A method and system for perovskite photovoltaic module femtosecond laser encapsulation
By real-time monitoring and adjustment of thermal stress and focus deviation in the femtosecond laser encapsulation of perovskite photovoltaic modules, and by employing a dual-parameter determination method based on acoustic emission and spectral data, the problems of abnormal thermal stress and focus deviation in the encapsulation process of existing technologies have been solved, thereby improving the encapsulation quality.
Patent Information
- Application Number
- CN202611200852.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-10
- Publication Date
- 2026-11-17
AI Technical Summary
Existing femtosecond laser packaging technology lacks the ability to sense transient physical changes in the processing area in perovskite photovoltaic modules, leading to abnormal thermal stress and focus deviation, which cannot be monitored and corrected in real time, thus affecting the packaging yield.
The encapsulation method using a controller generates a random skip processing sequence for the local filling area by acquiring the centerline trajectory of the sealing tape, extracts acoustic emission signals and spectral data in real time, calculates the cumulative thermal impact index, performs dual-parameter joint state determination, and performs focus compensation and laser energy adjustment to generate a random skip processing sequence to disperse heat input.
It enables real-time monitoring and correction of thermal stress and focus deviation during the packaging process, improving the hermeticity and mechanical bonding strength of the packaging, and reducing the risk of damage to the sealing material or substrate.
Smart Images

Figure CN123124580A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of laser processing and optoelectronic device packaging technology, specifically to a femtosecond laser packaging method and system for perovskite photovoltaic modules. Background Technology
[0002] The photoelectric active layer of perovskite photovoltaic modules is extremely sensitive to moisture and oxygen in the external environment. Moisture or oxidation can cause a sharp decline in photoelectric conversion efficiency, thus requiring highly reliable hermetically sealed packaging processes to ensure the module's lifespan. Among existing packaging technologies, femtosecond laser packaging, which precisely focuses the energy of an ultrashort pulse laser onto the packaging dielectric layer between the substrate and the cover plate, causing localized melting of the dielectric and achieving a dense connection between the upper and lower layers, is currently a commonly used packaging method in the industry.
[0003] Existing femtosecond laser packaging operations typically rely on open-loop processing modes with preset parameters. In this continuous processing, the continuous scanning of the laser beam along a fixed trajectory often leads to excessive localized heat concentration in the packaging area. The resulting abnormal thermal stress can easily cause microcracks to initiate inside the packaging medium or the surrounding glass substrate. Simultaneously, the glass substrates and cover plates used in large-area perovskite photovoltaic modules objectively exhibit flatness errors and spatial deformations during the initial processing and subsequent lamination and assembly stages. These physical fluctuations cause the preset laser focus to deviate from the actual sealing interface, directly resulting in uneven laser energy absorption at the processing interface and localized poor fusion.
[0004] Because existing open-loop control systems can only mechanically execute fixed trajectories and energy parameters, lacking the ability to sense transient physical changes in the processing area, the equipment cannot monitor local heat accumulation and focal point spatial deviation in real time during operation. When thermal stress microcracks or incomplete fusion points caused by focal point deviation initially appear at the processing interface, the system cannot promptly distinguish the specific causes of the anomaly, nor can it adjust processing parameters or compensate for the trajectory in situ based on the actual sealing state. This processing mode, lacking a feedback mechanism, makes micro-defects prone to propagation under subsequent continuous laser energy input, ultimately compromising the hermeticity and mechanical bonding strength of the entire closed encapsulation area, severely restricting the improvement of perovskite photovoltaic module encapsulation yield. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this invention provides a femtosecond laser encapsulation method and system for perovskite photovoltaic modules. It solves the problems that local heat accumulation during the laser encapsulation process of perovskite thin film devices can easily lead to abnormal thermal stress and microcracks, and that fluctuations in the flatness of the substrate and cover plate can easily cause the processing focus to deviate from the sealing interface. Existing open-loop processing modes cannot simultaneously monitor and correct thermal stress and defocusing defects.
[0006] To address the above problems, the present invention provides the following technical solution: The first aspect of this invention provides a femtosecond laser encapsulation method for perovskite photovoltaic modules, executed by a controller, comprising the following steps: The process involves acquiring the centerline trajectory of the closed sealing strip of the device to be packaged, generating a random skip processing sequence for multiple local filling areas, reading the coordinates of the current target local filling area and performing positioning, activating laser output, extracting the effective acoustic emission signal within the dynamic time window corresponding to the current target local filling area to extract the root mean square value of the bandpass acoustic emission signal, and simultaneously acquiring spectral data to calculate the absolute deviation of the full width at half maximum (FWHM) of the characteristic emission spectral lines. Based on the local filling areas centered on the current target local filling area and already subjected to laser action, the spatiotemporal thermal effect accumulation index is calculated, and the dynamic acoustic emission anomaly judgment applicable to the current target local filling area is determined based on the spatiotemporal thermal effect accumulation index. Threshold; Based on the ratio of the root mean square value of the bandpass acoustic emission signal to the dynamic acoustic emission anomaly judgment threshold, and the absolute deviation of the full width at half maximum (FWHM), a dual-parameter joint state judgment is performed on the current target local filling area, and Z-axis focus compensation, laser single-pulse energy adjustment, and / or supplementary processing are performed on the current target local filling area according to the judgment result; The quality status of the current target local filling area is updated according to the judgment result and the online retest result. When the quality status is qualified, the sequence index of the random jump processing sequence of the local filling area is incremented, and the next target local filling area is located and processed until all local filling areas in the random jump processing sequence of the local filling area are in the qualified processing state.
[0007] Furthermore, a random jump processing sequence for local filling areas, comprising multiple local filling areas, is generated. Specifically, this includes: generating an original set of coordinate points containing multiple coordinate points along the centerline trajectory of the closed sealing strip; randomly sorting the coordinate points in the original set of coordinate points according to a preset minimum center distance threshold, such that the distance between the centers of two consecutively executed local filling areas is not less than the minimum center distance threshold, thus obtaining a random jump processing sequence for local filling areas; and generating corresponding local filling areas centered on each coordinate point in the random jump processing sequence for local filling areas, such that adjacent local filling areas overlap in a predetermined manner along the extension direction of the closed sealing strip, and generating a local reciprocating raster scanning trajectory within each local filling area.
[0008] Furthermore, within the dynamic time window corresponding to the current target local filling area, effective acoustic emission signals are extracted to obtain the root mean square (RMS) value of the bandpass acoustic emission signal. Specifically, this includes: calculating the minimum and maximum propagation distances to the acoustic emission sensor based on the closest and farthest scanning positions in the local scanning trajectory within the current target local filling area, respectively; converting these distances to the minimum and maximum propagation times using pre-stored equivalent acoustic wave propagation speeds; subtracting the pre-set time margin from the sum of the trigger time of the first processing laser pulse in the current target local filling area, the minimum acoustic wave propagation time, and the fixed system delay, as the start time of the dynamic time window; and using the sum of the trigger time of the last processing laser pulse, the maximum acoustic wave propagation time, the fixed system delay, and the post-set time margin, as the end time of the dynamic time window; extracting acoustic emission data segments from the continuously acquired acoustic emission signals according to the dynamic time window; performing digital bandpass filtering on the extracted acoustic emission data segments and removing filtered transition data; and then performing RMS calculation to obtain the RMS value of the bandpass acoustic emission signal.
[0009] Furthermore, the calculation methods for the spatiotemporal thermal impact accumulation index and the dynamic acoustic emission anomaly judgment threshold are as follows: The energy equivalence coefficient, spatial weight, and temporal weight corresponding to each local filling area located within the preset search radius and already subjected to laser action are multiplied to obtain the spatiotemporal thermal impact accumulation index; the initial acoustic emission anomaly judgment threshold and the lower limit of the dynamic acoustic emission threshold corresponding to the current laser single pulse energy level are called, and the initial acoustic emission anomaly judgment threshold is lowered according to the preset exponential decay relationship and the spatiotemporal thermal impact accumulation index, and the lowered threshold is not lower than the lower limit of the dynamic acoustic emission threshold, thus obtaining the dynamic acoustic emission anomaly judgment threshold applicable to the current target local filling area.
