Chip-level LED device packaging method and system

Abstract

The invention relates to the technical field of chip-scale packaging, in particular to a chip-scale LED device packaging method and system, and the method comprises the steps: collecting bonding point surface topography data, extracting key parameters, calculating a fracture toughness predicted value, generating a bonding pressure dynamic compensation amount when the fracture toughness predicted value is lower than a lower threshold value, and generating a bonding angle composite correction amount when the fracture toughness predicted value exceeds an upper threshold value. And acquiring light intensity distribution data based on the correction amount, generating a dynamic correction amount of lens curvature, calculating a packaging body inclination angle adjustment proportion, and generating a global adaptation amount. By collecting surface morphology data of the bonding point and extracting multiple key parameters, the fracture toughness prediction accuracy is improved, the bonding pressure is dynamically adjusted to strengthen the interface stability, the bonding angle is corrected in combination with morphology feedback, local stress concentration is relieved, the lens curvature is corrected in real time based on light intensity distribution change, and light spot center offset is accurately controlled; the inclination angle parameter is synchronously corrected in combination with the mapping relation between the packaging body inclination angle and the bonding point density, and the packaging optical consistency and the structural integrity are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of chip-level packaging, and in particular to a chip-level LED device packaging method and system. Background Art

[0002] The field of chip-scale packaging technology encompasses the packaging process and structural design of semiconductor devices at the wafer or chip level. The core of this field involves achieving high-density integration and protection of miniaturized electronic components through semiconductor manufacturing technologies such as photolithography, deposition, and etching, with a focus on redistribution layer formation, microbump preparation, plastic encapsulation material filling, and dicing and separation processes in wafer-level packaging. The systematic technical process involves constructing electrical interconnect structures on the wafer surface, optimizing signal transmission paths using thin-film deposition and patterning techniques, and achieving mechanical connection and electrical continuity between the chip and substrate through thermal compression bonding or reflow soldering, ultimately improving device performance stability and reliability.

[0003] Among them, the chip-level LED device packaging method refers to a specific technical solution for packaging light-emitting diode chips based on semiconductor technology. The technical matters targeted by this patent subject include eutectic welding parameter control during the chip die bonding stage, gold wire bonding angle and pressure adjustment during the wire bonding process, and thermal expansion coefficient matching and filling uniformity control of the plastic encapsulation material. Technical means include using high-precision chip placement equipment to achieve positioning and calibration of the chip and substrate, reducing signal transmission impedance by optimizing the wire bonding path design, and using a layered curing process during the plastic encapsulation process to reduce internal stress, thereby completing the coordinated regulation of device structure packaging and optical performance.

[0004] Existing chip-scale packaging technology primarily relies on wafer surface rewiring, soldering, and molding processes, which present significant deficiencies in micro-defect monitoring and dynamic compensation. Wafer surface topography data is often not collected in real time and analyzed comprehensively across multiple parameters, leading to fluctuations in bond quality and the potential for localized stress concentrations. Bonding pressure control often relies on static settings, lacking the ability to adjust in real time to changes in micromechanical properties, creating potential weaknesses within the bond interface. During the wire bonding process, angle settings are often based on standard parameters, lacking personalized optimization based on microscopic feedback, which can easily lead to localized bond wire fracture or deformation. Optical path calibration relies on fixed parameters, making it difficult to address dynamic drift, resulting in a cumulative effect of spot shift and impacting device optical consistency. Package tilt adjustment strategies are often based on empirical corrections, which cannot synchronize with real-time topographic changes and light flux variations. This can easily lead to localized package stress concentrations, which can cause premature device failure. These deficiencies limit the consistency of overall packaging quality and the stability of device performance. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a chip-level LED device packaging method and system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a chip-level LED device packaging method, comprising the following steps: S1: Collect the surface topography data of the bonding point, extract the arithmetic mean height parameter of the surface roughness, the maximum concave depth parameter and the profile curvature radius parameter, calculate the critical stress intensity factor parameter, and generate the fracture toughness prediction value; S2: When the fracture toughness prediction value is less than a preset fracture toughness lower limit threshold, calculating a bonding pressure adjustment amount, increasing the bonding pressure to a target value, and generating a bonding pressure dynamic compensation amount; S3: combining the feedback effect of the dynamic compensation amount of the bonding pressure on the bonding morphology, when the fracture toughness prediction value exceeds the preset fracture toughness upper limit threshold, calculating the bonding angle compensation ratio and generating a composite bonding angle correction amount; S4: Based on the influence of the dynamic correction amount of the bonding angle on the optical path transmission, light intensity distribution data is obtained through the photoelectric detection array, the lateral offset and the longitudinal offset between the light spot coordinates and the target position are calculated, and the dynamic correction amount of the lens curvature is generated; S5: Based on the influence of the dynamic correction amount of the lens curvature on the luminous flux, combined with the dynamic correspondence rule between the distribution density of the bonding points and the geometric inclination angle of the package, the adjustment ratio of the package inclination angle is calculated, the inclination angle parameters are corrected in real time, and the accumulated amount of the light spot offset is monitored synchronously to generate the global adaptation amount of the package inclination angle.

[0007] As a further solution of the present invention, the fracture toughness prediction value includes a surface roughness arithmetic mean height parameter, a maximum concave depth parameter, a profile curvature radius parameter, and a critical stress intensity factor parameter; the bonding pressure dynamic compensation amount includes a pressure compensation reference value, a pressure compensation adjustment amount, and a target pressure value; the bonding angle composite correction amount includes an angle correction ratio, a dynamic correction parameter, and an angle adjustment reference value; the lens curvature dynamic correction amount includes a lateral offset, a longitudinal offset, and the center coordinates of the light spot; and the package inclination global adaptation amount includes an inclination adjustment ratio, a package inclination parameter, and a cumulative light spot offset.

[0008] As a further solution of the present invention, the specific steps of S1 are: S101: Collecting the surface topography data of the bonding point, obtaining the surface height distribution curve based on the three-dimensional profilometer, extracting the absolute value of the longitudinal coordinate to calculate the arithmetic mean height, identifying the vertical distance between the coordinates of the maximum concave point and the average height reference plane, analyzing the curvature radius of the coordinates of the extreme point of the profile curvature, and generating a morphological feature parameter set; S102: calling the morphological characteristic parameter set, combining the fracture mechanics stress field distribution equation with the crack propagation direction correction relationship, establishing a critical stress state equation, determining the critical stress condition through equilibrium point analysis, and generating a critical stress intensity factor parameter; S103: Based on the critical stress intensity factor parameter, combined with the material elastic modulus and Poisson's ratio, the fracture characteristics after material constant correction are determined through the fracture toughness calculation relationship under plane strain conditions, and a fracture toughness prediction value is generated.

