Pin coplanarity intelligent leveling system and process in high-density QFN (Quad Flat No-lead) packaging

Abstract

The invention relates to the technical field of semiconductor packaging, in particular to an intelligent pin coplanarity leveling system and process in high-density QFN packaging, and the method comprises the steps: building a theoretical model based on element design data, obtaining multi-dimensional scanning data of a QFN element through a multi-dimensional scanner, building a physical simulation model through the combination of material physical attributes, and obtaining a QFN element coplanarity leveling model; the physical simulation model is compared with the theoretical model to obtain an initial difference sequence, an initial correction instruction is generated for the initial difference sequence through the correction decision intelligent model, simulation correction is conducted on the physical simulation model to obtain an optimized final correction instruction, and finally the final correction instruction is executed through the piezoelectric actuator. And whether the element is qualified or not is detected circularly. According to the invention, the coplanarity detection and repair efficiency, accuracy and intelligence of the QFN element pins can be improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of semiconductor packaging, in particular to a high-density QFN packaging pin coplanarity intelligent leveling system and process. BACKGROUND

[0002] In today's electronic manufacturing industry, QFN technology is widely used in various fields such as smart phones and wearable devices due to its small size, light weight, superior heat dissipation performance and electrical performance. In QFN technology, pin coplanarity is a key factor that determines the quality of subsequent welding. A small coplanarity deviation can cause serious defects such as virtual welding and open circuit, which directly damages the performance and reliability of the final product. Therefore, accurate detection and repair of QFN component pin coplanarity is a core technical link to ensure electronic manufacturing yield and enterprise market competitiveness.

[0003] Traditional detection methods mainly rely on manual detection and manual repair. However, this method is not only inefficient and costly, but also easily affected by the subjective experience and fatigue state of the operator, with poor consistency and reliability, which can easily cause secondary damage to fragile pins or even cause the component to be directly scrapped. Therefore, the efficiency, accuracy and intelligence of this method need to be improved. SUMMARY

[0004] The application provides a high-density QFN packaging pin coplanarity intelligent leveling system and process, which mainly aims to improve the efficiency, accuracy and intelligence of QFN component pin coplanarity detection and repair.

[0005] To achieve the above purpose, the application provides a high-density QFN packaging pin coplanarity intelligent leveling process, which comprises:

[0006] Obtaining a QFN component, component design data, a precision clamp, a multi-dimensional scanner, a correction decision intelligent model and a piezoelectric actuator, wherein the multi-dimensional scanner comprises a three-dimensional scanner and a high-definition camera;

[0007] Based on the direction mark pre-constructed in the QFN component and the precision clamp, the QFN component is fixed to obtain a positioned component, wherein the positioned component comprises a plurality of component pins;

[0008] Based on the component design data and the pre-constructed simulation software, a theoretical model is obtained, wherein the theoretical model comprises a plurality of theoretical model pins, and the theoretical model pins correspond one-to-one to the component pins;

[0009] The multi-dimensional scanner is used to perform multi-dimensional scanning operation on the positioned component to obtain actual scanning data;

[0010] Construct a physical simulation model based on actual scanning data and simulation software, and obtain an initial difference sequence based on a theoretical model and the physical simulation model;

[0011] Perform a coplanarity judgment operation based on the initial difference sequence and a preset allowable deviation parameter to obtain a coplanarity judgment result;

[0012] If the coplanarity judgment result is a preset coplanarity abnormality, confirm the positioned component as a component to be corrected, and obtain an initial correction instruction based on the initial difference sequence and a correction decision intelligent model;

[0013] Perform a simulation correction operation on the physical simulation model based on the initial correction instruction to obtain a simulation correction difference sequence, and obtain a final correction instruction based on the simulation correction difference sequence and the correction decision intelligent model;

[0014] Perform a pin correction operation on the component to be corrected based on a piezoelectric actuator and the final correction instruction to obtain a leveled component, and return to the step of performing a multi-dimensional scanning operation on the positioned component by using the multi-dimensional scanner, with the leveled component as the positioned component;

[0015] If the coplanarity judgment result is a preset coplanarity normality, confirm the positioned component as a qualified component, and complete a pin coplanarity intelligent leveling process in high-density QFN packaging based on the qualified component.

[0016] Optionally, the multi-dimensional scanning operation on the positioned component by using the multi-dimensional scanner to obtain actual scanning data comprises:

[0017] Perform a three-dimensional scanning operation on the positioned component by using a three-dimensional scanner to obtain original point cloud data;

[0018] Perform a point cloud noise reduction operation on the original point cloud data by using a pre-constructed point cloud noise reduction algorithm to obtain three-dimensional point cloud data;

[0019] Capture a high-definition image of the positioned component by using a high-definition camera;

[0020] Perform a grayscale conversion operation on the high-definition image by using a pre-constructed grayscale conversion algorithm to obtain a grayscale image;

[0021] Perform a contrast enhancement operation on the grayscale image by using a pre-constructed contrast enhancement algorithm to obtain an enhanced image;

[0022] Perform an image noise reduction operation on the enhanced image by using a pre-constructed image noise reduction algorithm to obtain a noise-reduced image;

[0023] Perform an edge detection operation on the noise-reduced image based on a pre-constructed segmentation threshold algorithm and a pre-constructed edge detection algorithm to obtain an edge detection image;

[0024] The three-dimensional point cloud data and the edge detection image are aggregated to obtain actual scanning data.

[0025] Optionally, the actual scanning data and the simulation software are used to construct a physical simulation model, including:

[0026] An initial three-dimensional model is constructed based on the simulation software and the three-dimensional point cloud data in the actual scanning data.

[0027] A mapping relationship matrix is obtained, and the edge detection image in the actual scanning data is mapped to the initial three-dimensional model based on the mapping relationship matrix to obtain a textured three-dimensional model.

[0028] Material physical properties in the component design data are extracted, and a physical simulation model is constructed based on the material physical properties and the textured three-dimensional model, wherein the physical simulation model includes a plurality of physical simulation model pins, and the physical simulation model pins correspond one-to-one to the component pins.

[0029] Optionally, the mapping relationship matrix is obtained, including:

[0030] A three-dimensional corner point coordinate set is obtained based on a pre-constructed corner point detection algorithm and the three-dimensional point cloud data.

[0031] A two-dimensional corner point coordinate set is obtained based on the corner point detection algorithm and the edge detection image.

[0032] A preliminary matching relationship matrix in the multi-dimensional scanner is extracted, and the two-dimensional corner point coordinate set is mapped to the initial three-dimensional model based on the preliminary matching relationship matrix to obtain a mapped corner point coordinate set.

[0033] A matrix adjustment parameter is obtained based on the three-dimensional corner point coordinate set, the mapped corner point coordinate set, and a pre-constructed nearest point registration algorithm.

[0034] The mapping relationship matrix is obtained based on the preliminary matching relationship matrix and the matrix adjustment parameter.

[0035] Optionally, the initial difference sequence is obtained based on the theoretical model and the physical simulation model, including:

[0036] A spatial registration operation is performed on the theoretical model and the physical simulation model to obtain a merged model, wherein the merged model includes a theoretical model layer and a physical simulation model layer.

[0037] A theoretical model pin point cloud sequence is obtained based on a theoretical model pin in the theoretical model layer, and a physical simulation model pin point cloud sequence is obtained based on a physical simulation model pin in the physical simulation model layer, wherein the theoretical model pin point cloud sequence includes a plurality of theoretical model pin point clouds, the theoretical model pin point clouds correspond one-to-one to the theoretical model pins, the physical simulation model pin point cloud sequence includes a plurality of physical simulation model pin point clouds, and the physical simulation model pin point clouds correspond one-to-one to the physical simulation model pins.