[0010] Furthermore, a dual-parameter joint state determination is performed. Specific determination rules include: calculating the ratio between the root mean square value of the bandpass acoustic emission signal and the dynamic acoustic emission anomaly determination threshold as the acoustic emission threshold ratio; when the absolute deviation of the full width at half maximum (FWHM) does not exceed the preset spectral deviation threshold and the acoustic emission threshold ratio is not greater than 1, it is determined to be a normal processing state; when the absolute deviation of the FWHM exceeds the preset spectral deviation threshold and the acoustic emission threshold ratio is not greater than 1, it is determined to be a local defocus state, and the Z-axis focus compensation amount is determined based on the defocus amount and the FWHM offset calibration curve to perform Z-axis focus compensation and supplementary processing ... If the deviation does not exceed the preset spectral deviation threshold and the acoustic emission threshold ratio is greater than 1 but not greater than 1.3, it is judged as an acoustic emission abnormality. The laser output is turned off, and a preset thermal relaxation waiting time is performed. The laser single pulse energy level is reduced and low-energy supplementary processing is performed. If the absolute deviation of the full width at half maximum exceeds the preset spectral deviation threshold and the acoustic emission threshold ratio is greater than 1 but not greater than 1.3, it is judged as a composite abnormality. Z-axis focus compensation, laser single pulse energy reduction and supplementary processing are performed in sequence. If the acoustic emission threshold ratio is greater than 1.3, it is judged as a severe acoustic emission abnormality. Processing is stopped and a suspected crack alarm is output.
[0011] Furthermore, before the dual-parameter joint state determination, the process also includes the steps of initializing the processing state table and the focal plane height estimation table. Specifically, this includes: establishing the processing state table and the focal plane height estimation table in memory; the processing state table is used to record the laser action status field, quality status field, most recent processing timestamp field, trial pulse record field, focus compensation count field, supplementary processing and online retest count field, and overlapping area influence marker field for each local filling area; the focal plane height estimation table is used to record the estimated value of the focal plane height of the sealing interface at different processing coordinates and its height estimation reliability index. When the height estimation reliability index corresponding to the target processing coordinate is lower than the preset value, the optical focusing component is called to measure the actual height before the formal processing begins to update the estimated value of the focal plane height of the sealing interface.
[0012] Furthermore, after updating the quality status of the current target local filling area, the process also includes a neighborhood-weighted update of the focal plane height estimation table: after the current target local filling area has completed the supplementary processing and has been judged as qualified by online retesting, the focal plane height estimation values of the grid nodes that have not yet been processed in the preset neighborhood around the current processing coordinate in the focal plane height estimation table are updated using Gaussian distance weight.
[0013] Furthermore, during the process of activating laser output and extracting signals, the controller or independent hardware timing module generates a local filling area processing start trigger signal, a laser pulse trigger signal, a data acquisition time marker signal, and a spectral integration trigger signal based on a unified hardware timing reference, so that the laser pulse trigger time, the acoustic emission signal sampling time, and the spectral integration time have a convertible unified time reference.
[0014] Furthermore, acquiring spectral data to calculate the absolute deviation of the full width at half maximum (FWHM) of the characteristic emission spectral lines specifically includes: acquiring spectral data collected by the spectrometer during the scanning time of the local filling region of the current target; sequentially performing dark background subtraction, spectral acquisition efficiency correction, and amplitude normalization on the acquired spectral data; and correcting the spectral line broadening caused by the instrument's own line shape according to the instrument broadening correction parameters; fitting the characteristic emission spectral lines using the Lorentz function to obtain the measured FWHM; and calculating the difference between the measured FWHM and the reference FWHM corresponding to the current laser single-pulse energy level to obtain the absolute deviation of the FWHM.
[0015] The second aspect of the present invention provides a femtosecond laser packaging system for perovskite photovoltaic modules, including a controller, and a laser, modulator, beam expander and collimator, optical galvanometer, focusing field lens, Z-axis piezoelectric displacement stage, processing stage, acoustic emission sensor, preamplifier, bandpass filter module, data acquisition module, spectral acquisition lens group, fiber optic probe, optical focusing component and spectrometer, all connected to the controller. A modulator, a beam expander and collimator, an optical galvanometer, and a focusing field lens are sequentially arranged in the output optical path of the laser to focus the laser onto the interface of the sealing material between the substrate and the cover plate on the processing stage. An acoustic emission sensor is arranged on the side of the processing stage, and its output end is sequentially connected to a preamplifier, a bandpass filter module, and a data acquisition module to continuously acquire analog acoustic emission signals and convert them into digital acoustic emission signals to be sent to the controller. A spectral acquisition mirror group is arranged off-axis on one side of the focusing field lens, with its object-side focal point coinciding with the laser processing focal point. A fiber optic probe is arranged at the image-side focal point of the spectral acquisition mirror group and is connected to the input end of the spectrometer via an optical fiber to acquire the plasma emission signal generated in the processing focal area and send it to the controller. The controller is configured to call pre-stored processing trajectory parameters and calibration parameters to execute the femtosecond laser packaging method for perovskite thin films provided in the first aspect of the present invention.
[0016] This invention provides a femtosecond laser encapsulation method and system for perovskite photovoltaic modules. It has the following beneficial effects: 1. This invention constructs a dual-parameter joint state determination mechanism by simultaneously extracting the root mean square value of the bandpass acoustic emission signal and the absolute deviation of the full width at half maximum (FWHM) of the characteristic emission spectrum. This mechanism decouples the local thermal stress state during processing from the degree of focus deviation, enabling the system to accurately distinguish between different states such as normal, defocused, abnormal thermal stress, and composite anomalies. Based on this, it automatically performs Z-axis focus compensation or laser single-pulse energy adjustment, thereby effectively controlling the expansion of packaging defects and improving the hermeticity and mechanical bonding strength of the device packaging.
[0017] 2. This invention utilizes a spatiotemporal thermal effect accumulation index to dynamically set the acoustic emission anomaly detection threshold. During processing, the system comprehensively calculates the spatial distance, time interval, and energy input of the surrounding processed areas, and dynamically lowers the acoustic emission threshold according to an exponential decay relationship. This dynamic adjustment method compensates for the signal baseline drift caused by heat accumulation during processing, avoids misjudging normal thermal effects as processing defects, and improves the detection accuracy of the system in complex heat conduction environments.
[0018] 3. This invention employs constrained random sorting to generate a random skip processing sequence for local filling areas. By setting a minimum center distance threshold, the continuous laser scanning trajectory is spatially discretized, forcing adjacent laser processing areas to maintain a predetermined distance. This path planning method interrupts the continuous heat conduction in local areas, directly dispersing the laser heat input and reducing the risk of damage to the sealing material or substrate due to excessive local heat accumulation. Attached Figure Description
[0019] Figure 1 This is a schematic diagram of a laser packaging system for perovskite thin films according to an embodiment of the present invention; Figure 2 This is a flowchart of a laser encapsulation method for perovskite thin films according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the composite scan trajectory generation logic according to an embodiment of the present invention; Figure 4 This is a flowchart illustrating the multidimensional state space initialization process according to an embodiment of the present invention. Figure 5 This is a timing diagram for dynamic gating data acquisition according to an embodiment of the present invention; Figure 6 This is a schematic diagram illustrating the dynamic stress reference calculation principle according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the two-parameter decoupling and control logic according to an embodiment of the present invention; Figure 8 This is a flowchart of state update and system iteration according to an embodiment of the present invention. Detailed Implementation
[0020] 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.
[0021] See attached document Figure 1 The perovskite thin film laser packaging system of this embodiment includes a controller, and a laser, modulator, beam expander and collimator, optical galvanometer, focusing field lens, Z-axis piezoelectric displacement stage, processing stage, acoustic emission sensor, preamplifier, bandpass filter module, data acquisition module, spectral acquisition lens group, fiber optic probe, optical focusing assembly and spectrometer connected to the controller.
[0022] The laser output from the laser passes sequentially through a modulator, beam expander and collimator, optical galvanometer, and focusing field lens before being focused onto the interface of the sealing material between the substrate and the cover plate. A processing stage supports and fixes the substrate and cover plate, while a Z-axis piezoelectric displacement stage moves the device to be packaged along the Z-axis. In this embodiment, its effective stroke is 200 micrometers, and its closed-loop positioning resolution is no greater than 0.1 micrometers. An optical focusing assembly measures the surface height of the cover plate, and the controller obtains the focal plane height of the sealing interface based on a pre-calibrated height conversion relationship.
[0023] The acoustic emission sensor maintains acoustic coupling with the processing platform through an acoustic coupling layer. Its output is sequentially connected to a preamplifier, a bandpass filter module, and a data acquisition module. In this embodiment, the effective frequency response of the acoustic emission sensor is 100 kHz to 1 MHz, the preamplifier gain is 40 dB, the bandpass range is 200 kHz to 800 kHz, and the data acquisition frequency is 5 MHz sampling points per second. The controller calls the equivalent sound wave propagation velocity, fixed system delay, and acoustic emission detection calibration parameters corresponding to the actual installation method of the acoustic emission sensor.
[0024] The spectral acquisition mirror assembly is positioned off-axis to one side of the focusing field mirror, with its object-side focal point coinciding with the laser processing focal point. It is connected to the spectrometer via a fiber optic probe and fiber optic cable to acquire the plasma emission signal generated in the processing focal region. In this embodiment, the spectral measurement range is 350 nm to 750 nm, with a spectral resolution of no more than 0.02 nm. The controller is pre-calibrated for spectral acquisition efficiency and spectrometer broadening at different processing coordinates and performs corresponding corrections during spectral processing.