[0009] As a further solution of the present invention, the specific steps of S2 are: S201: Obtaining the fracture toughness prediction value and a preset fracture toughness lower limit threshold, comparing the predicted value with the threshold, determining whether the fracture toughness prediction value is lower than the preset fracture toughness lower limit threshold, and generating a fracture toughness state determination value; S202: triggering a pressure compensation rule based on the fracture toughness state determination value, adjusting the bonding pressure reference by applying a compensation coefficient according to the compensation logic defined in the compensation rule and the current bonding pressure value, and generating a bonding pressure adjustment amount under the constraints of the compensation rule; S203: calling the bonding pressure adjustment amount, combining it with the current bonding pressure value, determining the compensated target bonding pressure value, extracting the deviation between the target value and the current value, and generating a dynamic bonding pressure compensation amount.

[0010] As a further solution of the present invention, the specific calculation formula for the deviation between the extracted target value and the current value is: ; Where ΔP represents the current bonding pressure deviation compensation amount, Pt represents the target bonding pressure value after compensation, and Pc represents the current bonding pressure value. represents the sliding average of historical pressure deviations, γ represents the temperature fluctuation compensation factor, λ represents the material deformation resistance coefficient, and η represents the process dynamic adaptation coefficient.

[0011] As a further solution of the present invention, the specific steps of S3 are: S301: obtaining a real-time value of the dynamic compensation amount of the bonding pressure, combining it with the bonding topography feedback data, extracting the topography offset features associated with the pressure compensation, analyzing and integrating the dynamic relationship between the features and the compensation amount, and generating a deformation offset; S302: applying the deformation offset to a fracture toughness assessment process to obtain a current fracture toughness value, comparing it with a preset fracture toughness upper threshold, and if the value exceeds the threshold, extracting the excess difference and marking it as a fracture toughness state identifier; S303: Based on the fracture toughness state identifier, the compensation adjustment calculation formula in the angle correction process is used, combined with the dynamic compensation amount, and processed according to the set ratio and reference value in the process to generate a composite correction amount for the bonding angle.

[0012] As a further solution of the present invention, the specific steps of S4 are: S401: Acquire the dynamic correction value of the bonding angle and the light intensity distribution data of the photoelectric detection array, calibrate the pixel coordinates output by the photoelectric detection array, construct an initial spatial distribution matrix of the light spot based on the calibrated pixel data and the measured distance data, and generate the initial spatial distribution matrix value of the light spot; S402: Calling the initial spatial distribution matrix value of the light spot, extracting the lateral displacement and the longitudinal displacement of the light spot center coordinate, separating the lateral offset and the longitudinal offset based on the extracted displacement data and the target preset position data, and generating a light spot center offset sequence; S403: Match the bonding angle dynamic correction value and the offset data according to the light spot center offset sequence, extract the corresponding curvature change, establish the overall lens curvature dynamic adjustment coefficient based on the correlation between the curvature change and the offset trend, and generate the lens curvature dynamic correction value.

[0013] As a further solution of the present invention, the specific calculation formula for separating the lateral offset and the longitudinal offset based on the extracted displacement data and the target preset position data is: ; in, Representative The lateral displacement of the center of the light spot extracted is Representative The longitudinal displacement of the center of the light spot extracted Represents the horizontal coordinate value of the preset target position, Represents the vertical coordinate value of the preset target position, represents the displacement product correction factor, represents the normalized adjustment factor, Representative The offset separation factor of the secondary separation process.

[0014] As a further solution of the present invention, the specific steps of S5 are: S501: Obtaining the dynamic correction value of the lens curvature and the luminous flux change value, collecting luminous flux dynamic response data according to the lens curvature change interval, combining the corresponding relationship between the curvature change and the luminous flux change, extracting the current luminous flux change characteristics, and generating the luminous flux change increase / decrease rate; S502: calling the luminous flux change increase / decrease rate, detecting the bonding point distribution density and the package geometric tilt angle change value, extracting the package tilt angle adjustment characteristics based on the corresponding relationship between the distribution density and the tilt angle change, and generating a package tilt angle adjustment ratio value; S503: calling the package tilt adjustment ratio value, synchronously monitoring the cumulative change of the light spot offset over time, establishing a tilt correction sequence according to the offset accumulation and the tilt adjustment characteristics, and generating a global package tilt adaptation value.

[0015] A chip-level LED device packaging system, comprising: The surface topography acquisition module acquires the surface topography data of the chip LED device bonding point, collects the surface roughness arithmetic mean height parameter, maximum concave depth parameter and profile curvature radius parameter, calculates the critical stress intensity factor value of the bonding point, and generates the fracture toughness prediction value based on the characteristic index set; The bonding pressure compensation module calls the fracture toughness prediction value, compares the numerical difference between the fracture toughness prediction value and the fracture toughness lower limit threshold value according to a preset fracture toughness lower limit threshold value, extracts the current bonding pressure parameter value, improves and adjusts the bonding pressure parameter according to the difference magnitude, and generates a dynamic bonding pressure compensation amount; The bonding angle correction module calls the bonding pressure dynamic compensation amount, extracts the bonding area topography feedback data, detects the fracture toughness change trend, extracts the existing bonding angle parameter value based on the extent by which the fracture toughness exceeds the preset upper limit threshold, adjusts the angle value according to the excess extent ratio, and generates a composite bonding angle correction amount; The light spot offset detection module calls the composite correction value of the bonding angle, collects the light intensity distribution data of the photoelectric detection array, detects the coordinate position of the center of the light spot, calculates the lateral and longitudinal offsets between the center of the light spot and the target position, analyzes the change of the lens morphology based on the trend of the offset data, and generates a dynamic correction value for the lens curvature; The package tilt adaptation module calls the dynamic correction value of the lens curvature, collects the current tilt parameter value of the package, extracts the bonding point distribution density data and the package geometric tilt data, analyzes the package tilt change trend based on the cumulative offset trend of the light spot and the current tilt parameter, calculates the tilt adjustment ratio, and generates the global adaptation value of the package tilt.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are: The present invention collects surface morphology data of bonding points and extracts multiple key parameters to improve the accuracy of fracture toughness prediction, dynamically adjusts bonding pressure to enhance interface stability, corrects bonding angles based on morphology feedback, alleviates local stress concentration, corrects lens curvature in real time based on changes in light intensity distribution, accurately controls the center offset of the light spot, and synchronously corrects inclination parameters based on the mapping relationship between the package inclination angle and the bonding point density, effectively improving the optical consistency and structural integrity of the package. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the steps of the present invention.