[0038] obtaining a theoretical model pin plane normal vector sequence and a physical simulation model pin plane normal vector sequence based on the theoretical model pin plane sequence and the physical simulation model pin plane sequence;

[0039] obtaining a theoretical model pin plane normal vector sequence and a physical simulation model pin plane normal vector sequence based on the theoretical model pin plane sequence and the physical simulation model pin plane sequence;

[0040] obtaining a theoretical model pin plane normal vector sequence and a physical simulation model pin plane normal vector sequence based on the theoretical model pin plane sequence and the physical simulation model pin plane sequence;

[0041] obtaining a theoretical model pin plane normal vector sequence and a physical simulation model pin plane normal vector sequence based on the theoretical model pin plane sequence and the physical simulation model pin plane sequence;

[0042] obtaining a theoretical model pin plane normal vector sequence and a physical simulation model pin plane normal vector sequence based on the theoretical model pin plane sequence and the physical simulation model pin plane sequence;

[0043] obtaining a theoretical model pin plane normal vector sequence and a physical simulation model pin plane normal vector sequence based on the theoretical model pin plane sequence and the physical simulation model pin plane sequence;

[0044] obtaining a theoretical model pin plane normal vector sequence and a physical simulation model pin plane normal vector sequence based on the theoretical model pin plane sequence and the physical simulation model pin plane sequence;

[0045] obtaining a theoretical model pin plane normal vector sequence and a physical simulation model pin plane normal vector sequence based on the theoretical model pin plane sequence and the physical simulation model pin plane sequence;

[0046] Optionally, based on the initial difference sequence and a preset allowable deviation parameter, a coplanarity judgment operation is performed to obtain a coplanarity judgment result, including:

[0047] For each pin of the positioned components, the following operations are performed:

[0048] extracting the pin plane angle difference and the pin distance difference corresponding to the pin in the initial difference sequence;

[0049] extracting a distance allowable deviation range in the allowable deviation parameter, if the pin distance difference is located in the distance allowable deviation range, marking the pin as a distance difference normal, otherwise, marking the pin as a distance difference abnormal;

[0050] extracting an angle allowable deviation value in the allowable deviation parameter, if the pin plane angle difference is less than or equal to the angle allowable deviation value, marking the pin as an angle difference normal, otherwise, marking the pin as an angle difference abnormal;

[0051] If the distance difference and angle difference of the component pin are normal, mark the component pin as a normal pin; if the distance difference or angle difference of the component pin is abnormal, mark the component pin as an abnormal pin.

[0052] If every component pin in the located components is a normal pin, the coplanarity judgment result is confirmed as normal. If there is one component pin in the located components that is an abnormal pin, the coplanarity judgment result is confirmed as abnormal.

[0053] Optionally, obtaining the initial correction instruction based on the initial difference sequence and the intelligent correction decision model includes:

[0054] Extract the pin data of all abnormal pins in the initial difference sequence to obtain the pin dataset to be corrected. The pin data includes: pin distance difference, pin plane normal vector of physical simulation model and pin plane angle difference.

[0055] The material resilience rate is obtained based on the intelligent model of corrective decision-making and the physical properties of the material.

[0056] For each pin data in the pin dataset to be calibrated, perform the following operation:

[0057] The target correction amount is obtained based on the pin distance difference, the pin plane normal vector of the physical simulation model, and the pin plane angle difference.

[0058] The actual correction amount is obtained based on the target correction amount and the material resilience.

[0059] Based on the actual calibration amount and the intelligent model for calibration decision, a single pin calibration command is obtained, which includes the calibration point coordinates and calibration pressure.

[0060] Summarize the individual pin calibration instructions to obtain a single pin calibration instruction set, and obtain the initial calibration instruction based on the single pin calibration instruction set.

[0061] Optionally, the step of performing simulated calibration operations on the physical simulation model based on the initial calibration command to obtain a simulated calibration difference sequence, and obtaining the final calibration command based on the simulated calibration difference sequence and the calibration decision intelligent model, includes:

[0062] A simulated piezoelectric actuator was constructed based on simulation software.

[0063] By using a simulated piezoelectric actuator and initial calibration instructions, a simulated pin calibration operation is performed on each abnormal pin in the physical simulation model to obtain a simulated calibration element model.

[0064] Based on the theoretical model and the simulated correction element model, the simulated correction difference sequence is obtained;

[0065] Based on the simulated correction difference sequence and the intelligent correction decision model, the command correction parameters are obtained;

[0066] Based on the initial calibration command and command correction parameters, the final calibration command is obtained.

[0067] Optionally, the step of performing pin calibration on the element to be calibrated based on the piezoelectric actuator and the final calibration command to obtain a leveled element includes:

[0068] For each faulty pin in the component to be calibrated, perform the following operation:

[0069] Extract the final single-pin calibration instruction corresponding to the abnormal pin from the final calibration instruction. The final single-pin calibration instruction includes: the final calibration point coordinates and the final calibration pressure.

[0070] Using a pre-built multi-axis motion platform, the piezoelectric actuator is moved to the final calibration point coordinates to obtain the piezoelectric actuator to be started;

[0071] Based on the final calibration pressure and the preset voltage-pressure conversion equation, the initial driving voltage is obtained. The initial driving voltage is used to perform a single-pin calibration operation on the piezoelectric actuator to be started, resulting in a calibrated piezoelectric actuator. The pressure of the calibrated piezoelectric actuator is then monitored to obtain the actual pressure.

[0072] The pressure difference ratio is obtained based on the actual pressure and the corrected pressure.

[0073] If the pressure difference ratio is not within the preset allowable pressure difference ratio range, the initial drive voltage is adjusted based on the pressure difference ratio to obtain a new drive voltage. The new drive voltage is then used as the initial drive voltage, and the process returns to the step of performing a single pin calibration operation on the piezoelectric actuator to be started using the initial drive voltage.

[0074] If the pressure difference ratio is within the preset allowable pressure difference ratio range, the single pin calibration operation is completed based on the preset calibration pressure action time, resulting in a leveled component.

[0075] To achieve the above objectives, the present invention also provides a smart leveling system for pin coplanarity in a high-density QFN package, comprising:

[0076] The basic data acquisition module is used to acquire QFN components, component design data, precision fixtures, multi-dimensional scanners, intelligent calibration decision models, and piezoelectric actuators. The multi-dimensional scanner includes a 3D scanner and a high-definition camera. Based on the pre-built orientation marks in the QFN component and the precision fixture, the QFN component is fixed to obtain a positioned component. The positioned component includes multiple component pins. Based on the component design data and pre-built simulation software, a theoretical model is obtained. The theoretical model includes multiple theoretical model pins, which correspond one-to-one with the component pins. The multi-dimensional scanner is used to perform multi-dimensional scanning operations on the positioned component to obtain actual scan data.

[0077] The pin coplanarity analysis module is used to construct a physical simulation model based on actual scanning data and simulation software, obtain an initial difference sequence based on the theoretical model and the physical simulation model, and perform a coplanarity judgment operation based on the initial difference sequence and preset allowable deviation parameters to obtain the coplanarity judgment result.