[0025] The controller establishes control and / or data connections with the laser, modulator, optical galvanometer, Z-axis piezoelectric displacement stage, data acquisition module, spectrometer, and optical focusing assembly to perform laser start / stop and energy adjustment, XY-Z positioning, acoustic emission data acquisition, spectral triggering and data acquisition, and focal plane height measurement. The laser, data acquisition module, and spectrometer use a unified hardware timing reference. The controller or an independent hardware timing module generates a local filling area processing start trigger signal, a laser pulse trigger signal, a data acquisition time stamp signal, and a spectral integration trigger signal, respectively, ensuring that the laser pulse, acoustic emission sampling, and spectral integration have a calculable unified time reference.
[0026] See attached document Figure 2 This invention provides a laser encapsulation method for perovskite thin films. The laser encapsulation method for perovskite thin films in this embodiment involves a controller calling pre-stored processing trajectory parameters, acoustic emission calibration parameters, spectral calibration parameters, and focal plane height data, and executing the following steps: S1, Generate composite scanning trajectory. The controller generates multiple local filling areas based on the centerline trajectory of the closed sealing strip, randomly sorts each local filling area to meet the minimum center distance constraint, and generates a local scanning trajectory within the local filling area.
[0027] S2, Initialize the machining status table M and the focal plane height estimation table. It also loads acoustic emission detection parameters, spectral detection parameters, and defocus calibration parameters corresponding to the acoustic emission sensor installation method and laser single pulse energy level.
[0028] S3, perform XY-Z positioning, determine the dynamic time window based on the propagation distance between the current target local filling area and the acoustic emission sensor, and extract the root mean square value of the bandpass acoustic emission signal from the continuous acoustic emission signal. Simultaneously acquire spectral data to calculate the absolute deviation of the full width at half maximum (FWHM) of the characteristic emission lines. .
[0029] S4, Calculate the spatiotemporal thermal impact cumulative index based on the historical records of laser effects already received in the neighborhood of the current target local filling area. and according to Determine the appropriate dynamic acoustic emission anomaly detection threshold for the current local filling area.
[0030] S5 performs a dual-parameter joint state determination based on the ratio of the root mean square value of the bandpass acoustic emission signal to the dynamic acoustic emission anomaly determination threshold and the absolute deviation of the full width at half maximum (FWHM). Based on the determination result, it performs Z-axis focus compensation, thermal relaxation, laser single-pulse energy adjustment, reprocessing, and online retesting.
[0031] S6. Update the processing status table M based on the judgment result and the online retest result; when the current local filling area is qualified, increment the sequence index and process the next local filling area; when there is a state to be compensated, keep the current index and perform the corresponding processing; when processing fails or serious abnormality occurs, stop processing until all local filling areas meet the processing completion conditions.
[0032] See attached document Figure 3 The specific implementation process of generating the composite scan trajectory in step S1 can be further included as follows: S101 generates the set of original coordinate points of the center point of the local fill area.
[0033] The controller acquires the centerline trajectory of the closed seal strip located around the periphery of the perovskite functional layer, generates local filling area center points along the centerline at preset intervals, and determines the direction of the local filling area based on the local tangent and normal directions at each center point. In this embodiment, the width of the closed seal strip is 50 micrometers, the center point spacing in straight and gently curved areas is 40 micrometers, the size of the local filling area is 50 micrometers × 50 micrometers, and adjacent local filling areas overlap by approximately 10 micrometers; smaller sampling intervals are used in corners or areas with greater curvature to ensure continuous coverage of the closed seal strip.
[0034] The controller generates N local fill area center points according to the above sampling method, and combines the N local fill area center points into an original coordinate point set: ; in, to These represent the coordinates of the center point of each local filling area in the coordinate system of the machining support table.
[0035] S102, Generate a random jump processing sequence for the local filling region.
[0036] The controller employs a constrained pseudo-random sorting method to sort the center points of the local filling area in the original coordinate point set S, thereby obtaining a random jump processing sequence for the local filling area: ; The sorting process only changes the processing order of each local filling area, without changing the physical coordinates of the center point of the local filling area, the geometric dimensions of the local filling area, or the spatial overlap between adjacent local filling areas.
[0037] When generating a random skip processing sequence for local filling regions, the controller applies a minimum center distance constraint to any two consecutively executed local filling regions. For the sequence... The center point of the i-th local filling region and the center point of the (i+1)th local filling region The center distance between the two satisfies: ; In the formula, and These represent the X-axis and Y-axis coordinates of the center point of the i-th local filling area in the coordinate system of the machining support table, respectively. and These represent the X-axis and Y-axis coordinates of the center point of the (i+1)th local filling region in the same coordinate system, respectively. This represents the minimum center distance threshold between two consecutively executed local filling regions.
[0038] In this embodiment, the minimum center distance threshold Set to 100 micrometers.
[0039] During the specific sorting process, the controller selects candidate center points from those not yet added to the sequence, ensuring that the distance between these center points and the current sequence's end center point is not less than [a certain value]. The controller identifies candidate center points and adds them to the local filling region random jump processing sequence. When there are multiple candidate center points that meet the distance condition, the controller selects one of them based on the value output by the pseudo-random number generator.
[0040] S103 generates the scan trajectory within each local filling area.
[0041] The controller processes the sequence of random jumps according to the local filling area. The order of arrangement is used to read the center points of each local fill area in turn. The controller uses the current local fill area center point as the reference point. Using the geometric center as the reference point, and taking the local tangent direction of the centerline of the closed sealing strip at that center point as the length direction and the local normal direction as the width direction, a local filling area is generated. .
[0042] Local filling area The length along the local tangent direction is 50 micrometers, and the width along the local normal direction is 50 micrometers. When the local filling area... When a portion extends beyond the boundary of the closed sealing strip, the controller adjusts the local filling area according to the boundary of the closed sealing strip. Cut the tape to retain only the portion within the closed sealing strip.
[0043] In the local fill area Inside, the controller generates a local reciprocating grid scan trajectory. This trajectory consists of multiple parallel scan lines, with adjacent lines scanning in opposite directions and a spacing of 4 micrometers between them. The local scan speed is set to 80 millimeters per second, and the laser focal spot diameter is 8 micrometers, resulting in a 50% lateral overlap of focal spots between adjacent scan lines.
[0044] In this embodiment, the scanning time of a single local filling region Set to 7 milliseconds. Scan duration. This includes the scanning time during the effective laser output within the local filling region and the reversal time between adjacent scan lines, but excludes the jump time of the optical galvanometer between the center points of different local filling regions, and also excludes the thermal relaxation waiting time inserted due to failure to meet the minimum center distance constraint.
[0045] In this embodiment, the substrate is made of 0.7 mm thick ITO conductive borosilicate glass, and the cover plate is made of 0.5 mm thick borosilicate glass. A sodium-containing lead-free bismuth borate low-melting-point glass sealing material is disposed between the substrate and the cover plate, with a mass percentage composition of 65% bismuth oxide, 15% boron oxide, 10% zinc oxide, 8% sodium oxide, and 2% copper oxide; wherein sodium oxide is used to provide detectable sodium atom characteristic emission lines, and copper oxide is used to improve the absorption of 1030 nm processing laser.
[0046] A low-melting-point glass sealing material forms a continuous closed prefabricated sealing layer with a width of about 50 micrometers and a thickness of about 15 micrometers. The cover plate and the substrate are aligned and bonded together, and a predetermined clamping force is maintained.
[0047] The laser is a femtosecond pulsed laser with a center wavelength of 1030 nm, a pulse width of 300 femtoseconds, and a repetition frequency of 200 kHz. The initial single-pulse energy is 3.0 microjoules, and the laser focal spot diameter is 8 micrometers. The controller adjusts the single-pulse energy through a modulator. In this embodiment, the energy difference between adjacent laser single-pulse levels is 0.15 microjoules.
[0048] Scan duration of a single local fill region The settings can be adjusted based on the shortest effective integration time of the spectrometer, the effective extraction time of the acoustic emission signal, and the thermal response time of the sealing material. The controller stores the generated random skip processing sequence of local filling areas, the geometric parameters of each local filling area, and the corresponding local scanning trajectory in the controller's memory for subsequent processing steps.
[0049] See attached document Figure 4 The specific implementation process of initializing the multidimensional state space in step S2 can be further included as follows: As a specific implementation of step S2, the controller establishes the processing status table M and the focal plane height estimation table according to the following sub-steps. And complete the initialization of acoustic emission detection parameters and spectral detection parameters: S201, Establish and initialize the processing status table M.