[0018] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0021] See also Figure 1 , a chip-level LED device packaging method, comprising the following steps: S1: Collect the surface topography data of the bonding point, extract the arithmetic mean height parameter of the surface roughness, the maximum concave depth parameter and the profile curvature radius parameter, calculate the critical stress intensity factor parameter of the bonding point, and generate the fracture toughness prediction value; S2: When the predicted fracture toughness value is less than the preset fracture toughness lower limit threshold, the bonding pressure adjustment amount is calculated according to the pressure compensation rule, the bonding pressure is increased to the target value, and the bonding pressure dynamic compensation amount is generated; S3: Combined with the feedback effect of the dynamic compensation of the bonding pressure on the bonding morphology, when the fracture toughness prediction value exceeds the preset fracture toughness upper limit threshold, the bonding angle compensation ratio is calculated using the angle correction rule to generate a composite bonding angle correction value; S4: Based on the dynamic correction of the bonding angle and its influence on the optical path transmission, the light intensity distribution data is obtained through the photoelectric detection array, the lateral and longitudinal offsets between the center coordinates of the light spot and the target position are calculated, and the dynamic correction of the lens curvature is generated; S5: Based on the influence of the dynamic correction of the lens curvature on the luminous flux, combined with the dynamic correspondence between the distribution density of the bonding points and the geometric inclination of the package, the package inclination adjustment ratio is calculated and the inclination parameters are corrected in real time. The accumulated amount of the light spot offset is simultaneously monitored to generate the global adaptation amount of the package inclination.

[0022] The fracture toughness prediction value includes the surface roughness arithmetic mean height parameter, the maximum concave depth parameter, the profile curvature radius parameter, and the critical stress intensity factor parameter. The bonding pressure dynamic compensation includes the pressure compensation reference value, the pressure compensation adjustment amount, and the target pressure value. The bonding angle composite correction includes the angle correction ratio, the dynamic correction parameter, and the angle adjustment reference value. The lens curvature dynamic correction includes the lateral offset, the longitudinal offset, and the center coordinates of the light spot. The global adaptation of the package inclination includes the inclination adjustment ratio, the package inclination parameter, and the cumulative light spot offset.

[0023] See also Figure 1 , the specific steps of S1 are: S101: Collecting the surface topography data of the bonding point, obtaining the surface height distribution curve based on the three-dimensional profilometer, extracting the absolute value of the longitudinal coordinate to calculate the arithmetic mean height, identifying the vertical distance between the coordinates of the maximum concave point and the average height reference plane, analyzing the curvature radius of the coordinates of the extreme point of the profile curvature, and generating a morphological feature parameter set; When collecting the surface topography data of the bonding point, first use a high-precision three-dimensional profile measuring instrument for scanning, fix the sample on the platform, select a 5 mm × 5 mm area for scanning and set the sampling interval to 1 micron. After the scan is completed, the surface height data is exported, the longitudinal coordinate is extracted and its absolute value is taken, and then the sum is divided by the total number of sampling points to obtain the arithmetic average height. Sample A is selected for the operation, and the number of sampling points is set to 2500 points. The height value of each point is between 0.4 and 0.6 microns. The final average height is about 0.5 microns. The lowest point is located in all the sampling point data. Its original height is negative 2.8 microns, relative to the average height reference plane of 0.5 microns. The vertical distance from the lowest point to the reference plane is 3.3 microns. Then the curvature analysis is performed on the surface contour curve. The local curvature radius of each point is estimated by continuous three-point sampling, and the extreme curvature point is selected as the representative feature. The curvature radius of a certain point in the example is 20 microns. When dividing the curvature interval, the area less than 10 microns is set as the high curvature area, the area between 10 microns and 50 microns is set as the medium curvature area, and the area greater than 50 microns is set as the low curvature area. Finally, the radius values ​​corresponding to all the extreme curvature points are collected to form a morphological feature parameter set. During the execution process, all height values ​​and curvature radii are extracted through actual measurement, and no estimation method is used. All benchmark values ​​are determined by batch averaging through actual measurements.

[0024] S102: Calling the morphological characteristic parameter set, combining the fracture mechanics stress field distribution equation with the crack propagation direction correction relationship, establishing the critical stress state equation, determining the critical stress condition through equilibrium point analysis, and generating the critical stress intensity factor parameter; When calling the morphological feature parameter set, the maximum depression depth, average height and curvature extreme radius are selected as the crack initiation features, and the stress calculation is performed using the stress field distribution relationship commonly used in fracture mechanics. The initial crack is assumed to be 10 microns deep and the direction is set to zero degrees. Based on the stress change law at different positions from the crack tip, the stress value is solved point by point, and according to the crack propagation direction correction relationship, the new crack propagation direction is calculated by calculating the stress components at different angles of the crack tip. With each direction correction, the crack path and the corresponding stress state are updated in real time. When establishing the critical stress state equation, the crack length and the surface limit stress value are substituted to derive The corresponding critical stress intensity parameter, if the surface limit stress is set to 1500 MPa and the crack length is 10 microns, the corresponding critical stress intensity parameter can be obtained as approximately 8.4 MPa square root of m. The crack length range is set, less than 50 microns is a microcrack, between 50 and 200 microns is a medium crack, and greater than 200 microns is a macrocrack. At the same time, for the critical stress intensity parameter, less than 5 MPa square root of m is classified as brittle material, between 5 and 20 MPa square root of m is classified as medium material, and greater than 20 MPa square root of m is classified as ductile material. All crack length, stress value and other data are extracted based on the morphological characteristic parameter set and stress calculation, and no simplified estimation is performed.

[0025] S103: Based on the critical stress intensity factor parameter, combined with the material elastic modulus and Poisson's ratio, the fracture toughness calculation relationship under plane strain conditions is used to determine the fracture characteristics after the material constant correction and generate a fracture toughness prediction value; Based on the critical stress intensity parameters, combined with the elastic modulus and Poisson's ratio of the material, the fracture toughness relationship under plane strain conditions is used to determine the fracture properties of the material. The elastic modulus is measured by an ultrasonic measuring instrument, and the Poisson's ratio is obtained by a standard tensile test. In this example, if the elastic modulus of the material is 200 gigapascals and the Poisson's ratio is 0.3, the plane strain elastic modulus is converted to about 219.78 gigapascals. The fracture energy is set to 500 joules per square meter. Under this condition, the predicted fracture toughness value of the material is derived to be about 331.77 MPa multiplied by the square root of meter. When classifying materials, elastic modulus less than 100 gigapascals is classified as low modulus material, between 100 and 300 gigapascals is classified as medium modulus material, greater than 300 gigapascals is classified as high modulus material, Poisson's ratio less than 0.25 is classified as high brittle material, between 0.25 and 0.35 is classified as conventional material, and greater than 0.35 is classified as high toughness material. All elastic modulus, Poisson's ratio, and fracture energy data are obtained through actual testing, and the reasonable range is confirmed by comparing the actual measured values ​​with the material standard library. All set benchmark values ​​match the actual measured values ​​to ensure that the data source is traceable.