[0078] The intelligent decision-making and simulation correction module is used to identify the located component as the component to be corrected if the coplanarity judgment result is a preset coplanarity anomaly. Based on the initial difference sequence and the intelligent correction decision model, it obtains the initial correction instruction, performs a simulation correction operation on the physical simulation model based on the initial correction instruction, obtains the simulation correction difference sequence, and obtains the final correction instruction based on the simulation correction difference sequence and the intelligent correction decision model.

[0079] The pin calibration and output module is used to perform pin calibration operations on the component to be calibrated based on the piezoelectric actuator and the final calibration command, to obtain a leveled component. The leveled component is then used as the positioned component. The module returns to the step of performing multi-dimensional scanning operation on the positioned component using a multi-dimensional scanner. If the coplanarity judgment result is the preset coplanarity normal, the positioned component is confirmed as a qualified component. Based on the qualified component, the intelligent pin coplanarity leveling process in the high-density QFN package is completed.

[0080] To address the above problems, the present invention also provides an electronic device, the electronic device comprising:

[0081] Memory, storing at least one instruction;

[0082] The processor executes instructions stored in the memory to implement the intelligent pin coplanarity leveling process in the high-density QFN package described above.

[0083] To address the aforementioned issues, the present invention also provides a computer-readable storage medium storing at least one instruction, which is executed by a processor in an electronic device to implement the aforementioned intelligent leveling process for pin coplanarity in a high-density QFN package.

[0084] To address the problems described in the background art, this invention acquires a QFN element, element design data, a precision fixture, a multi-dimensional scanner, a calibration decision intelligent model, and a piezoelectric actuator. The multi-dimensional scanner includes a 3D scanner and a high-definition camera. This invention, by integrating precision hardware and intelligent software, provides a complete and necessary technical foundation for achieving high-precision automated detection and calibration. Based on pre-constructed orientation markers and a precision fixture within the QFN element, the QFN element is fixed to obtain a positioned element. The positioned element includes multiple element pins. This invention ensures the consistency of the element's position during scanning and repair, eliminates interference from positioning errors on measurement results, and guarantees data accuracy and subsequent calibration precision. Using component design data and pre-built simulation software, a theoretical model is obtained. This theoretical model includes multiple theoretical model pins, each corresponding one-to-one with a component pin. This demonstrates that the invention establishes a precise and error-free ideal benchmark model, providing a reliable reference standard for subsequent calculations of the actual physical deviations of each pin. A multi-dimensional scanner is used to perform multi-dimensional scanning operations on the located component, obtaining actual scan data. This invention employs a non-contact measurement method, enabling rapid and non-destructive acquisition of complete and realistic three-dimensional spatial contours and high-definition visual data of the component pins, providing comprehensive data input for accurate modeling. Based on the actual scan data and simulation software, a physical simulation model is constructed. Based on the theoretical model and the physical simulation model, an initial difference sequence is obtained. As can be seen, this invention can transform actual scan data into an analyzable digital model and quantify the deviation value of each pin from the ideal position, providing data basis for subsequent qualification judgment and correction instruction generation. Based on the initial difference sequence and preset allowable deviation parameters, a coplanarity judgment operation is performed to obtain the coplanarity judgment result. This invention achieves automated and standardized judgment of component pin coplanarity, improving detection efficiency and intelligence. If the coplanarity judgment result is a preset coplanarity anomaly, the located component is confirmed as a component to be corrected. Based on the initial difference sequence and the intelligent correction decision model, an initial correction instruction is obtained. This invention utilizes an artificial intelligence model to generate a targeted preliminary correction scheme based on deviation data, achieving… The intelligent calibration decision-making process involves performing simulated calibration operations on the physical simulation model based on initial calibration instructions to obtain a simulated calibration difference sequence. Based on this sequence and the intelligent calibration decision-making model, the final calibration instruction is obtained. This invention improves the calibration success rate and prevents component damage caused by calibration by pre-simulating and optimizing the calibration instructions in a virtual environment. Based on the piezoelectric actuator and the final calibration instruction, pin calibration operations are performed on the component to be calibrated to obtain a leveled component. Using this leveled component as the positioned component, the process returns to the step of performing a multi-dimensional scan of the positioned component using a multi-dimensional scanner. This invention executes optimized instructions using a high-precision actuator and ensures that the coplanarity of the component pins meets the acceptable standard through an iterative approach.This invention ensures high-quality final products. If the coplanarity assessment result is within the preset range, the positioned component is confirmed as a qualified component. Based on the qualified component, the intelligent leveling process for pin coplanarity in high-density QFN packaging is completed. Therefore, this invention outputs components with normal coplanarity as qualified components, thus completing the intelligent leveling process for pin coplanarity in high-density QFN packaging. Consequently, this invention improves the efficiency, accuracy, and intelligence of QFN component pin coplanarity detection and repair. Attached Figure Description

[0085] Figure 1 This is a schematic flowchart of the intelligent pin coplanarity leveling process in a high-density QFN package provided in an embodiment of the present invention;

[0086] Figure 2 This is a functional block diagram of a pin coplanarity intelligent leveling system in a high-density QFN package provided in an embodiment of the present invention;

[0087] Figure 3 This is a schematic diagram of an electronic device that implements the intelligent leveling process for pin coplanarity in the high-density QFN package, according to an embodiment of the present invention.

[0088] Explanation of reference numerals in the attached figures:

[0089] 10. Electronic device; 11. Processor; 12. Memory; 13. Bus.

[0090] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0091] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0092] This application provides a pin coplanarity intelligent leveling process in a high-density QFN package. The executing entity for this high-density QFN package pin coplanarity intelligent leveling process includes, but is not limited to, at least one of the following electronic devices that can be assigned to execute the process provided in this application: a server, a terminal, etc. In other words, the high-density QFN package pin coplanarity intelligent leveling process can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0093] Reference Figure 1 The diagram shown is a flowchart illustrating the intelligent pin coplanarity leveling process in a high-density QFN package according to an embodiment of the present invention. In this embodiment, the intelligent pin coplanarity leveling process in the high-density QFN package includes:

[0094] S1. Acquire QFN components, component design data, precision fixtures, multi-dimensional scanners, calibration decision-making intelligent models, and piezoelectric actuators. The multi-dimensional scanners include: a 3D scanner and a high-definition camera.

[0095] Understandably, the QFN component refers to a packaged component of an integrated circuit; QFN stands for Quad Flat No-Leader Package. The component design data refers to the design data of the QFN component during the design phase, containing theoretical information on all geometric and physical specifications, including dimensions, shape, pin positions, and materials. The precision fixture refers to a device used to fix the QFN component in a precise and known position, ensuring operational stability and accuracy. The multi-dimensional scanner refers to a device capable of performing 3D and 2D scanning of objects, consisting of a 3D scanner and a high-definition camera. The 3D scanner refers to a device that acquires spatial data of an object using methods such as lasers. The high-definition camera refers to a camera capable of capturing high-resolution images, used to capture surface images of the QFN component. The intelligent correction decision model refers to a model based on artificial intelligence algorithms (such as deep neural networks, gradient boosting decision trees, etc.) that can intelligently calculate how to correct the currently detected pin deviation data based on a large amount of historical experience with pin leveling. The piezoelectric actuator is a device that converts electrical energy into pressure by utilizing the piezoelectric effect (applying voltage to a specific material to cause deformation), enabling precise pressure control for accurately pushing and calibrating the pins of a QFN element.

[0096] S2. Based on the pre-built orientation markers and precision fixtures in the QFN element, fix the QFN element to obtain the positioned element, wherein the positioned element includes multiple element pins.