[0050] The controller establishes a status record corresponding to each local filling area based on the number of local filling areas. Each record includes a laser action status field, a quality status field, a recent processing timestamp field, a test pulse record field, a focus compensation count field, a supplementary processing and online retest count field, and an overlap area influence marker field. During initialization, it is set to no formal processing and no quality judgment, the corresponding count is cleared to zero, the test pulse record is set to empty, and the overlap area influence marker is initialized to an unaffected state.
[0051] S202, Establish and initialize the focal plane height estimation table. .
[0052] Before processing begins, the controller selects multiple initial focusing reference points. It measures the height of the upper surface of the cover plate using an optical focusing assembly and obtains the focal plane height of the sealing interface at each reference point based on a pre-calibrated conversion relationship. The controller then fits the coordinates and focal plane heights of each reference point to obtain the initial focal plane height distribution and establishes a focal plane height estimation table according to a preset grid. . Each grid node in the system records at least the node coordinates, the estimated value of the focal plane height, and the height estimation reliability index. The height estimation reliability index is used to determine whether optical focusing or low-energy probing needs to be re-executed before formal processing.
[0053] When the height estimation reliability index corresponding to the target processing coordinates is less than 0.5, the controller drives the optical focusing component to measure the height of the upper surface of the cover plate at the target processing coordinates before starting the formal processing of the local filling area, and obtains the corresponding sealing interface focal plane height according to the sealing interface height conversion relationship.
[0054] When the optical focusing assembly cannot obtain the required height measurement result, or when it is necessary to further determine the focus deviation direction, the controller outputs low-energy test pulses at two preset positions above and below the Z-axis, centered on the current estimated focal plane height, and simultaneously acquires the characteristic emission spectrum at the corresponding positions. By comparing the full width at half maximum (FWHM) measured at the two positions with the reference FWHM corresponding to the current laser single pulse energy level, the focus adjustment direction is determined. The energy and number of low-energy test pulses are pre-calibrated using the sample to ensure that effective characteristic emission spectra are obtained without irreversible melting or cracking.
[0055] The controller writes the single-pulse energy, number of pulses, output height position, and output time of the low-energy probe pulses into the probe pulse record field of the corresponding local filling area in the processing status table M. This is used in subsequent calculations of the space-time thermal effect cumulative index. At that time, the controller incorporates the thermal impact generated by the test pulse into the spatiotemporal thermal impact accumulation index according to the energy and pulse quantity ratio between the test pulse and the formal processing pulse. .
[0056] S203, initialize acoustic emission detection parameters and spectral detection parameters.
[0057] The controller performs low-energy laser trial processing at multiple known propagation distances, records the arrival time of the acoustic emission signal relative to the laser pulse triggering moment, and then... Calibration was performed to obtain the equivalent sound wave propagation speed corresponding to the actual installation method. and fixed system delay In this embodiment, =5600 m / s, =8 microseconds; recalibrate the corresponding parameters when the propagation structure changes.
[0058] The controller is set with the root mean square reference value of acoustic emission for normal processing corresponding to each laser single pulse energy level. and the initial acoustic emission anomaly detection threshold .
[0059] Corresponding to four laser single-pulse energy levels of 3.0 microjoules, 2.85 microjoules, 2.70 microjoules, and 2.55 microjoules, the root mean square reference value of acoustic emission during normal processing is... The values are 0.18 V, 0.17 V, 0.16 V, and 0.15 V, respectively; initial acoustic emission anomaly detection thresholds. The values are 0.32 volts, 0.30 volts, 0.28 volts, and 0.26 volts, respectively.
[0060] The voltage values mentioned above are all root mean square values calculated within the dynamic time window corresponding to the local filling region after the acoustic emission sensor output signal is pre-amplified by 40 dB, band-pass filtered from 200 kHz to 800 kHz.
[0061] The controller also stores the lower limit of the dynamic acoustic emission threshold corresponding to each laser single pulse energy level. The lower limit of the dynamic acoustic emission threshold is used to limit the minimum allowable value of the threshold when calculating the dynamic acoustic emission anomaly determination threshold.
[0062] The spectral detection used the characteristic emission line of sodium atoms with a center wavelength of 588.995 nm. The controller performed linear background subtraction on the spectral data in the wavelength range of 588.5 nm to 589.5 nm, and used the Lorentz function to fit the characteristic emission line of sodium atoms. The full width at half maximum (FWHM) of the fitted peak was used as the spectral parameter for determining the defocus state.
[0063] Corresponding to the four laser single-pulse energy levels of 3.0 μJ, 2.85 μJ, 2.70 μJ, and 2.55 μJ, the reference full width at half maximum (FWHM) at the reference focal plane is... The nanometers are 0.62 nm, 0.59 nm, 0.56 nm and 0.53 nm, respectively.
[0064] The controller will set the full width at half maximum (FWHM) absolute deviation threshold. Set to 0.05 nm. For the current locally filled region, the absolute deviation of the full width at half maximum (FWHM) is... Calculate according to the following formula: ; In the formula, This represents the full width at half maximum (FWHM) of the characteristic emission line measured in the current local filled region; It represents the reference full width at half maximum (FWHM) corresponding to the current laser single pulse energy level.
[0065] The controller judges the validity of the spectral results based on the fluctuation of the full width at half maximum (FWHM) repeated measurement and the Lorentz fitting quality. When the spectral results do not meet the preset validity criteria, the controller does not directly perform defocus judgment, but instead re-acquires the spectrum or performs Z-axis perturbation detection.
[0066] The controller, centered on the reference focal plane, acquires characteristic emission spectra at various laser single-pulse energy levels at multiple known offset positions along the Z-axis. After background subtraction, instrument broadening correction, and Lorentz fitting, a defocusing amount-full width offset calibration curve is established, and the calibration relationship on both sides of the reference focal plane is saved. The absolute deviation of the full width at half maximum (FWHM) is used to determine the degree of defocusing, while the focus compensation direction is determined by combining the optical focusing results, calibration data on both sides of the reference focal plane, or Z-axis perturbation detection results.
[0067] Within the approximately linear compensation range selected in this embodiment, the conversion factor between the full width at half maximum (FWHM) offset and the Z-axis defocus distance is 40 micrometers / nanometer; when the range is exceeded, interpolation is performed based on the calibration curve.
[0068] The controller calls upon the corresponding root mean square reference value of acoustic emission during normal processing based on the current laser single pulse energy level. Initial acoustic emission anomaly detection threshold Lower limit of dynamic acoustic emission threshold Reference half-peak full width And the defocus amount-half-peak full width offset calibration curve.
[0069] The controller incorporates the equivalent sound wave propagation speed, fixed system delay, acoustic emission detection parameters corresponding to each laser single pulse energy level, spectral detection parameters, and a focal plane height estimation table. The processing status table M is stored in memory for subsequent sound propagation time calculation, dynamic time window determination, dual-parameter joint state determination, and Z-axis focus compensation.
[0070] See attached document Figure 5As a specific implementation of step S3, the controller determines the dynamic time window of the acoustic emission signal corresponding to the current local filling area according to the following sub-steps, and simultaneously acquires the acoustic emission signal and spectral signal: S301, calculate the sound propagation time range corresponding to the current local filling area.
[0071] The controller reads the current i-th local fill region. The local reciprocating grid scanning trajectory, the current machining focal plane height, and the three-dimensional installation coordinates of the acoustic emission sensor in the machining platform coordinate system.
[0072] The k-th scan position within the current local fill region Three-dimensional propagation distance of acoustic emission sensor Calculate according to the following formula: ; Minimum propagation distance corresponding to the current local filling region and maximum propagation distance They are respectively: ; ; In the formula, This represents the current i-th local fill region; This indicates the k-th scan position in the current local fill region scan trajectory; , and These represent the X-axis, Y-axis, and Z-axis coordinates of the k-th scan position in the coordinate system of the machining platform, respectively. , and These represent the X-axis, Y-axis, and Z-axis installation coordinates of the acoustic emission sensor in the same coordinate system, respectively.
[0073] The controller is based on the equivalent sound wave propagation speed corresponding to the actual installation method of the acoustic emission sensor. Calculate the minimum sound wave propagation time corresponding to the current local filling area. and maximum sound wave propagation time : ; ; In the formula, This represents the time required for the acoustic emission signal generated in the current local filling region to propagate from the scan position closest to the acoustic emission sensor to the acoustic emission sensor; This indicates the time required for the acoustic emission signal generated in the current local filling area to propagate from the scan position furthest from the acoustic emission sensor to the acoustic emission sensor.
[0074] In this embodiment, the equivalent sound wave propagation speed corresponds to the actual installation method of the acoustic emission sensor. 5600 meters per second, fixed system delay It takes 8 microseconds.
[0075] S302, determine the dynamic time window of the acoustic emission signal corresponding to the current local filling area.
[0076] The controller obtains the trigger time of the first processing laser pulse in the current local filling area. and the triggering time of the last processing laser pulse .