[0026] See also Figure 1 , the specific steps of S2 are: S201: Obtaining a fracture toughness prediction value and a preset fracture toughness lower limit threshold, comparing the predicted value with the threshold, determining whether the fracture toughness prediction value is lower than the preset fracture toughness lower limit threshold, and generating a fracture toughness state determination value; When obtaining the fracture toughness prediction value and the preset fracture toughness lower limit threshold, it is first necessary to extract the crack extension data of the sample through material mechanics testing. The test basis is usually based on the ASTM E399 standard, and a linear elastic fracture mechanics test is performed. The actual stress is calculated based on the crack length, fracture load and specimen cross-sectional area data, and the fracture toughness prediction value is obtained in combination with the geometric correction factor. When the crack length is 2 mm, the fracture load is 500 Newtons, and the specimen cross-sectional area is 10 square millimeters, the stress is calculated to be 50 MPa. Then, based on the geometric correction factor 1.12 and the crack length, the fracture toughness prediction value is deduced to be approximately 198.5 MPa root m. The preset fracture toughness lower limit threshold is set according to the material type. For example, the lower limit for titanium alloy materials is 180 MPa. MPa-meter, compare the predicted value with the lower limit threshold. If the predicted value is greater than the lower limit, the state is judged to be normal. If it is lower than the lower limit, the state is judged to be low. Taking ceramic materials as an example, the preset lower limit is 200 MPa-meter. At this time, the predicted value is 198.5 MPa-meter, and the state is judged to be low. Finally, the fracture toughness state judgment value is generated. According to different material standards, the fracture toughness range is defined. For example, metal materials are in the normal range between 150 and 250 MPa-meters, and are in a degraded state if the value is lower than 150 MPa-meter. Combined with actual test data, the fracture toughness state is judged.

[0027] S202: triggering a pressure compensation rule based on the fracture toughness state determination value, adjusting the bonding pressure reference by applying a compensation coefficient according to the compensation logic defined in the compensation rule and the current bonding pressure value, and generating a bonding pressure adjustment amount under the constraints of the compensation rule; When triggering the pressure compensation rule based on the fracture toughness state determination value, the fracture toughness state determination result is first read. If the state is low, a compensation coefficient is set according to the compensation rule. The compensation coefficient is defined as zero when the state is normal and 5% when the state is low. The current bonding pressure value, for example, 1,000 Newtons, is extracted. At the same time, the original baseline pressure value in the bonding process file is called, for example, 950 Newtons. The compensation logic calculates the compensation amount as the product of the baseline pressure value and the compensation coefficient. In this example, the compensation amount is 47.5 Newtons. If the fracture toughness state is normal, the compensation amount is zero. The pressure compensation amount is used to update the subsequent target bonding pressure setting value. This action ensures that the pressure is appropriately increased when the fracture toughness is insufficient, making the overall bonding effect more stable and reliable. In actual manufacturing scenarios, such as semiconductor bonding, if the wafer fracture toughness is detected to be reduced, the bonding pressure is automatically adjusted from 950 Newtons to approximately 997.5 Newtons to compensate for material fragility.

[0028] S203: calling the bonding pressure adjustment value, combining it with the current bonding pressure value, determining the compensated target bonding pressure value, extracting the deviation between the target value and the current value, and generating a dynamic bonding pressure compensation value; The specific calculation formula for extracting the deviation between the target value and the current value is: ; Where ΔP represents the current bonding pressure deviation compensation (unit: N), Represents the target bonding pressure value after compensation (unit: N), Represents the current bonding pressure value (unit: N), represents the sliding average of historical pressure deviations (unit: N), γ represents the temperature fluctuation compensation factor (dimensionless), λ represents the material deformation resistance coefficient (unit: N^{-1}), and η represents the process dynamic adaptation coefficient (dimensionless); According to the formula structure, the parameter values ​​are as follows: the current bonding pressure value Pc is measured in real time by the pressure sensor as 42.3N, the compensated target bonding pressure value Pt is calculated as 47.5N by adding the bonding pressure adjustment amount to the current value, and the historical pressure deviation sliding average The moving average of the absolute values ​​of the first 10 pressure deviations was calculated to be 0.56 N. The temperature fluctuation compensation factor γ was mapped to 0.23 based on the temperature-pressure compensation comparison table after monitoring the current ambient temperature fluctuation amplitude (±2.1°C) through the temperature sensor. The material deformation resistance coefficient λ was fitted to 0.048 N⁻¹ through the slope of the relationship between the deformation variable and pressure in the material deformation test experiment. The process dynamic adaptation coefficient η was set to 1.1 according to the dynamic weight range recommended in the equipment process manual.

[0029] The formula calculation process is expanded step by step: calculate : ; ; calculate : ; ; ; Substitute into the formula to calculate ΔP: ; The result shows that the current bonding pressure deviation compensation is 13.18N, which is determined by the absolute deviation between the target pressure and the current pressure, the sliding average of the historical deviation, and the comprehensive correction term of temperature fluctuation and material deformation. After being adjusted by the process dynamic adaptation coefficient, it is directly used to generate the bonding pressure dynamic compensation.

[0030] See also Figure 1 , the specific steps of S3 are: S301: Acquire a real-time value of the dynamic compensation amount of the bonding pressure, combine it with the bonding topography feedback data, extract the topography offset features associated with the pressure compensation, analyze and integrate the dynamic relationship between the features and the compensation amount, and generate the deformation offset; When obtaining the real-time value of the dynamic compensation of the bonding pressure, it is necessary to configure a high-frequency response pressure sensor on the bonding device, such as a high-sensitivity micro pressure sensor installed on the bonding head, which collects pressure data at a sampling rate of 0.01 seconds to form a continuous time series. During the real-time monitoring process, the pressure difference between each two adjacent sampling points is used as the unit time change, and the pressure dynamic compensation is updated in real time according to the pressure rise or fall in different stages. Subsequently, combined with the bonding morphology feedback data processing, a high-precision laser confocal microscope is used to continuously obtain the microscopic morphology of the bonding interface. In the image processing, an edge detection method is selected, such as the Sobel algorithm to extract the local height change area, and the local maximum height and minimum height are calculated. The difference is used to determine the morphological offset characteristics of each local area, and the average offset value of each local feature is further used as the representation value of the overall morphological offset. These morphological offset characteristics are dynamically matched with the pressure compensation data at the corresponding time point, and the relationship between the two is described by linear regression modeling. The fitting process adopts the minimum mean square error principle to ensure that the sum of square errors of each sample is minimized, and a mathematical relationship between the morphological offset and the pressure compensation amount is established. For example, when the real-time pressure compensation amount sampled in the bonding process is 15 kilopascals, and the corresponding overall morphological offset is 2.5 microns on average, it can be deduced from the linear relationship that the deformation offset at this time is approximately 2.7 microns. The deformation offset is generated as the final dynamic data for subsequent fracture toughness evaluation processing.