[0097] It should be explained that the orientation mark refers to a physical mark pre-installed on the body of the QFN component during manufacturing to identify its orientation and reference position. The positioned component refers to the QFN component that has been accurately placed and locked in a precision fixture according to the orientation mark. The component pins refer to the metal contacts on the bottom edge of the QFN component used for electrical connection with the circuit board.

[0098] S3. Based on component design data and pre-built simulation software, obtain a theoretical model, which includes multiple theoretical model pins, and the theoretical model pins correspond one-to-one with the component pins.

[0099] As is clear, the simulation software refers to computer software capable of creating virtual models and simulating physical behavior based on input data (such as component design data). Examples include engineering analysis software such as ANSYS and ABAQUS. The theoretical model refers to a three-dimensional digital model representing a defect-free QFN component, created in the simulation software based on the component design data. The theoretical model pins refer to the pins on the theoretical model that correspond to the pins of the actual QFN component.

[0100] S4. Use a multi-dimensional scanner to perform multi-dimensional scanning on the positioned component to obtain actual scan data.

[0101] Understandably, the multi-dimensional scanning operation refers to the operation of collecting information from multiple aspects of the located element, including the acquisition of three-dimensional data and two-dimensional data. The actual scan data refers to the set of data reflecting the true physical state of the located element in the real world, collected through the multi-dimensional scanning operation, including three-dimensional point cloud data and edge detection images.

[0102] Furthermore, the step of using a multi-dimensional scanner to perform a multi-dimensional scanning operation on the located element to obtain actual scan data includes:

[0103] The original point cloud data is obtained by performing a 3D scanning operation on the located components using a 3D scanner.

[0104] Using a pre-built point cloud denoising algorithm, point cloud denoising is performed on the original point cloud data to obtain three-dimensional point cloud data;

[0105] High-definition images of the positioned components are captured using a high-definition camera;

[0106] A pre-built grayscale conversion algorithm is used to perform grayscale conversion on high-definition images to obtain grayscale images;

[0107] A pre-built contrast enhancement algorithm is used to enhance the contrast of a grayscale image, resulting in an enhanced image.

[0108] A pre-built image denoising algorithm is used to perform image denoising on the enhanced image to obtain a denoised image;

[0109] Based on a pre-constructed segmentation threshold algorithm and a pre-constructed edge detection algorithm, edge detection is performed on the denoised image to obtain an edge-detected image;

[0110] By combining 3D point cloud data and edge detection images, the actual scan data is obtained.

[0111] As is clear, the 3D scanning operation refers to the operation of acquiring 3D topographic data of an object's surface using a 3D scanner. The raw point cloud data refers to the set of 3D coordinate points output by the 3D scanner after directly scanning the located element, without any processing. The point cloud denoising algorithm refers to an algorithm used to remove noise from the data, such as statistical filtering, radius filtering, and Gaussian filtering. The point cloud denoising operation refers to the operation of automatically identifying and removing invalid or erroneous data points (i.e., noise) that may exist in the raw point cloud data due to system errors or other reasons, thereby improving data quality. The 3D point cloud data refers to the set of coordinate points that more accurately reflects the true 3D shape of the located element after point cloud denoising processing. The high-definition image refers to a high-resolution photograph of the located element taken by a high-definition camera. The grayscale conversion algorithm refers to an algorithm used to convert a color high-definition image into a black-and-white grayscale image, and this grayscale conversion algorithm is an existing technology, which will not be elaborated further here.

[0112] Understandably, the grayscale conversion operation refers to the operation of converting a high-definition image containing color information into a grayscale image without color information using a grayscale conversion algorithm. A grayscale image is an image that does not contain color information and is represented only by grayscale values, which simplifies subsequent computer analysis. The contrast enhancement algorithm is an image processing technique whose purpose is to amplify the grayscale difference between the brightest and darkest areas of an image, making the boundary between component pins and the package body more distinct. For example, commonly used contrast enhancement techniques include histogram equalization and linear stretching. The package body refers to the black square plastic part on a QFN component used to enclose the internal silicon chip. The contrast enhancement operation refers to the operation of enhancing the details and sense of depth in a grayscale image using a contrast enhancement algorithm. The enhanced image is the image obtained after contrast enhancement processing of a grayscale image. The image denoising algorithm is an algorithm used to eliminate or reduce interfering pixels (i.e., image noise) in an enhanced image, such as median filtering and bilateral filtering. The image denoising operation refers to the operation of removing noise that may exist in an enhanced image using an image denoising algorithm. The denoised image is the image obtained after image denoising processing. The segmentation thresholding algorithm is an algorithm used in edge detection processes. Its output is a threshold used to distinguish whether a pixel is an edge, such as the Otsu algorithm. The edge detection algorithm refers to an algorithm used to identify and mark the contours (i.e., edges) of objects in an image, such as the Canny operator and the Sobel operator. The edge detection operation refers to the operation of extracting the contour lines (and edges) of objects in an image using the segmentation thresholding algorithm and the edge detection algorithm. The edge-detected image refers to the output image generated after applying the edge detection algorithm to a denoised image, represented as a black background with white lines, where the white lines represent detected edges.

[0113] S5. Based on actual scanning data and simulation software, construct a physical simulation model, and based on the theoretical model and physical simulation model, obtain the initial difference sequence.

[0114] It should be explained that the physical simulation model refers to a three-dimensional digital model that accurately reproduces the geometric state and physical properties of the positioned component in the real world. The initial difference sequence refers to the set of different types of deviation values ​​calculated after comparing the theoretical model with the physical simulation model, including the pin plane angle difference sequence and the pin distance difference sequence.

[0115] Furthermore, the construction of a physical simulation model based on actual scanning data and simulation software includes:

[0116] An initial 3D model is constructed based on 3D point cloud data from simulation software and actual scanning data.

[0117] Obtain the mapping relationship matrix, and based on the mapping relationship matrix, map the edge detection images in the actual scan data to the initial 3D model to obtain the textured 3D model;

[0118] Material physical properties are extracted from component design data. Based on the material physical properties and textured 3D model, a physical simulation model is constructed. The physical simulation model includes multiple physical simulation model pins, and each physical simulation model pin corresponds one-to-one with a component pin.

[0119] It should be explained that the initial 3D model refers to a 3D model constructed in simulation software using only denoised 3D point cloud data, which only reflects the geometry of the located components. The mapping matrix refers to the mathematical transformation matrix that shows how each pixel in the edge detection image corresponds to a spatial coordinate point on the surface of the initial 3D model.

[0120] In detail, the textured 3D model refers to the model obtained by accurately mapping the edge detection image to the initial 3D model through a mapping relationship matrix. The material physical properties refer to the parameters defined in the component design data that describe the various materials (such as package plastic and lead metal) constituting the QFN component. The physical simulation model pins refer to the pins on the physical simulation model that correspond to the pins of the actual QFN component.

[0121] Furthermore, obtaining the mapping matrix includes:

[0122] Based on a pre-built corner detection algorithm and 3D point cloud data, obtain a set of 3D corner coordinates;

[0123] Based on corner detection algorithms and edge detection images, a two-dimensional set of corner coordinates is obtained;

[0124] Extract the preliminary matching relationship matrix from the multidimensional scanner, and map the two-dimensional corner coordinate set to the initial three-dimensional model based on the preliminary matching relationship matrix to obtain the mapped corner coordinate set;

[0125] Matrix adjustment parameters are obtained based on the 3D corner coordinate set, the mapped corner coordinate set, and the pre-constructed nearest point registration algorithm;

[0126] Based on the initial matching relationship matrix and matrix adjustment parameters, the mapping relationship matrix is ​​obtained.