[0077] The controller will determine the trigger time of the first processing laser pulse. Minimum sound wave propagation time and fixed system delay The sum, minus the lead time margin , as the starting moment of the current dynamic time window The formula for its calculation is: ; The controller will determine the trigger time of the last processing laser pulse. Maximum sound wave propagation time Fixed system delay and post-time margin The sum of these values serves as the end point of the current dynamic time window. : ; Therefore, the dynamic time window corresponding to the current i-th local filling region Represented as: ; In this embodiment, the lead time margin Set to 15 microseconds, with a time margin after the setting. The time margin is set to 20 microseconds. The pre-time margin is used to retain baseline data and digital filtering transition data before the arrival of the earliest valid acoustic emission signal, and the post-time margin is used to cover the valid acoustic emission signal generated by the last processing laser pulse and its attenuation process.
[0078] The starting boundary of the dynamic time window uses the minimum sound wave propagation time. The boundary is determined by the maximum sound wave propagation time. The controller uses the same time window definition in subsequent data acquisition and signal processing, and does not replace the minimum and maximum sound wave propagation times with the single sound wave propagation time.
[0079] S303, extract the root mean square value of the bandpass acoustic emission signal and calculate the absolute deviation of the full width at half maximum (FWHM) of the spectrum.
[0080] The data acquisition module continuously acquires the signal output from the acoustic emission sensor at a sampling frequency of 5 MHz per second, and records the sampling data with time stamps according to a unified hardware timing reference. The controller determines the time based on the current dynamic time window. Extract the acoustic emission data segment corresponding to the current local filling area from the continuous sampling data.
[0081] The controller performs digital bandpass filtering on the captured acoustic emission data segments from 200 kHz to 800 kHz.
[0082] After completing the digital bandpass filtering from 200 kHz to 800 kHz, the controller removes the filtered transition data at both ends of the dynamic time window. In this embodiment, 10 microseconds of data are removed from the start and end ends respectively, and the remaining data segment covering the effective acoustic emission signal is retained for root mean square calculation.
[0083] The controller performs root mean square (RMS) calculation on the acoustic emission data segment after removing filtered transition data to obtain the RMS value of the bandpass acoustic emission signal in the current local filling region. : ; In the formula, This represents the nth acoustic emission signal sample value of the current local filling region after bandpass filtering and removal of filtered transition data; This indicates the number of sampling points involved in the root mean square operation.
[0084] root mean square value of bandpass acoustic transmission signal It is used to represent the overall amplitude level of acoustic emission signals in the frequency range of 200 kHz to 800 kHz during the current local filling area processing, and is used for subsequent dynamic acoustic emission anomaly determination.
[0085] During acoustic emission acquisition, the spectrometer is triggered according to a unified hardware timing sequence, and its integration time corresponds to the scanning duration of a single local filling region, which is 7 milliseconds in this embodiment. After spectral integration, dark background subtraction, spectral acquisition efficiency correction, instrument broadening correction, and amplitude normalization are performed sequentially. Finally, Lorentz fitting is performed on the characteristic emission line of sodium atoms at 588.995 nm to obtain the measured full width at half maximum (FWHM). Then calculate the absolute deviation of the full width at half maximum (FWHM). If the spectral results do not meet the validity criteria, re-acquisition or perform Z-axis perturbation detection.
[0086] The controller will use the machining coordinates of the current local filling area, the dynamic time window, and the root mean square value of the bandpass acoustic transmission signal. Measure the full width at half maximum (FWHM) absolute deviation of full width at half maximum The corresponding laser single-pulse energy level is associated with and stored for subsequent dynamic threshold calculation and dual-parameter joint state determination.
[0087] See attached document Figure 6 As a specific implementation of step S4, the controller calculates the spatiotemporal thermal impact accumulation index corresponding to the current local filling area according to the following sub-steps, and determines the dynamic acoustic emission anomaly judgment threshold accordingly: S401, calculate the spatiotemporal thermal impact cumulative index corresponding to the current local filling area.
[0088] The controller uses the current i-th local filling area as the center and queries the laser action records that occurred within a preset search radius Rs before processing began in the current local filling area. The laser actions include formal processing, supplementary processing, and low-energy trial pulse actions. Each laser action record includes at least the action location, single-pulse energy, actual number of pulses, and completion time. When multiple laser actions occur in the same local filling area, each laser action is included in the calculation separately. In this embodiment, the search radius Rs is 1.5 mm.
[0089] Let the coordinates of the position corresponding to the j-th laser action record within the search range be... The center coordinates of the current local fill region are The center distance between the two Calculate according to the following formula: ; The controller records the center distance between the j-th laser action record and the current local filling area. Calculate the corresponding spatial weights : ; In the formula, The feature distance represents the spatial weights. In this embodiment, the feature distance... It is 0.75 mm.
[0090] Let the planned start time of the current local filling area be . The completion time of the j-th laser action record is Then the time interval from the j-th laser action record to the current processing time is... for: ; Among them, only those that meet the requirements will be included. The laser's effect is recorded and included in the calculation. The controller calculates based on time intervals. Calculate the corresponding time weights : ; In the formula, The decay time constant represents the time weight. In this embodiment, the decay time constant... It takes 0.5 seconds.
[0091] The controller also calculates the energy equivalence coefficient corresponding to the laser action record based on the actual single-pulse energy and pulse number of the j-th laser action record. : ; In the formula, This represents the actual single pulse energy corresponding to the j-th laser action record; This indicates the actual number of pulses involved in the laser action; Indicates the reference single pulse energy; This indicates the number of machining pulses contained in a standard local fill area under reference machining conditions.
[0092] In this embodiment, the reference single pulse energy The initial single-pulse energy is 3.0 microjoules, and the reference pulse number is... It is determined based on the pulse repetition frequency and the effective laser output duration within the standard local filling area.
[0093] For the local filling area cut by the boundary of the closed sealing strip, the low-energy test pulse, and the reprocessing process, the controller uses the actual output single pulse energy and the actual number of pulses to calculate the energy equivalence coefficient, and does not use the theoretical number of pulses within the preset scan time to replace the actual number of pulses.
[0094] The controller accumulates the products of the energy equivalence coefficient, spatial weight, and temporal weight corresponding to each laser action record within the search range to obtain the spatiotemporal thermal impact accumulation index corresponding to the current local filling area. : ; In the formula, Indicates that it is located within the search radius The collection of all laser action records completed before the processing of the current local filling area begins; This represents the energy equivalence coefficient of the j-th laser action record; Indicates spatial weights; Indicates time weight.
[0095] Spatiotemporal thermal cumulative index It is a dimensionless computational quantity, and its value increases with the number of laser actions in the neighborhood of the current local filling area, the laser input energy, and the spatial proximity, and decreases with the increase of time elapsed after each laser action.
[0096] S402, determine the threshold for judging dynamic acoustic emission anomalies based on the spatiotemporal thermal effect cumulative index.
[0097] The controller determines the laser single-pulse energy level used in the current local filling area. The initial acoustic emission anomaly detection threshold preset in step S203 is invoked. and the lower limit of dynamic acoustic emission threshold .
[0098] The controller is based on the spatiotemporal thermal effect cumulative index. The initial acoustic emission anomaly detection threshold is lowered according to a preset exponential decay relationship to obtain the dynamic acoustic emission anomaly detection threshold corresponding to the current local filling area. : ; In the formula, This indicates the current i-th local filling region at the laser single-pulse energy level. The corresponding dynamic acoustic emission anomaly detection threshold; This represents the initial acoustic emission anomaly detection threshold at the corresponding laser single-pulse energy level; This represents the cumulative spatiotemporal thermal impact index corresponding to the current local filling area; Indicates the threshold attenuation coefficient; This indicates the lower limit of the dynamic acoustic emission threshold at the corresponding laser single-pulse energy level.
[0099] because For non-negative values, the dynamic acoustic emission anomaly detection threshold calculated according to the above relationship should not be higher than the corresponding initial acoustic emission anomaly detection threshold. The lower limit of the dynamic acoustic emission threshold is used to prevent the exponential decay calculation result from being too low when there are many laser action records in the neighborhood, which would cause normal processing acoustic emission signals to be misjudged as abnormal.
[0100] Threshold attenuation coefficient By pre-determining calibrated samples of similar materials, different results can be obtained by varying the spatial distance, time interval, and laser input energy between locally filled areas. The critical value of acoustic emission anomaly under the given conditions, and based on the anomaly critical value... The relationship between the changes is fitted to determine α. In this embodiment... Take 0.05.
[0101] The lower limit of the dynamic acoustic emission threshold corresponding to each laser single pulse energy level is pre-calibrated based on the ambient background noise, the root mean square reference value of acoustic emission during normal processing and its fluctuation range, and set to be lower than the initial acoustic emission anomaly judgment threshold of the corresponding energy level, so as to avoid the dynamic threshold dropping too low when the spatiotemporal thermal influence is large, which would cause the normal processing signal to be misjudged as abnormal.