[0031] S302: Applying the deformation offset to the fracture toughness assessment process to obtain the current fracture toughness value, and comparing it with the preset fracture toughness upper threshold. If the value exceeds the threshold, extracting the excess difference and marking it as a fracture toughness state identifier; When applying deformation offset to the fracture toughness evaluation process, it is first necessary to select a suitable fracture toughness evaluation model. Usually, an evaluation formula based on crack propagation theory is selected for calculation, which includes parameters such as shape factor, material tensile strength and crack length. In the initial stage, the ultimate tensile strength of the material is set to 500 MPa, and the initial crack length is 0.5 mm. Then, the deformation offset generated in the previous section is directly added to the original crack length to form a new crack propagation length. The current fracture toughness value is calculated based on the new crack size and the established shape factor. In actual scenarios, the shape factor usually ranges from 1.12 to 1.5. The appropriate value is selected according to the geometric characteristics of the sample. For example, 1.12 is taken as the standard. After completion After the crack is corrected, a new fracture toughness value is obtained. When the calculated result is 628 MPa-mm, it is directly compared with the preset upper threshold of 600 MPa-mm. If the calculated value exceeds the upper threshold, the excess part is extracted as the fracture toughness state identifier, and the excess range is further divided into intervals. For example, when the fracture toughness value does not exceed 600 MPa-mm, it is in a normal state. If the fracture toughness value is between 600 and 650 MPa-mm, it is classified as a warning interval. When it exceeds 650 MPa-mm, it is classified into a danger interval. In this example, since the fracture toughness value is 628 MPa-mm, it falls into the warning interval, so the corresponding warning identifier is generated to provide a basis for subsequent compensation adjustments.

[0032] S303: Based on the fracture toughness state identifier, the compensation adjustment calculation formula in the angle correction process is used, combined with the dynamic compensation amount, and processed according to the ratio set in the process and the reference value to generate a composite correction amount for the bond angle; According to the fracture toughness state identifier, compensation adjustment is performed in the angle correction process. It is necessary to first determine the standard bonding angle and the current actual measurement angle, and then perform dynamic correction according to the set compensation adjustment calculation method. The specific steps are to determine the corresponding weight adjustment coefficient according to the fracture toughness state identifier. For example, when in the warning range, the weight is set to 1.2. The dynamic compensation amount is combined with the pressure change amplitude to form the pressure correction contribution. Based on the deviation between the current measurement angle and the reference angle, the weight correction is superimposed, and the angle compensation corresponding to the pressure change is added to form a preliminary correction angle. In the actual case, when the reference angle is set to 90 degrees, the actual measurement angle is 92 degrees, and the pressure dynamic compensation amount is 15 kPa, according to the set regulations First, the adjustment amount is calculated based on the deviation, that is, the result of subtracting the reference angle from the measured angle is multiplied by a weight of 1.2. Then, the additional angle offset caused by the pressure is calculated based on the ratio of each kiloPascal pressure compensation amount to a 0.05-degree angle change. The two are added together to obtain the preliminary correction angle. Finally, it is processed according to the set proportional coefficient, usually set to 0.8, to balance the response speed and compensation stability. The final bonding angle obtained after correction is approximately 92.52 degrees, and the composite correction amount is the difference between the reference angle and the corrected angle, which is 2.52 degrees. In the process, the weight adjustment coefficient, proportional coefficient and pressure angle conversion coefficient must be determined based on actual equipment parameters, material properties and empirical data, and are continuously adjusted and optimized according to specific test data in application.

[0033] See also Figure 1 , the specific steps of S4 are: S401: Acquire the dynamic correction value of the bonding angle and the light intensity distribution data of the photoelectric detection array, calibrate the pixel coordinates output by the photoelectric detection array, construct the initial spatial distribution matrix of the light spot based on the calibrated pixel data and the measured distance data, and generate the initial spatial distribution matrix value of the light spot; When obtaining the dynamic correction value of the bonding angle and the light intensity distribution data of the photoelectric detection array, the angle change of the optical component is first continuously collected through a high-precision gyroscope device, and a sensor with an accuracy better than 0.1% is used to keep the data continuously and stably recorded. Then, the photoelectric detection array equipped with a high-sensitivity photosensitive chip is called to collect the light intensity value of each pixel in the array. Through preliminary data processing, the influence of the array dark current and noise is eliminated, and normalization is performed to make the light intensity value uniform between zero and one. Then, the pixel coordinate calibration operation of the photoelectric detection array is carried out. A standard laser source is used in conjunction with a micro-displacement platform to gradually move the plane with a fixed step distance. At the same time, the array response changes are synchronously recorded. By calculating the actual displacement changes corresponding to the pixel points, a one-to-one correspondence table is established. Ten samples are taken during the measurement process to obtain the corresponding values. The average value is used to reduce the system error. For example, when the standard moving step is 0.1 mm, the corresponding pixel column is shifted by five pixels, that is, the actual displacement of each pixel corresponding to 0.02 mm is calculated. After the calibration is completed, combined with the preset ranging parameters, for example, the target distance is set to 500 mm, and the allowable error range is 0.5 mm. According to the calibration data and the ranging data, the initial spatial distribution matrix of the light spot is constructed. In a specific example, for example, the pixel point numbered 100 rows and 200 columns is normalized to obtain a light intensity of 0.85. At the same time, the calibration table indicates that its corresponding physical position is 10 mm horizontally and 20 mm vertically. Then the pixel point is recorded in the preliminary spatial matrix as 10 mm horizontally, 20 mm vertically, and a light intensity of 0.85. All pixel data are processed in this way to complete the generation of the initial spatial distribution matrix of the light spot.

[0034] S402: Calling the initial spatial distribution matrix value of the light spot, extracting the horizontal displacement and vertical displacement of the light spot center coordinate, separating the horizontal offset and the vertical offset based on the extracted displacement data and the target preset position data, and generating a light spot center offset sequence; Based on the extracted displacement data and the target preset position data, the specific calculation formula for separating the lateral offset and the longitudinal offset is: ; in, Representative The lateral displacement of the center of the light spot extracted is Representative The longitudinal displacement of the center of the light spot extracted Represents the horizontal coordinate value of the preset target position, Represents the vertical coordinate value of the preset target position, represents the displacement product correction factor, represents the normalized adjustment factor, Representative The offset separation coefficient of the secondary separation process; In the formula, For the The lateral displacement of the secondary spot center is extracted from the initial spatial distribution matrix of the spot through the spot center detection algorithm. The actual measurement value range is ±5μm based on the accuracy of the laser calibration equipment. ; For the The longitudinal displacement of the secondary spot center is detected in the same way as the lateral displacement. ; is the preset target horizontal coordinate value, which is defined by the system configuration parameters. ; is the preset target longitudinal coordinate value, the example takes ; is the displacement product correction factor, which is set based on the mean variance of the spot offset in the historical data. , whose value increases with the displacement product term The fluctuation range is adjusted, and the fluctuation range is ; is the normalized adjustment factor, based on the displacement ratio The distribution characteristics are set as , whose value is limited by the normalization requirement to .