[0127] This can be explained as follows: the corner detection algorithm refers to an algorithm used to automatically identify and locate corner points (i.e., points where the direction of an edge or surface changes abruptly) in a geometric shape. For example, for two-dimensional images, Harris corner detection is typically used; for three-dimensional point clouds, corner points are usually determined by analyzing local curvature changes. The three-dimensional corner coordinate set refers to the set of three-dimensional spatial coordinates corresponding to corner points found in three-dimensional point cloud data using the corner detection algorithm. The two-dimensional corner coordinate set refers to the set of two-dimensional planar coordinates corresponding to corner points found in an edge detection image (i.e., an edge detection image) using the corner detection algorithm. The preliminary matching matrix refers to the transformation matrix determined and stored during the initial setup of the multi-dimensional scanner based on the relative positional relationship between the 3D scanner and the high-definition camera. Its purpose is to describe the correspondence between the 2D image plane captured by the camera and the space scanned by the 3D scanner. Since deviations in the relative positions between the 3D scanner and the high-definition camera cannot be completely avoided in actual use, the preliminary matching matrix is ​​only used to determine a rough mapping relationship and still needs further refinement. The preliminary matching matrix has a three-row, four-column structure, where the first three columns collectively represent the scaling factor and rotation direction of the mapping, and the data in the fourth column represent the pixel points along... axis, axis, The translation distance of the axis, for example, the coordinates of a pixel in a two-dimensional image are... The preliminary matching relationship matrix is ​​shown below:

[0128]

[0129] The mapped 3D coordinates are shown below:

[0130]

[0131]

[0132]

[0133] That is, coordinate points in a two-dimensional image The mapped 3D coordinates are .

[0134] It is clear that the mapped corner coordinate set refers to a new set of three-dimensional coordinate points obtained by projecting each two-dimensional coordinate point in the two-dimensional corner coordinate set onto the space of the initial three-dimensional model through mathematical transformation using a preliminary matching relation matrix. The nearest point registration algorithm refers to an algorithm used to find an optimal rigid transformation matrix (rotation and translation) that minimizes the sum of distances between corresponding points in the three-dimensional corner coordinate set and the mapped corner coordinate set, such as the iterative nearest point algorithm. The matrix adjustment parameters refer to a set of specific transformation values ​​calculated after running the nearest point registration algorithm, including: along... The distance the axis needs to be adjusted, along The distance the axis needs to be adjusted, along The distance and winding of the axis need to be adjusted. The angle and rotation of the shaft need to be adjusted. The angle and rotation of the shaft need to be adjusted. The angle that the axis needs to be adjusted is used to describe how to perform a rigid transformation on the mapped corner coordinate set so that the mapped corner coordinate set coincides with the three-dimensional corner coordinate set.

[0135] Specifically, obtaining the mapping relationship matrix based on the preliminary matching relationship matrix and matrix adjustment parameters means:

[0136] The distances that need to be adjusted along the X-axis, Y-axis, and Z-axis are constructed into a 3x4 matrix (called the atomic matrix), as shown below:

[0137]

[0138] in, Represents an atomic matrix. Indicates along The distance the axis needs to be adjusted. Indicates along The distance the axis needs to be adjusted. Indicates along The distance the axis needs to be adjusted.

[0139] Will around The angle that the axis needs to be adjusted is constructed into a three-row, four-column matrix (called...). (axis rotation matrix) The axis rotation matrix is ​​shown below:

[0140]

[0141] in, express Axis rotation matrix, Indicates circling The angle that the shaft needs to be adjusted Represents the cosine function. This represents the sine function.

[0142] Will around The angle that the axis needs to be adjusted is constructed into a three-row, four-column matrix (called...). (axis rotation matrix) The axis rotation matrix is ​​shown below:

[0143]

[0144] in, express Axis rotation matrix, Indicates circling The angle that the shaft needs to be adjusted.

[0145] Will around The angle that the axis needs to be adjusted is constructed into a three-row, four-column matrix (called...). (axis rotation matrix) The axis rotation matrix is ​​shown below:

[0146]

[0147] in, express Axis rotation matrix, Indicates circling The angle that the shaft needs to be adjusted.

[0148] The method for calculating the mapping matrix is ​​as follows:

[0149]

[0150] in, Represents the mapping matrix, This represents the initial matching relationship matrix.

[0151] Furthermore, obtaining the initial difference sequence based on the theoretical model and physical simulation model includes:

[0152] Spatial registration is performed on the theoretical model and the physical simulation model to obtain a merged model, which includes a theoretical model layer and a physical simulation model layer.

[0153] The theoretical model pin point cloud sequence is obtained based on the theoretical model pins in the theoretical model layer, and the physical simulation model pin point cloud sequence is obtained based on the physical simulation model pins in the physical simulation model layer. The theoretical model pin point cloud sequence includes multiple theoretical model pin point clouds, and each theoretical model pin point cloud corresponds to a theoretical model pin. The physical simulation model pin point cloud sequence includes multiple physical simulation model pin point clouds, and each physical simulation model pin point cloud corresponds to a physical simulation model pin.

[0154] Based on the pin point cloud sequence of the theoretical model and a pre-built plane fitting algorithm, the pin plane equation sequence of the theoretical model is obtained.

[0155] Based on the pin point cloud sequence of the physical simulation model and the plane fitting algorithm, the pin plane equation sequence of the physical simulation model is obtained.

[0156] Based on the pin plane equation sequence of the theoretical model and the pin plane equation sequence of the physical simulation model, calculate the pin plane normal vector sequence of the theoretical model and the pin plane normal vector sequence of the physical simulation model.

[0157] The pin plane angle difference sequence is obtained based on the pin plane normal vector sequence of the theoretical model and the pin plane normal vector sequence of the physical simulation model. The pin plane angle difference sequence includes multiple pin plane angle differences, and the pin plane angle differences correspond one-to-one with the component pins.

[0158] Based on the pin point cloud sequence of the theoretical model, obtain the coordinate sequence of the pin center point of the theoretical model.

[0159] Based on the pin point cloud sequence of the physical simulation model, obtain the coordinate sequence of the pin center point of the physical simulation model;

[0160] Based on the pin center point coordinate sequence of the theoretical model and the pin center point coordinate sequence of the physical simulation model, the pin distance difference sequence is obtained. The pin distance difference sequence includes multiple pin distance differences, and each pin distance difference corresponds one-to-one with a component pin.

[0161] The initial difference sequence is obtained by summing the pin plane angle difference sequence and the pin distance difference sequence.

[0162] In detail, the spatial registration operation refers to an operation that aligns the theoretical model and the physical simulation model to a unified coordinate system through transformations such as rotation and translation. The merged model refers to the model obtained in the simulation software after spatial registration of the theoretical model and the physical simulation model, containing two model layers. The theoretical model layer refers to the part belonging to the theoretical model in the merged model. The physical simulation model layer refers to the part belonging to the physical simulation model in the merged model. The theoretical model pin point cloud sequence refers to the set of three-dimensional coordinates of all data points extracted from the surface of each theoretical model pin in the theoretical model layer. The physical simulation model pin point cloud sequence refers to the set of three-dimensional coordinates of all data points extracted from the surface of each physical simulation model pin in the physical simulation model layer. The plane fitting algorithm refers to an algorithm that can calculate a plane that best fits a given set of three-dimensional spatial points (point cloud), such as the least squares method. The theoretical model pin plane equation sequence refers to the set of all plane equations obtained by applying the plane fitting algorithm to the point cloud data of each pin in the theoretical model. The sequence of pin plane equations in the physical simulation model refers to the set of mathematical equations for a series of planes obtained by applying a plane fitting algorithm to the point cloud data of each pin in the physical simulation model. The sequence of pin plane normal vectors in the theoretical model refers to the set of pin plane normal vectors in the theoretical model. The pin plane normal vectors in the theoretical model are the normal vectors of the pin planes in the theoretical model calculated based on the pin plane equations.