[0102] The controller will calculate the dynamic acoustic emission anomaly detection threshold. This is related to the sequence index of the current local filling area, processing coordinates, laser single-pulse energy level, and spatiotemporal thermal effect accumulation index. The threshold is associated with the storage and then provided to the two-parameter joint state determination process in step S5.
[0103] See attached document Figure 7 As a specific implementation of step S5, the controller performs the following sub-steps: based on the root mean square value of the bandpass acoustic emission signal and the absolute deviation of the full width at half maximum (FWHM) of the characteristic emission spectrum, it jointly determines the processing status of the current local filling area, and performs Z-axis focus compensation, laser single-pulse energy adjustment, reprocessing, and online retesting based on the determination results: S501, execute the two-parameter joint state determination.
[0104] The controller reads the root mean square value of the current local fill zone bandpass acoustic emission signal obtained in step S303. Absolute deviation of full width at half maximum And read the dynamic acoustic emission anomaly determination threshold calculated in step S402. .
[0105] The controller can only perform the dual-parameter joint state determination in this step if both the acoustic emission data and the spectral data meet the corresponding validity criteria. When either parameter is invalid, the controller does not output the quality determination result of the current local filling area, but instead re-acquires the corresponding data or performs Z-axis perturbation detection according to the provisions of step S303.
[0106] The controller calculates the acoustic emission threshold ratio of the current local filling region. : ; In the formula, This represents the root mean square value of the bandpass acoustic emission signal in the current local filled area; Indicates the current laser single pulse energy level The following threshold is applied to determine dynamic acoustic emission anomalies in the current local filling area.
[0107] The controller will compare the acoustic emission threshold ratio Absolute deviation of full width at half maximum It is compared with a pre-set judgment condition. The absolute deviation threshold of the full width at half maximum (FWHM) is denoted as... In this embodiment, It is 0.05 nanometers.
[0108] when Not greater than ,and When the value is not greater than 1, the controller determines that the current local filling area is in a normal processing state.
[0109] when Greater than ,and When the value is not greater than 1, the controller determines that the current local filling area is in a local defocus state.
[0110] when Not greater than ,and When the value is greater than 1 but not greater than 1.3, the controller determines that the current local filling area is in an abnormal acoustic emission state. An abnormal acoustic emission state is used to indicate that there is a suspected thermal stress anomaly in the current processing area, but does not directly indicate that a crack has formed.
[0111] when Greater than ,and When the value is greater than 1 but not greater than 1.3, the controller determines that the current local filling area is in a composite abnormal state. A composite abnormal state indicates that local defocusing and acoustic emission abnormalities occur simultaneously.
[0112] when When it is greater than 1.3, regardless of the absolute deviation of the full width at half maximum (FWHM) Whether the absolute deviation threshold of the full width at half maximum (FWHM) is exceeded, the controller will prioritize determining that the current local filling area is in a state of severe acoustic emission abnormality, stop the laser output, and execute the abnormal termination protection.
[0113] When a severe acoustic emission anomaly is accompanied by an absolute deviation of the full width at half maximum (FWHM) exceeding the threshold, the controller adds and saves a local defocus mark in the anomaly record, but does not perform focus compensation or reprocessing before stopping processing protection.
[0114] S502 performs Z-axis focus compensation and updates the focal plane height estimation table when it is determined to be in partial out of focus.
[0115] When the current local filling area is determined to be in a local defocus state, or is determined to be in a complex abnormal state including local defocus, the controller adjusts the absolute deviation of the full width at half maximum (FWHM) according to the defocus amount-full width offset calibration curve corresponding to the current laser single pulse energy level. Converted to the absolute value of the defocus distance.
[0116] When the absolute deviation of the full width at half maximum (FWHM) is within the approximate linear compensation range of step S203, the absolute value of the defocus distance is calculated according to the following formula: ; In the formula, This represents the absolute value of the defocus distance calculated from the spectral measurement results; The conversion factor between the full width at half maximum (FWHM) offset and the Z-axis defocus distance is 40 micrometers per nanometer in this embodiment.
[0117] When the absolute deviation of the full width at half maximum exceeds the approximate linear compensation range, the controller performs piecewise interpolation based on the defocusing amount-full width at half maximum offset calibration curve under the corresponding laser single pulse energy level to obtain the absolute value of the defocusing distance.
[0118] Since the full width at half maximum (FWHM) absolute deviation does not include information about the focus deviation direction, the controller does not... The value directly determines the Z-axis focus compensation direction.
[0119] controller combination The Z-axis focus compensation direction is determined by the estimated focal plane height and the actual height obtained by the optical focusing assembly. When the optical focusing result cannot reliably determine the direction, low-energy trial detection is performed at two preset positions above and below the current processing height on the Z-axis. The full width at half maximum (FWHM) of the characteristic emission spectral line at the two positions is compared with the corresponding reference FWHM. The direction of movement that makes the FWHM closer to the reference value is determined as the Z-axis focus compensation direction.
[0120] The controller generates Z-axis focus compensation amounts with positive and negative directions based on the determined compensation direction and the absolute value of the defocus distance. The absolute value of a single Z-axis focus compensation is no greater than 5 micrometers. ; The number of Z-axis focus compensations allowed to be performed in a single local filling area shall not exceed 3, and the sum of the absolute values of each Z-axis focus compensation shall not exceed 20 micrometers.
[0121] ; In the formula, This indicates the number of Z-axis focus compensation operations performed in the current local fill area; This represents the Z-axis focus compensation amount for the m-th iteration.
[0122] Before performing focus compensation, the controller checks whether the single compensation amount, cumulative compensation amount, number of compensations, and target Z-axis position meet preset limits. In this embodiment, the absolute value of the single compensation amount is no greater than 5 micrometers, the number of compensations for a single local filling area is no more than 3, and the sum of the absolute values of the cumulative compensation amounts is no greater than 20 micrometers. If any limit is exceeded, the current local filling area is updated to a processing failure state; if the limit is met, the corresponding Z-axis compensation is performed.
[0123] The controller updates the focal plane height estimation table based on the Z-axis focal point compensation amount. The estimated height of the focal plane corresponding to the current processing coordinates. After the current local filling area is completed and deemed qualified by online retesting, the controller updates the height estimation reliability index of the current node to 0.9.
[0124] For mesh nodes within a preset neighborhood around the current processing coordinates that have not yet been processed, the controller uses Gaussian distance weights to perform neighborhood-weighted updates on the estimated focal plane height. Any neighboring node... Gaussian distance weight for: ; In the formula, This indicates the center coordinates of the currently defocused local fill area; This represents the coordinates of the neighboring nodes to be updated; Distance parameter representing the range of influence of height compensation.
[0125] The estimated focal plane height of the corresponding neighboring nodes is updated according to the following formula: ; In the formula, This represents the estimated focal plane height before the update; This represents the updated estimated focal plane height.
[0126] Distance parameters The spatial correlation range of the focal plane height at the sealing interface is predetermined.
[0127] S503, Perform the next target local filling area positioning.
[0128] Once the current local filling area is processed satisfactorily and the next target local filling area is determined, the controller reads the XY coordinates of the next target local filling area and... The corresponding estimated focal plane height is used to drive the optical galvanometer and Z-axis piezoelectric displacement stage for parallel positioning. When the reliability of the height estimation is lower than a preset threshold, optical focusing or low-energy probing is performed before formal processing. When the current local filling area is still in a state of needing compensation or retesting, the current sequence index remains unchanged, and the next target local filling area is not located.
[0129] S504 performs thermal relaxation waiting, laser single-pulse energy adjustment, reprocessing, and online retesting based on the acoustic emission abnormality or combined abnormality.
[0130] The controller pre-stores four laser single-pulse energy levels: 3.0 microjoules, 2.85 microjoules, 2.70 microjoules, and 2.55 microjoules. The difference between two adjacent laser single-pulse energy levels is 0.15 microjoules, which is 5% of the initial single-pulse energy of 3.0 microjoules.
[0131] When the current local filling region is determined to be in an acoustic emission abnormal state, that is, when the following conditions are met: ; The controller treats the anomaly as a recoverable acoustic emission anomaly, immediately shuts down the laser output, and performs a 30-millisecond thermal relaxation wait.
[0132] After the thermal relaxation period ends, if the current laser single pulse energy has not yet reached 2.55 microjoules, the controller will reduce the laser single pulse energy by one level.
[0133] After changing the laser single-pulse energy level, the controller calls the root mean square reference value of acoustic emission for normal processing corresponding to the new energy level. Initial acoustic emission anomaly detection threshold Lower limit of dynamic acoustic emission threshold Reference half-peak full width And the defocus amount-half-peak full width offset calibration curve.