[0035] Substitute the numerical value into the example for calculation: Calculate the numerator : = ; Calculate the denominator : ; Calculate the value of a fraction within a square root: ; Compute the square root term: ; Calculate the second term : ; Combining the two calculation results: ; This result shows that the offset separation coefficient The value is 127.919, which directly reflects the comprehensive separation degree of the horizontal and vertical offsets. The larger the value, the greater the deviation between the actual spot center position and the preset target. The calculation result of the formula is used to generate the spot center offset sequence. Compare with the preset threshold to determine whether to trigger the position calibration operation.

[0036] S403: Matching the dynamic correction value of the bonding angle with the offset data according to the spot center offset sequence, extracting the corresponding curvature change, establishing the overall lens curvature dynamic adjustment coefficient based on the correlation between the curvature change and the offset trend, and generating the dynamic correction value of the lens curvature; According to the center offset sequence of the light spot, all the offset data are first smoothed using fixed window smoothing, that is, two points before and after the current data point are taken, a total of five points are averaged, and the current point is replaced to eliminate noise interference. The sliding window size is fixed to five groups of data. After processing, a smooth center offset sequence is generated, and then matched according to the correspondence between the offset and the dynamic correction amount of the bonding angle. The correction amount and the offset are matched according to a preset linear proportional relationship. For example, an offset of 0.1 mm corresponds to a correction angle of 0.05 degrees, and an offset of 0.2 mm corresponds to 0.1 degrees. Based on this, a proportional table is established to extract the correction angle data corresponding to each group of offsets. On this basis, according to the set value of the optical path distance, for example, 500 mm, the curvature change is converted by dividing the angle change by the distance. The curvature change is less than 5 times A slight change is considered when the change is 10-6 per millimeter, a medium change is considered when it is between 5 times 10-6 and 2 times 10-5 per millimeter, and a significant change is considered when it is greater than 2 times 10-5 per millimeter. Based on the relationship between the curvature change trend and time evolution, the least squares fitting method is used to determine the curvature change adjustment coefficient. The dynamic adjustment function is set to a linear expression. The fitting coefficient is obtained by minimizing the sum of squared errors in the fitting process. For example, the correction angle corresponding to a certain moment is 0.1 degrees, which is converted to radians of approximately 0.001745. Divided by the optical path distance of 500 mm, the curvature change is calculated to be approximately 3.49 times 10-6 per millimeter. Finally, the dynamic correction amount of the overall lens curvature is fitted based on the change amount at all times, and recorded for subsequent lens adjustment execution module calls.

[0037] See also Figure 1 , the specific steps of S5 are: S501: Obtaining the dynamic correction value of the lens curvature and the luminous flux change value, collecting the luminous flux dynamic response data according to the lens curvature change interval, combining the corresponding relationship between the curvature change and the luminous flux change, extracting the current luminous flux change characteristics, and generating the luminous flux change increase / decrease rate; When obtaining the dynamic correction amount of lens curvature and the change value of luminous flux, it is first necessary to determine the initial curvature radius of the lens through a high-precision laser displacement sensor, for example, with 10 mm as the initial value. Then, under controlled conditions, use a precision mechanical device to gradually apply control force to the lens surface with a tiny step amount of 0.01 mm per step, and detect the curvature change value after each application in real time. At the same time, use an integrating sphere and a high-precision optical power meter to measure the luminous flux value under the corresponding state. After each application, a set of curvature change and luminous flux change values ​​are obtained. By comparing the curvature change and luminous flux change before and after application, the dynamic response relationship of the luminous flux change is summarized according to the change trend. In the specific operation process, the curvature change can be controlled within plus or minus 0.5 mm, and the luminous flux change range can be controlled within plus or minus Within 100 lumens, avoid exceeding the normal working range. For the luminous flux change after each curvature fine-tuning, the change rate can be calculated by a simple ratio method. For example, when the curvature decreases by 0.1 mm, the luminous flux decreases by 10 lumens, and the change rate is 100 lumens per millimeter. Using multiple groups of data, the overall change characteristics are extracted by linear fitting method, and abnormal values ​​of changes are eliminated. For example, data with a change rate of more than 120 lumens per millimeter are defined as abnormal points and eliminated. In practical applications, the focusing lens material can be polymethyl methacrylate with a thickness of 2 mm, and the force point is 5 mm away from the center of the optical axis. After collecting the curvature change of each group and the corresponding luminous flux change data, the current luminous flux change characteristics are comprehensively extracted, and finally a dynamic change increase or decrease rate is formed as the basis for subsequent adjustments.

[0038] S502: Invoking the luminous flux change rate to detect the bonding point distribution density and the package body geometric tilt angle change value, extracting the package body tilt angle adjustment characteristics based on the corresponding relationship between the distribution density and the tilt angle change, and generating the package body tilt angle adjustment ratio value; After calling the luminous flux change increase or decrease rate, first select a typical working point based on the relationship between the curvature change and the luminous flux change, such as the luminous flux change rate corresponding to a curvature change of 0.05 mm, for example, a decrease of 95 lumens per millimeter. Then it is necessary to detect the distribution density of the bonding points on the surface of the package, obtain the surface image through high-magnification microscopy, use image analysis software to perform grayscale segmentation, threshold segmentation and mark and identify the number of all bonding points, and count the number of bonding points within 1 square millimeter, such as obtaining a result of 150 points. At the same time, the initial geometric inclination of the package is measured by a three-dimensional contour scanner. For example, the initial inclination is measured to be 2.5 degrees. In the process of gradually changing the amount of curvature applied, the change in the inclination is monitored in real time. For example, when the curvature increases by 0.05 mm, the inclination changes to 0.56 degrees. By averaging the bonding points, the initial geometric inclination of the package is measured. By introducing the degree and the luminous flux change rate into the tilt change model for correlation analysis, a simple linear relationship can be established. For example, each increase in the bonding point density by 1 point corresponds to a 0.005-degree tilt change, and each increase in the luminous flux change rate causes a 0.002-degree tilt change. By substituting actual parameters for calculation, for example, if the distribution density is 150 points per square millimeter and the change rate decreases by 95 lumens per millimeter, the calculated tilt change is 0.75 minus 0.19, or 0.56 degrees. Based on the initial tilt angle of 2.5 degrees, the calculated tilt adjustment ratio is 22.4%. Through this processing method, the package tilt adjustment characteristics are extracted for use in subsequent global adjustments of the package. It is suitable for silicon carbide-based packages with a size of 10 mm by 10 mm and a height of 2 mm, and the surface roughness is controlled within 0.3 microns.