[0163] Specifically, the sequence of pin plane normal vectors in the physical simulation model refers to the set of pin plane normal vectors in the physical simulation model. The pin plane normal vector is the normal vector of the pin plane in the physical simulation model calculated based on the pin plane equation. The pin plane angle difference is the angle calculated between the theoretical model pin plane normal vector and the corresponding physical simulation model pin plane normal vector, where the angle is typically calculated using the vector dot product formula. The pin plane angle difference sequence is the set of pin plane angle differences. The sequence of theoretical model pin center point coordinates is the set of theoretical model pin center point coordinates, where the theoretical model pin center point coordinates are the three-dimensional coordinates obtained by calculating the geometric center of the theoretical model pin point cloud. The sequence of physical simulation model pin center point coordinates is a set of three-dimensional coordinates representing the actual center position of each pin, obtained by calculating the geometric center of the point cloud data of each physical simulation model pin. The pin distance difference is the straight-line distance between the calculated theoretical model pin center point coordinates and the corresponding physical simulation model pin center point coordinates, representing the offset of the pin center point relative to its ideal position, where the ideal position is the position of the geometric center of the corresponding pin in the theoretical model. The pin distance difference sequence refers to the set of pin distance differences.

[0164] S6. Based on the initial difference sequence and the preset allowable deviation parameters, perform the coplanarity judgment operation to obtain the coplanarity judgment result.

[0165] It is clear that the allowable deviation parameters refer to a set of pre-set critical thresholds used to determine whether a pin is qualified, based on engineering standards or product specifications, including allowable distance deviation ranges and allowable angle deviation values. The coplanarity judgment operation refers to comparing the initial difference sequence with the allowable deviation parameters and ultimately providing an assessment of whether the overall coplanarity of the component is qualified. The coplanarity judgment result refers to the final conclusion output by the coplanarity judgment operation, including normal coplanarity and abnormal coplanarity. Normal coplanarity means that every pin in the QFN component is determined to be a normal pin. Abnormal coplanarity means that one or more abnormal pins exist in the QFN component.

[0166] Furthermore, the coplanarity judgment operation, based on the initial difference sequence and a preset allowable deviation parameter, is performed to obtain the coplanarity judgment result, including:

[0167] Perform the following operation on each pin of the located components:

[0168] Extract the pin plane angle difference and pin distance difference corresponding to the component pins from the initial difference sequence;

[0169] Extract the distance tolerance range from the tolerance parameters. If the pin distance difference is within the distance tolerance range, mark the component pin as having a normal distance difference; otherwise, mark the component pin as having an abnormal distance difference.

[0170] Extract the angle tolerance value from the tolerance parameters. If the angle difference between the pin planes is less than or equal to the angle tolerance value, mark the component pin as having a normal angle difference; otherwise, mark the component pin as having an abnormal angle difference.

[0171] If the distance difference and angle difference of the component pin are normal, mark the component pin as a normal pin; if the distance difference or angle difference of the component pin is abnormal, mark the component pin as an abnormal pin.

[0172] If every component pin in the located components is a normal pin, the coplanarity judgment result is confirmed as normal. If there is one component pin in the located components that is an abnormal pin, the coplanarity judgment result is confirmed as abnormal.

[0173] Specifically, the allowable distance deviation range refers to the interval of pin distance differences in the allowable deviation parameters that will not cause problems such as cold solder joints or short circuits during component soldering. "Normal distance difference" means the judgment result when the pin distance difference is within the allowable distance deviation range. "Distance difference is common" means the judgment result when the pin distance difference is not within the allowable distance deviation range. The allowable angle deviation value refers to the value of the pin plane angle difference in the allowable deviation parameters that will not cause problems such as cold solder joints or short circuits during component soldering. "Normal angle difference" means the judgment result when the pin plane angle difference is less than or equal to the allowable angle deviation value. "Angle difference is common" means the judgment result when the pin plane angle difference is greater than the allowable angle deviation value. A normal pin refers to a pin that is marked as normal in the final synthesis state. When a pin is marked as normal in both distance and angle differences, it is confirmed as a normal pin. An abnormal pin refers to a pin that is marked as abnormal in the final synthesis state. When any one or both of the pin distance difference and pin plane angle difference of a pin are abnormal, it is confirmed as an abnormal pin.

[0174] S7. If the coplanarity judgment result is a preset coplanarity anomaly, the located element is confirmed as the element to be corrected, and the initial correction instruction is obtained based on the initial difference sequence and the intelligent correction decision model.

[0175] It is clear that the initial calibration command refers to the preliminary calibration scheme formed by summarizing the individual pin calibration commands of all abnormal pins.

[0176] Furthermore, the process of obtaining the initial correction instruction based on the initial difference sequence and the intelligent correction decision model includes:

[0177] Extract the pin data of all abnormal pins in the initial difference sequence to obtain the pin dataset to be corrected. The pin data includes: pin distance difference, pin plane normal vector of physical simulation model and pin plane angle difference.

[0178] The material resilience rate is obtained based on the intelligent model of corrective decision-making and the physical properties of the material.

[0179] For each pin data in the pin dataset to be calibrated, perform the following operation:

[0180] The target correction amount is obtained based on the pin distance difference, the pin plane normal vector of the physical simulation model, and the pin plane angle difference.

[0181] The actual correction amount is obtained based on the target correction amount and the material resilience.

[0182] Based on the actual calibration amount and the intelligent model for calibration decision, a single pin calibration command is obtained, which includes the calibration point coordinates and calibration pressure.

[0183] Summarize the individual pin calibration instructions to obtain a single pin calibration instruction set, and obtain the initial calibration instruction based on the single pin calibration instruction set.

[0184] Specifically, the dataset of pins to be corrected refers to the set of pin data containing all abnormal pins, filtered from the initial difference sequence. The material resilience rate refers to the proportion of the material's shape recovered after deformation under external force, calculated using a correction decision intelligent model based on material physical properties (e.g., elastic modulus, yield strength, Poisson's ratio, etc.), when the external force is removed, relative to the total deformation. The target correction amount refers to the adjustment amount applied to the abnormal pin required to repair the component pin, with the aim of eliminating the distance and angle differences of the pin. The calculation formula for the actual correction amount in the step of obtaining the actual correction amount based on the target correction amount and material resilience rate is as follows:

[0185]

[0186] in, Indicates the actual correction amount. Indicates the target correction amount. This indicates the material's springback rate.

[0187] Specifically, the single-pin calibration command refers to a specific action command generated by the calibration decision intelligent model for a particular abnormal pin, including calibration point coordinates and calibration pressure. The calibration point coordinates refer to the three-dimensional spatial coordinates of the optimal force application point on the pin surface calculated by the intelligent model. The calibration pressure refers to the precisely calculated force value that the piezoelectric actuator needs to apply at the calibration point coordinates to achieve the deformation corresponding to the actual calibration amount. The single-pin calibration command set refers to the collection obtained by summarizing the single-pin calibration commands. Obtaining the initial calibration command based on the single-pin calibration command set means integrating all the single-pin calibration commands in the single-pin calibration command set to obtain a comprehensive command (i.e., the initial calibration command).