[0134] The controller also considers the spatiotemporal thermal impact cumulative index of the current local filling area. Based on the changed laser single-pulse energy level, the dynamic acoustic emission anomaly judgment threshold corresponding to the current local filling area is recalculated.
[0135] After the parameter update is completed, the controller uses the reduced laser single pulse energy to perform supplementary processing on the current local filling area, and re-acquires acoustic emission signals and spectral signals, and performs dual-parameter joint state determination again according to step S501.
[0136] When the online retest result shows a normal processing status, the controller updates the quality status of the current local filling area to "processing qualified".
[0137] When the current laser single pulse energy has reached 2.55 microjoules and the current local filling area is still judged to be in an abnormal acoustic emission state, the controller stops processing, updates the quality status of the current local filling area to processing failure, and outputs a processing energy lower limit alarm.
[0138] When the online system determines that the severe acoustic emission abnormality condition specified in step S501 has been triggered, the controller immediately stops the processing of the current and subsequent local filling areas, updates the quality status of the current local filling area to processing failure, and outputs a suspected crack alarm, without performing the supplementary processing after reducing the energy.
[0139] When the current local filling area is determined to be in a composite abnormal state, the controller first determines the Z-axis focus compensation direction according to step S502 and performs Z-axis focus compensation. After the focus compensation is completed, if the current laser single pulse energy has not yet reached 2.55 microjoules, the controller reduces the laser single pulse energy by one level and then performs supplementary processing and online retesting on the current local filling area.
[0140] After Z-axis focus compensation, reduction of laser single-pulse energy, and reprocessing, the combined dual-parameter judgment is re-executed for composite anomalies. If online retesting restores the processing to normal, the quality status is updated to "processing qualified." If local defocusing persists, compensation is repeated as long as the number of focus compensation attempts and cumulative compensation amount have not reached the upper limit. If recoverable acoustic emission anomalies remain, thermal relaxation is performed, and energy is further reduced within permissible limits before reprocessing. The total number of reprocessing attempts and online retests for a single localized filling area does not exceed three. If normal processing cannot be restored after reaching the upper limit of focus compensation, cumulative compensation amount, minimum laser single-pulse energy, or number of retests, or if severe acoustic emission anomalies occur, the quality status is updated to "processing failed," and processing is stopped.
[0141] See attached document Figure 8 As a specific implementation of step S6, the controller updates the processing status table according to the following sub-steps, determines the execution position of the random jump processing sequence in the local filling area, and outputs a laser sealing processing completion signal after all local filling areas have been processed: S601, Update processing status table M.
[0142] After completing the laser action, status determination, and compensation control of the current local filling area, the controller updates the processing status table M according to the sequence index. When the current local filling area receives formal processing or supplementary processing, its laser action status, the most recent processing timestamp, and the corresponding laser action history are updated; when only receiving low-energy test pulses, it remains in the non-formal processing state, and the test pulse parameters are written to the test pulse record field.
[0143] The actual energy input of formal processing, supplementary processing, and low-energy test pulses are all included in the calculation of subsequent spatiotemporal thermal impact cumulative index in accordance with the provisions of step S401.
[0144] When the processing trajectory of the current local filling area overlaps with an adjacent local filling area that is already in the processing qualified state, the controller updates the influence flag of the overlapping area of the adjacent local filling area from the unaffected state or the reviewed state to the affected and awaiting review state, and records the current local filling area sequence index that caused the state change.
[0145] Once the current local filling area has been processed or supplemented and has been deemed qualified by online retesting, the controller will update the overlap area influence mark of the adjacent local filling areas to the verified status.
[0146] When a local fill area fails the online retest, adjacent local fill areas that overlap with it and are marked as affected and awaiting retesting are not included in the area that has already been sealed. The controller maintains the corresponding affected and awaiting retesting status until the current local fill area completes compensation, reprocessing, and online retesting, or the processing process terminates abnormally.
[0147] The controller updates the quality status field of the current local filling area based on the judgment results of steps S501 to S504 and the online retest results.
[0148] Based on the judgments in steps S501 to S504 and the results of online retesting, the controller updates the quality status of the current local filling area accordingly to: processing qualified, defocused and awaiting reprocessing, acoustic emission abnormal and awaiting online retesting, composite abnormal, or processing failed.
[0149] When the current local filling area reaches the preset number of Z-axis focus compensations, cumulative compensation amount, number of reprocessing and online retests, or energy reduction limit and still cannot return to normal, the Z-axis target position exceeds the allowable working range, serious acoustic emission abnormalities occur, or the focus compensation direction cannot be reliably determined, the controller updates its quality status to processing failure status.
[0150] The controller will only include the local filling area in the area that has been sealed if the current quality status of the local filling area is qualified and there is no affected status pending review in the overlapping area influence mark related to the local filling area.
[0151] S602, Update the sequence index and determine whether the processing cycle has ended.
[0152] After completing the status update in step S601, the controller reads the current quality status of the local filling area.
[0153] When the current quality status of the local filling area is in a state of defocusing and waiting for reprocessing, an acoustic emission abnormality and waiting for online retesting, or a combined abnormality state, the controller keeps the current sequence index i of the random jump processing sequence of the local filling area unchanged, and returns to execute the Z-axis focus compensation, thermal relaxation waiting, laser single pulse energy adjustment, reprocessing and online retesting in step S3 or step S5 according to the corresponding status.
[0154] When the quality status of the current local filling area is in a processing failure state, the controller stops laser output, terminates the processing of subsequent local filling areas, and outputs an abnormal termination alarm.
[0155] The controller will update the sequence index from i to i+1 only if the quality status of the current local filling area is qualified and there is no affected state pending review in the overlapping area influence markers related to the current local filling area.
[0156] When the updated sequence index i is less than or equal to the total number of local fill regions N, the controller reads the random jump processing sequence of the local fill regions. Find the center coordinates and scan trajectory of the i-th target local filling area, and return to step S3.
[0157] The controller sequentially performs XY-Z parallel positioning, dynamic time window determination of acoustic emission signal, acquisition of acoustic emission signal and spectral signal, calculation of dynamic acoustic emission anomaly judgment threshold, and joint state judgment of two parameters for the target local filling area.
[0158] When the updated sequence index i is greater than the total number of local filling areas N, the controller traverses all state records in the processing state table M and determines whether the following condition is met: When the sequence index exceeds the total number of local filling areas N, the controller traverses the processing status table M; only when all local filling areas are in the processing qualified state, there are no related overlapping areas in the state to be reviewed, and the processing process has not triggered the abnormal termination condition, the closed sealing area is determined to be processed and a stop processing command is sent.
[0159] When all the above conditions are met, the controller determines that the laser processing of the closed sealing area is complete and sends a stop processing command to the laser, modulator, optical galvanometer and Z-axis piezoelectric displacement stage.
[0160] The controller saves the processing trajectory, detection parameters, status judgment, compensation and online retest records for this processing, and outputs a laser sealing processing completion signal.
[0161] After laser sealing, the closed sealing area can be tested for microscopic continuity, airtightness, and mechanical bonding strength. In this embodiment, an optical microscope is used for full-circumference continuity testing, and the sealing area has no unfused gaps or through cracks exceeding the preset size; helium mass spectrometry is used for airtightness testing, and the equivalent helium leakage rate is no greater than 1.0 × 10⁻⁶. -8 The shear strength test was conducted using samples from the same batch of processes, with an average shear strength of not less than 8 MPa and a single sample strength of not less than 6 MPa. This test was used to confirm the continuity, airtightness, and mechanical bonding performance after sealing.
[0162] If any of the above processing failures or serious abnormal conditions are triggered during the processing, the controller will stop laser output and subsequent local filling area processing, save the generated processing data, abnormal location, abnormal type and equipment status, and output the corresponding abnormal termination alarm.
[0163] 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 femtosecond laser encapsulation method for perovskite photovoltaic modules, characterized in that, Performed by the controller, the process includes the following steps: Obtain the centerline trajectory of the closed sealing tape of the device to be packaged, and generate a random jump processing sequence of local filling areas including multiple local filling areas; Read the coordinates of the local filling area of the current target and perform positioning, turn on laser output, extract the effective acoustic emission signal within the dynamic time window corresponding to the local filling area of the current target to extract the root mean square value of the bandpass acoustic emission signal, and simultaneously acquire spectral data to calculate the absolute deviation of the full width at half maximum of the characteristic emission spectral lines. Based on the local filling area centered on the current target local filling area and already subjected to laser action, calculate the spatiotemporal thermal effect accumulation index, and determine the dynamic acoustic emission anomaly judgment threshold applicable to the current target local filling area based on the spatiotemporal thermal effect accumulation index. Based on the ratio of the root mean square value of the bandpass acoustic emission signal to the dynamic acoustic emission anomaly determination threshold, and the absolute deviation of the full width at half maximum (FWHM), a dual-parameter joint state determination is performed on the current target local filling area, and Z-axis focus compensation, laser single pulse energy adjustment, and / or reprocessing are performed on the current target local filling area according to the determination result. The quality status of the current target local filling area is updated based on the judgment result and the online retest result. When the quality status is qualified, the sequence index of the random jump processing sequence of the local filling area is incremented, and the next target local filling area is located and processed until all local filling areas in the random jump processing sequence of the local filling area are in the qualified processing state.