[0039] S503: Calling the package body tilt adjustment ratio value, synchronously monitoring the cumulative change of the light spot offset over time, establishing a tilt correction sequence based on the offset accumulation and the tilt adjustment characteristics, and generating a global adaptation value of the package body tilt; After calling the package body tilt adjustment ratio value, it is necessary to synchronously monitor the cumulative offset of the light spot over time in the actual environment, set the initial moment and the initial light spot position, for example, the initial position is the coordinate origin, and the high-speed camera system records the center position of the light spot in real time at a speed of 1 frame per second. After continuous recording, the offset data of the light spot relative to the initial position at each moment can be obtained. For example, the offset is 5 microns in the first second, and the cumulative offset is 12 microns in the second second. The offset calculation uses the straight-line distance between the center position coordinates, and the total offset is obtained through continuous accumulation. During the monitoring process, if the total offset exceeds the set threshold, for example, the threshold is set to 50 microns, the tilt correction operation is triggered. During the tilt correction process, the cumulative offset is compared with the premise. The tilt adjustment ratio is calculated synchronously. For example, when the cumulative offset is 17 microns, the corresponding tilt correction coefficient is the adjustment ratio of 22.4% multiplied by the ratio of the offset to the threshold, that is, 17 divided by 50 multiplied by 22.4%. The calculated correction value is about 7.6%. This correction value is synchronously transmitted to the micro-electric actuator module for tilt adjustment. After each adjustment, the spot position continues to be recorded in real time to ensure that the correction action is updated synchronously with the actual offset change. It is suitable for alumina ceramic packages with a size of 8 mm by 8 mm, an overall thickness of less than 2 mm, and a surface finish of 0.2 microns. Through the above processing, the global adaptation amount of the package tilt is formed, providing basic parameters for subsequent further dynamic compensation.

[0040] See also Figure 2 , a chip-level LED device packaging system, comprising: The surface topography acquisition module acquires the surface topography data of the chip LED device bonding point, collects the surface roughness arithmetic mean height parameter, maximum concave depth parameter and profile curvature radius parameter, calculates the critical stress intensity factor value of the bonding point, and generates the fracture toughness prediction value based on the characteristic index set; The bonding pressure compensation module calls the fracture toughness prediction value, compares the numerical difference between the fracture toughness prediction value and the fracture toughness lower limit threshold value based on the preset fracture toughness lower limit threshold value, extracts the current bonding pressure parameter value, and adjusts the bonding pressure parameter based on the difference amplitude to generate the bonding pressure dynamic compensation amount; The bond angle correction module uses dynamic compensation for bond pressure, extracts feedback data on bond area topography, detects fracture toughness trends, extracts existing bond angle parameters based on the extent to which fracture toughness exceeds a preset upper threshold, adjusts the angle value in proportion to the excess, and generates a composite bond angle correction. The light spot offset detection module calls the composite correction value of the bonding angle, collects the light intensity distribution data of the photoelectric detection array, detects the coordinate position of the light spot center, calculates the lateral and longitudinal offsets between the light spot center and the target position, analyzes the lens morphology changes based on the offset data trend, and generates a dynamic correction value for the lens curvature; The package tilt adaptation module calls the dynamic correction value of the lens curvature, collects the current tilt parameter value of the package, extracts the bonding point distribution density data and the package geometric tilt data, analyzes the package tilt change trend based on the cumulative offset trend of the light spot and the current tilt parameter, calculates the tilt adjustment ratio, and generates the global adaptation value of the package tilt.

[0041] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A chip-level LED device packaging method, characterized in that: The following steps are involved: S1: Collect the surface topography data of the bonding point, extract the arithmetic mean height parameter of the surface roughness, the maximum concave depth parameter and the profile curvature radius parameter, calculate the critical stress intensity factor parameter, and generate the fracture toughness prediction value; S2: When the fracture toughness prediction value is less than a preset fracture toughness lower limit threshold, calculating a bonding pressure adjustment amount, increasing the bonding pressure to a target value, and generating a bonding pressure dynamic compensation amount; S3: combining the feedback effect of the dynamic compensation amount of the bonding pressure on the bonding morphology, when the fracture toughness prediction value exceeds the preset fracture toughness upper limit threshold, calculating the bonding angle compensation ratio and generating a composite bonding angle correction amount; S4: Based on the influence of the dynamic correction amount of the bonding angle on the optical path transmission, light intensity distribution data is obtained through the photoelectric detection array, the lateral offset and the longitudinal offset between the light spot coordinates and the target position are calculated, and the dynamic correction amount of the lens curvature is generated; S5: Based on the influence of the dynamic correction amount of the lens curvature on the luminous flux, combined with the dynamic correspondence rule between the distribution density of the bonding points and the geometric inclination angle of the package, the adjustment ratio of the package inclination angle is calculated, the inclination angle parameters are corrected in real time, and the accumulated amount of the light spot offset is monitored synchronously to generate the global adaptation amount of the package inclination angle.

2. The chip-level LED device packaging method according to claim 1, wherein: The fracture toughness prediction value includes a surface roughness arithmetic mean height parameter, a maximum concave depth parameter, a profile curvature radius parameter, and a critical stress intensity factor parameter; the bonding pressure dynamic compensation amount includes a pressure compensation reference value, a pressure compensation adjustment amount, and a target pressure value; the bonding angle composite correction amount includes an angle correction ratio, a dynamic correction parameter, and an angle adjustment reference value; the lens curvature dynamic correction amount includes a lateral offset, a longitudinal offset, and the center coordinates of the light spot; and the package body inclination global adaptation amount includes an inclination adjustment ratio, a package body inclination parameter, and a cumulative light spot offset.

3. The chip-level LED device packaging method according to claim 1, wherein: The specific steps of S1 are: S101: Collecting the surface topography data of the bonding point, obtaining the surface height distribution curve based on the three-dimensional profilometer, extracting the absolute value of the longitudinal coordinate to calculate the arithmetic mean height, identifying the vertical distance between the coordinates of the maximum concave point and the average height reference plane, analyzing the curvature radius of the coordinates of the extreme point of the profile curvature, and generating a morphological feature parameter set; S102: calling the morphological characteristic parameter set, combining the fracture mechanics stress field distribution equation with the crack propagation direction correction relationship, establishing a critical stress state equation, determining the critical stress condition through equilibrium point analysis, and generating a critical stress intensity factor parameter; S103: Based on the critical stress intensity factor parameter, combined with the material elastic modulus and Poisson's ratio, the fracture characteristics after material constant correction are determined through the fracture toughness calculation relationship under plane strain conditions, and a fracture toughness prediction value is generated.

