number, and laser writing position number. It also records design prior data, incoming material inspection data, input calibration light monitoring data, output time-domain optical power data, heating current data, attitude adjustment beam temperature data, laser pulse data, holding load release data, temperature cycle data, packaging cover plate connection data, and final full-port acceptance data.

Images