2. The method for femtosecond laser encapsulation of perovskite thin films according to claim 1, characterized in that, The generation of the local filling region random jump processing sequence, which includes multiple local filling regions, specifically includes: A set of original coordinate points containing multiple coordinate points is generated along the trajectory of the centerline of the closed sealing strip; According to the preset minimum center distance threshold, the coordinate points in the original coordinate point set are randomly sorted under constraints so that the distance between the centers of two consecutively executed local filling areas is not less than the minimum center distance threshold, thus obtaining the local filling area random jump processing sequence. Using each coordinate point in the random jump processing sequence of the local filling area as the center, corresponding local filling areas are generated respectively, so that adjacent local filling areas form a predetermined overlap along the extension direction of the closed sealing strip, and a local reciprocating grid scanning trajectory is generated in each local filling area.
3. The method for femtosecond laser encapsulation of perovskite thin films according to claim 1, characterized in that, The step of extracting the root mean square value of the bandpass acoustic emission signal by intercepting the effective acoustic emission signal within the dynamic time window corresponding to the current target local filling area specifically includes: Based on the scanning positions closest to and farthest from the acoustic emission sensor in the local scanning trajectory within the current target local filling area, the minimum and maximum propagation distances to the acoustic emission sensor are calculated respectively, and the minimum and maximum propagation times are obtained by converting them according to the pre-stored equivalent sound wave propagation speed. The start time of the dynamic time window is the sum of the trigger time of the first processing laser pulse in the current target local filling area, the minimum acoustic wave propagation time, and the fixed system delay, minus the pre-time margin; the end time of the dynamic time window is the sum of the trigger time of the last processing laser pulse, the maximum acoustic wave propagation time, the fixed system delay, and the post-time margin. Acoustic emission data segments are extracted from continuously acquired acoustic emission signals according to the dynamic time window. After digital bandpass filtering and removal of filtered transition data, root mean square (RMS) calculation is performed on the extracted acoustic emission data segments to obtain the RMS value of the bandpass acoustic emission signal.
4. The method for femtosecond laser encapsulation of perovskite thin films according to claim 1, characterized in that, The calculation methods for the spatiotemporal thermal effect accumulation index and the dynamic acoustic emission anomaly determination threshold are as follows: The spatiotemporal thermal effect cumulative index is obtained by summing the products of the energy equivalence coefficient, spatial weight, and temporal weight of each local filling area located within the preset search radius and already subjected to laser action. The initial acoustic emission anomaly judgment threshold and the dynamic acoustic emission threshold lower limit corresponding to the current laser single pulse energy level are called. According to the preset exponential decay relationship, the initial acoustic emission anomaly judgment threshold is lowered according to the spatiotemporal thermal effect accumulation index, and the lowered threshold is not lower than the dynamic acoustic emission threshold lower limit, so as to obtain the dynamic acoustic emission anomaly judgment threshold applicable to the local filling area of the current target.
5. The method for femtosecond laser encapsulation of perovskite thin films according to claim 1, characterized in that, The specific rules for performing the two-parameter joint state determination include: The ratio between the root mean square value of the bandpass acoustic emission signal and the dynamic acoustic emission anomaly determination threshold is calculated as the acoustic emission threshold ratio; When the absolute deviation of the full width at half maximum (FWHM) does not exceed the preset spectral deviation threshold and the acoustic emission threshold ratio is not greater than 1, it is determined to be a normal processing state. When the absolute deviation of the full width at half maximum exceeds the preset spectral deviation threshold and the acoustic emission threshold ratio is not greater than 1, it is determined to be a local defocus state, and the Z-axis focus compensation amount is determined according to the defocus amount - full width at half maximum offset calibration curve to perform Z-axis focus compensation and reprocessing. When the absolute deviation of the full width at half maximum (FWHM) does not exceed the preset spectral deviation threshold and the acoustic emission threshold ratio is greater than 1 but not greater than 1.3, it is determined to be an abnormal acoustic emission state. The laser output is turned off to perform a preset thermal relaxation wait, the laser single pulse energy level is reduced, and low-energy compensation processing is performed. When the absolute deviation of the full width at half maximum exceeds the preset spectral deviation threshold and the acoustic emission threshold ratio is greater than 1 but not greater than 1.3, it is determined to be a composite abnormal state, and Z-axis focus compensation, laser single pulse energy reduction and reprocessing are performed in sequence. When the acoustic emission threshold ratio is greater than 1.3, it is determined to be a serious acoustic emission abnormality, processing is stopped and a suspected crack alarm is output.
6. The method for femtosecond laser encapsulation of perovskite thin films according to claim 1, characterized in that, Before the dual-parameter joint state determination, the process also includes the steps of initializing the processing state table and the focal plane height estimation table, specifically including: Establish a processing status table and a focal plane height estimation table in memory; The processing status table is used to record the laser action status field, quality status field, most recent processing timestamp field, trial pulse record field, focus compensation count field, supplementary processing and online retest count field, and overlapping area influence mark field for each local filling area; The focal plane height estimation table is used to record the estimated values of the focal plane height of the sealing interface at different processing coordinates and its height estimation reliability index. When the height estimation reliability index corresponding to the target processing coordinate is lower than the preset value, the optical focusing component is called to measure the actual height before the formal processing begins to update the estimated value of the focal plane height of the sealing interface.
7. The method for femtosecond laser encapsulation of perovskite thin films according to claim 6, characterized in that, After updating the quality status of the current target local filling region, the method further includes a neighborhood-weighted update of the focal plane height estimation table: After the local filling area of the current target is completed and deemed qualified by online retesting, the estimated value of the focal plane height of the grid nodes that have not yet been processed in the preset neighborhood around the current processing coordinate in the focal plane height estimation table is updated using Gaussian distance weighting.
8. The method for femtosecond laser encapsulation of perovskite thin films according to claim 1, characterized in that, During the process of activating laser output and extracting signals, the controller or independent hardware timing module generates a local filling area processing start trigger signal, a laser pulse trigger signal, a data acquisition time marker signal, and a spectral integration trigger signal according to a unified hardware timing reference, so that the laser pulse trigger time, the acoustic emission signal sampling time, and the spectral integration time have a convertible unified time reference.
9. The method for femtosecond laser encapsulation of perovskite thin films according to claim 1, characterized in that, The acquisition of spectral data to calculate the absolute deviation of the full width at half maximum (FWHM) of the characteristic emission lines specifically includes: Acquire spectral data collected by the spectrometer within the scanning time of the local filling area of the current target. Then, perform dark background subtraction, spectral acquisition efficiency correction, and amplitude normalization on the collected spectral data in sequence. Finally, correct the spectral line broadening caused by the instrument line shape of the spectrometer itself according to the instrument broadening correction parameters. The characteristic emission spectrum is fitted using the Lorentz function to obtain the measured full width at half maximum (FWHM). The difference between the measured FWHM and the reference FWHM corresponding to the current laser single-pulse energy level is calculated to obtain the absolute deviation of the FWHM.
10. A femtosecond laser encapsulation system for perovskite photovoltaic modules, characterized in that, It includes a controller, and lasers, modulators, beam expanders and collimators, optical mirrors, focusing mirrors, Z-axis piezoelectric displacement stages, machining stages, acoustic emission sensors, preamplifiers, bandpass filters, data acquisition modules, spectral acquisition mirror groups, fiber optic probes, optical focusing components, and spectrometers, all connected to the controller. The modulator, beam expander and collimator, optical galvanometer and focusing field lens are sequentially arranged in the output optical path of the laser to focus the laser onto the interface of the sealing material between the substrate and the cover plate on the processing support table. The acoustic emission sensor is installed on the side of the processing support table, and its output terminal is connected in sequence to a preamplifier, a bandpass filter module and a data acquisition module, for continuously acquiring analog acoustic emission signals and converting them into digital acoustic emission signals to be sent to the controller. The spectral acquisition mirror assembly is mounted on one side of the focusing field mirror, with its object-side focal point coinciding with the laser processing focal point. The fiber optic probe is positioned at the image-side focal point of the spectral acquisition mirror assembly and is connected to the light input end of the spectrometer via an optical fiber. It is used to acquire the plasma emission signal generated in the processing focal area and send it to the controller. The controller is configured to call pre-stored processing trajectory parameters and calibration parameters to execute the femtosecond laser encapsulation method for perovskite thin films as described in any one of claims 1 to 9.