4. The chip-level LED device packaging method according to claim 3, characterized in that: The specific steps of S2 are: S201: Obtaining the fracture toughness prediction value and a preset fracture toughness lower limit threshold, comparing the predicted value with the threshold, determining whether the fracture toughness prediction value is lower than the preset fracture toughness lower limit threshold, and generating a fracture toughness state determination value; S202: triggering a pressure compensation rule based on the fracture toughness state determination value, adjusting the bonding pressure reference by applying a compensation coefficient according to the compensation logic defined in the compensation rule and the current bonding pressure value, and generating a bonding pressure adjustment amount under the constraints of the compensation rule; S203: calling the bonding pressure adjustment amount, combining it with the current bonding pressure value, determining the compensated target bonding pressure value, extracting the deviation between the target value and the current value, and generating a dynamic bonding pressure compensation amount.

5. The chip-level LED device packaging method according to claim 4, characterized in that: The specific calculation formula for the deviation between the extracted target value and the current value is: ； Where ΔP represents the current bonding pressure deviation compensation amount, Pt represents the target bonding pressure value after compensation, and Pc represents the current bonding pressure value. represents the sliding average of historical pressure deviations, γ represents the temperature fluctuation compensation factor, λ represents the material deformation resistance coefficient, and η represents the process dynamic adaptation coefficient.

6. The chip-level LED device packaging method according to claim 4, characterized in that: The specific steps of S3 are: S301: obtaining a real-time value of the dynamic compensation amount of the bonding pressure, combining it with the bonding topography feedback data, extracting the topography offset features associated with the pressure compensation, analyzing and integrating the dynamic relationship between the features and the compensation amount, and generating a deformation offset; S302: applying the deformation offset to a fracture toughness assessment process to obtain a current fracture toughness value, comparing it with a preset fracture toughness upper threshold, and if the value exceeds the threshold, extracting the excess difference and marking it as a fracture toughness state identifier; S303: Based on the fracture toughness state identifier, the compensation adjustment calculation formula in the angle correction process is used, combined with the dynamic compensation amount, and processed according to the set ratio and reference value in the process to generate a composite correction amount for the bonding angle.

7. The chip-level LED device packaging method according to claim 6, characterized in that: The specific steps of S4 are: S401: Acquire the dynamic correction value of the bonding angle and the light intensity distribution data of the photoelectric detection array, calibrate the pixel coordinates output by the photoelectric detection array, construct an initial spatial distribution matrix of the light spot based on the calibrated pixel data and the measured distance data, and generate the initial spatial distribution matrix value of the light spot; S402: Calling the initial spatial distribution matrix value of the light spot, extracting the lateral displacement and the longitudinal displacement of the light spot center coordinate, separating the lateral offset and the longitudinal offset based on the extracted displacement data and the target preset position data, and generating a light spot center offset sequence; S403: Match the bonding angle dynamic correction value and the offset data according to the light spot center offset sequence, extract the corresponding curvature change, establish the overall lens curvature dynamic adjustment coefficient based on the correlation between the curvature change and the offset trend, and generate the lens curvature dynamic correction value.

8. The chip-level LED device packaging method according to claim 7, characterized in that: The specific calculation formula for separating the lateral offset and the longitudinal offset based on the extracted displacement data and the target preset position data is: ； in, Representative The lateral displacement of the center of the light spot extracted is Representative The longitudinal displacement of the center of the light spot extracted Represents the horizontal coordinate value of the preset target position, Represents the vertical coordinate value of the preset target position, represents the displacement product correction factor, represents the normalized adjustment factor, Representative The offset separation factor of the secondary separation process.

9. The chip-level LED device packaging method according to claim 7, wherein: The specific steps of S5 are: S501: Obtaining the dynamic correction value of the lens curvature and the luminous flux change value, collecting luminous flux dynamic response data according to the lens curvature change interval, combining the corresponding relationship between the curvature change and the luminous flux change, extracting the current luminous flux change characteristics, and generating the luminous flux change increase / decrease rate; S502: calling the luminous flux change increase / decrease rate, detecting the bonding point distribution density and the package geometric tilt angle change value, extracting the package tilt angle adjustment characteristics based on the corresponding relationship between the distribution density and the tilt angle change, and generating a package tilt angle adjustment ratio value; S503: calling the package tilt adjustment ratio value, synchronously monitoring the cumulative change of the light spot offset over time, establishing a tilt correction sequence according to the offset accumulation and the tilt adjustment characteristics, and generating a global package tilt adaptation value.

10. A chip-level LED device packaging system, characterized in that: According to a chip-level LED device packaging method according to any one of claims 1 to 9, the system comprises: The surface topography acquisition module acquires the surface topography data of the chip LED device bonding point, collects the surface roughness arithmetic mean height parameter, maximum concave depth parameter and profile curvature radius parameter, calculates the critical stress intensity factor value of the bonding point, and generates the fracture toughness prediction value based on the characteristic index set; The bonding pressure compensation module calls the fracture toughness prediction value, compares the numerical difference between the fracture toughness prediction value and the fracture toughness lower limit threshold value according to a preset fracture toughness lower limit threshold value, extracts the current bonding pressure parameter value, improves and adjusts the bonding pressure parameter according to the difference magnitude, and generates a dynamic bonding pressure compensation amount; The bonding angle correction module calls the bonding pressure dynamic compensation amount, extracts the bonding area topography feedback data, detects the fracture toughness change trend, extracts the existing bonding angle parameter value based on the extent by which the fracture toughness exceeds the preset upper limit threshold, adjusts the angle value according to the excess extent ratio, and generates a composite bonding angle correction amount; The light spot offset detection module calls the composite correction value of the bonding angle, collects the light intensity distribution data of the photoelectric detection array, detects the coordinate position of the center of the light spot, calculates the lateral and longitudinal offsets between the center of the light spot and the target position, analyzes the change of the lens morphology based on the trend of the offset data, and generates a dynamic correction value for the lens curvature; The package tilt adaptation module calls the dynamic correction value of the lens curvature, collects the current tilt parameter value of the package, extracts the bonding point distribution density data and the package geometric tilt data, analyzes the package tilt change trend based on the cumulative offset trend of the light spot and the current tilt parameter, calculates the tilt adjustment ratio, and generates the global adaptation value of the package tilt.

Images