[0188] S8. Perform simulation correction operation on the physical simulation model based on the initial correction command to obtain the simulation correction difference sequence. Based on the simulation correction difference sequence and the intelligent correction decision model, obtain the final correction command.

[0189] As can be understood, the simulated calibration operation refers to performing a virtual, digital calibration operation on the physical simulation model in simulation software before formally performing physical calibration on the actual QFN component. Its purpose is to verify whether the initial calibration command can roughly achieve the expected calibration effect. The simulated calibration difference sequence refers to the new state and related deviation data exhibited by the model after performing the simulated calibration operation on the physical simulation model. The final calibration command refers to the final, trial-ready calibration scheme generated by the calibration decision intelligent model after further optimization and fine-tuning based on the simulated calibration difference sequence.

[0190] Furthermore, the step of performing simulated calibration operations on the physical simulation model based on the initial calibration command to obtain a simulated calibration difference sequence, and obtaining the final calibration command based on the simulated calibration difference sequence and the calibration decision intelligent model, includes:

[0191] A simulated piezoelectric actuator was constructed based on simulation software.

[0192] By using a simulated piezoelectric actuator and initial calibration instructions, a simulated pin calibration operation is performed on each abnormal pin in the physical simulation model to obtain a simulated calibration element model.

[0193] Based on the theoretical model and the simulated correction element model, the simulated correction difference sequence is obtained;

[0194] Based on the simulated correction difference sequence and the intelligent correction decision model, the command correction parameters are obtained;

[0195] Based on the initial calibration command and command correction parameters, the final calibration command is obtained.

[0196] It is clear that the simulated piezoelectric actuator refers to a virtual digital model created within the simulation software based on the physical characteristics of a real piezoelectric actuator (such as size, shape, and the relationship between output force and voltage). The simulated pin calibration operation refers to the operation of using an initial calibration command to drive the simulated piezoelectric actuator to correct abnormal pins in the physical simulation model. The simulated calibration element model refers to the three-dimensional model of the new state obtained after the physical simulation model has undergone simulated calibration operations in the simulation environment according to the initial calibration command. The simulated calibration element model represents the state of the QFN element after calibration by the initial calibration command. The simulated calibration difference sequence refers to the set of residual deviation values ​​calculated after the simulated calibration element model and the theoretical model representing the standard state are spatially registered and compared again. Its purpose is to predict whether each pin still has positional and angular deviations after the implementation of the quantized initial calibration scheme. The method for obtaining the simulated calibration difference sequence based on the theoretical model and the simulated calibration element model is the same as the method for obtaining the initial difference sequence based on the theoretical model and the physical simulation model, and will not be elaborated here. The instruction correction parameter refers to the parameter calculated by the intelligent correction decision model after analyzing the simulated correction difference sequence, used to fine-tune the initial correction instruction. For example, if the correction pressure of a certain pin cannot achieve the repair of pin coplanarity, resulting in a new distance deviation opposite to the original distance deviation, then the instruction correction parameter is "correction pressure -5%". The final correction instruction refers to the final instruction generated after applying the instruction correction parameter to the initial correction instruction, which is then used for actual execution.

[0197] S9. Based on the piezoelectric actuator and the final calibration command, perform pin calibration operation on the element to be calibrated to obtain the leveled element. Use the leveled element as the positioned element and return to the step of performing multi-dimensional scanning operation on the positioned element using a multi-dimensional scanner.

[0198] As is clear, the pin calibration operation refers to the precise physical adjustment of the pins identified as abnormal on a real QFN element using a real piezoelectric actuator, following the coordinates of each final calibration point and its corresponding final calibration pressure in the final calibration instruction. The leveled element refers to the real QFN element after a complete pin calibration operation.

[0199] Furthermore, the step of performing pin calibration on the element to be calibrated based on the piezoelectric actuator and the final calibration command to obtain a leveled element includes:

[0200] For each faulty pin in the component to be calibrated, perform the following operation:

[0201] Extract the final single-pin calibration instruction corresponding to the abnormal pin from the final calibration instruction. The final single-pin calibration instruction includes: the final calibration point coordinates and the final calibration pressure.

[0202] Using a pre-built multi-axis motion platform, the piezoelectric actuator is moved to the final calibration point coordinates to obtain the piezoelectric actuator to be started;

[0203] Based on the final calibration pressure and the preset voltage-pressure conversion equation, the initial driving voltage is obtained. The initial driving voltage is used to perform a single-pin calibration operation on the piezoelectric actuator to be started, resulting in a calibrated piezoelectric actuator. The pressure of the calibrated piezoelectric actuator is then monitored to obtain the actual pressure.

[0204] The pressure difference ratio is obtained based on the actual pressure and the corrected pressure.

[0205] If the pressure difference ratio is not within the preset allowable pressure difference ratio range, the initial drive voltage is adjusted based on the pressure difference ratio to obtain a new drive voltage. The new drive voltage is then used as the initial drive voltage, and the process returns to the step of performing a single pin calibration operation on the piezoelectric actuator to be started using the initial drive voltage.

[0206] If the pressure difference ratio is within the preset allowable pressure difference ratio range, the single pin calibration operation is completed based on the preset calibration pressure action time, resulting in a leveled component.

[0207] In detail, the final single-pin calibration command refers to the command extracted from the final calibration command to calibrate a single abnormal pin, including the final calibration point coordinates and the final calibration pressure. The final calibration point coordinates refer to the three-dimensional spatial coordinates of the point where the force is applied for the actual calibration operation. The final calibration pressure refers to the pressure value applied at the calibration point coordinates for the actual calibration. The multi-axis motion platform refers to a pre-installed and calibrated mechanical system capable of controlling the piezoelectric actuator to move with high precision in three-dimensional space. The piezoelectric actuator to be started refers to a piezoelectric actuator that has moved to the final calibration point coordinates but has not yet started pin calibration. The voltage-pressure conversion equation refers to a mathematical equation, pre-determined by the user, representing the correspondence between the driving voltage applied to the piezoelectric actuator and the resulting contact pressure. The initial driving voltage refers to the voltage value applied to the piezoelectric actuator, calculated based on the calibration pressure using the voltage-pressure conversion equation. The single-pin calibration operation refers to the calibration operation performed on a single abnormal pin using a piezoelectric actuator. The calibrating piezoelectric actuator refers to the piezoelectric actuator currently performing a single-pin calibration operation. The pressure monitoring refers to the operation of monitoring the pressure between the piezoelectric actuator and the component pin using a pressure sensor integrated at the end of the piezoelectric actuator (i.e., at the contact point with the pin). The actual pressure refers to the pressure value monitored in real time during the process of the piezoelectric actuator contacting the component pin and applying force. The pressure difference ratio is a relative error percentage calculated by comparing the actual pressure and the calibration pressure, and its calculation method is as follows:

[0208]

[0209] in, Indicates the pressure difference ratio. Indicates actual pressure. This indicates the correction pressure.

[0210] Specifically, the allowable pressure difference ratio range refers to a pre-set interval used to determine whether the deviation between the currently applied actual pressure and the target correction pressure will affect the correction effect. When the pressure difference ratio is not within the allowable pressure difference ratio range, it is considered to affect the correction effect. The preset correction pressure application time refers to a pre-set duration for maintaining the correction pressure stable in order to ensure sufficient and stable plastic deformation of the pin material. The new driving voltage refers to the updated voltage value obtained after the control system adjusts the initial driving voltage accordingly based on the pressure difference ratio when the pressure difference ratio exceeds the allowable range.

[0211] S10. If the coplanarity judgment result is the preset coplanarity normal, the positioned component is confirmed as a qualified component, and the intelligent leveling process of pin coplanarity in high-density QFN package is completed based on the qualified component.

[0212] It should be explained that the qualified component refers to a QFN component that has been finally confirmed to meet the coplanarity standard after undergoing a coplanarity judgment operation.

[0213] To address the problems described in the background art, this invention acquires a QFN element, element design data, a precision fixture, a multi-dimensional scanner, a calibration decision intelligent model, and a piezoelectric actuator. The multi-dimensional scanner includes a 3D scanner and a high-definition camera. This invention, by integrating precision hardware and intelligent software, provides a complete and necessary technical foundation for achieving high-precision automated detection and calibration. Based on pre-constructed orientation markers and a precision fixture within the QFN element, the QFN element is fixed to obtain a positioned element. The positioned element includes multiple element pins. This invention ensures the consistency of the element's position during scanning and repair, eliminates interference from positioning errors on measurement results, and guarantees data accuracy and subsequent calibration precision. Using component design data and pre-built simulation software, a theoretical model is obtained. This theoretical model includes multiple theoretical model pins, each corresponding one-to-one with a component pin. This demonstrates that the invention establishes a precise and error-free ideal benchmark model, providing a reliable reference standard for subsequent calculations of the actual physical deviations of each pin. A multi-dimensional scanner is used to perform multi-dimensional scanning operations on the located component, obtaining actual scan data. This invention employs a non-contact measurement method, enabling rapid and non-destructive acquisition of complete and realistic three-dimensional spatial contours and high-definition visual data of the component pins, providing comprehensive data input for accurate modeling. Based on the actual scan data and simulation software, a physical simulation model is constructed. Based on the theoretical model and the physical simulation model, an initial difference sequence is obtained. As can be seen, this invention can transform actual scan data into an analyzable digital model and quantify the deviation value of each pin from the ideal position, providing data basis for subsequent qualification judgment and correction instruction generation. Based on the initial difference sequence and preset allowable deviation parameters, a coplanarity judgment operation is performed to obtain the coplanarity judgment result. This invention achieves automated and standardized judgment of component pin coplanarity, improving detection efficiency and intelligence. If the coplanarity judgment result is a preset coplanarity anomaly, the located component is confirmed as a component to be corrected. Based on the initial difference sequence and the intelligent correction decision model, an initial correction instruction is obtained. This invention utilizes an artificial intelligence model to generate a targeted preliminary correction scheme based on deviation data, achieving… The intelligent calibration decision-making process involves performing simulated calibration operations on the physical simulation model based on initial calibration instructions to obtain a simulated calibration difference sequence. Based on this sequence and the intelligent calibration decision-making model, the final calibration instruction is obtained. This invention improves the calibration success rate and prevents component damage caused by calibration by pre-simulating and optimizing the calibration instructions in a virtual environment. Based on the piezoelectric actuator and the final calibration instruction, pin calibration operations are performed on the component to be calibrated to obtain a leveled component. Using this leveled component as the positioned component, the process returns to the step of performing a multi-dimensional scan of the positioned component using a multi-dimensional scanner. This invention executes optimized instructions using a high-precision actuator and ensures that the coplanarity of the component pins meets the acceptable standard through an iterative approach.This invention ensures high-quality final products. If the coplanarity assessment result is within the preset range, the positioned component is confirmed as a qualified component. Based on the qualified component, the intelligent leveling process for pin coplanarity in high-density QFN packaging is completed. Therefore, this invention outputs components with normal coplanarity as qualified components, thus completing the intelligent leveling process for pin coplanarity in high-density QFN packaging. Consequently, this invention improves the efficiency, accuracy, and intelligence of QFN component pin coplanarity detection and repair.

[0214] like Figure 2 The diagram shown is a functional block diagram of a pin coplanarity intelligent leveling system in a high-density QFN package provided by an embodiment of the present invention.

[0215] The intelligent pin coplanarity leveling system 100 in the high-density QFN package described in this invention can be installed in an electronic device. Depending on the functions implemented, the intelligent pin coplanarity leveling system 100 in the high-density QFN package may include a basic data acquisition module 101, a pin coplanarity analysis module 102, an intelligent decision-making and simulation correction module 103, and a pin correction and output module 104. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and which are stored in the memory of the electronic device.

[0216] The basic data acquisition module 101 is used to acquire QFN components, component design data, precision fixtures, multi-dimensional scanners, correction decision intelligent models, and piezoelectric actuators. The multi-dimensional scanner includes a 3D scanner and a high-definition camera. Based on the pre-built orientation marks in the QFN component and the precision fixture, the QFN component is fixed to obtain a positioned component. The positioned component includes multiple component pins. Based on the component design data and pre-built simulation software, a theoretical model is acquired. The theoretical model includes multiple theoretical model pins, and the theoretical model pins correspond one-to-one with the component pins. The multi-dimensional scanner is used to perform multi-dimensional scanning operations on the positioned component to obtain actual scan data.

[0217] The pin coplanarity analysis module 102 is used to construct a physical simulation model based on actual scanning data and simulation software, obtain an initial difference sequence based on the theoretical model and the physical simulation model, and perform a coplanarity judgment operation based on the initial difference sequence and preset allowable deviation parameters to obtain the coplanarity judgment result.

[0218] The intelligent decision-making and simulation correction module 103 is used to identify the located element as the element to be corrected if the coplanarity judgment result is a preset coplanarity anomaly, obtain the initial correction instruction based on the initial difference sequence and the correction decision intelligent model, perform a simulation correction operation on the physical simulation model based on the initial correction instruction to obtain the simulation correction difference sequence, and obtain the final correction instruction based on the simulation correction difference sequence and the correction decision intelligent model.

[0219] The pin calibration and output module 104 is used to perform pin calibration operation on the component to be calibrated based on the piezoelectric actuator and the final calibration command, to obtain the leveled component, and to use the leveled component as the positioned component. Then, it returns to the step of using a multi-dimensional scanner to perform multi-dimensional scanning operation on the positioned component. If the coplanarity judgment result is the preset coplanarity normal, the positioned component is confirmed as a qualified component. Based on the qualified component, the intelligent leveling process of pin coplanarity in high-density QFN package is completed.

[0220] In detail, the modules in the intelligent pin coplanarity leveling system 100 of the high-density QFN package described in this embodiment of the invention employ the same principles as described above during use. Figure 1 The high-density QFN package described herein uses the same intelligent leveling technology for pin coplanarity as described above, and can produce the same technical effect, so it will not be elaborated here.

[0221] like Figure 3 The diagram shown is a structural schematic of an electronic device that implements intelligent leveling technology for pin coplanarity in a high-density QFN package according to an embodiment of the present invention.

[0222] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a pin coplanarity intelligent leveling process program in a high-density QFN package.

[0223] The memory 11 includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of the electronic device 1, such as a portable hard drive. In other embodiments, the memory 11 can be an external storage device of the electronic device 1, such as a plug-in portable hard drive, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device 1. Furthermore, the memory 11 includes both internal storage units and external storage devices of the electronic device 1. The memory 11 can be used not only to store application software and various types of data installed on the electronic device 1, such as the code of the pin coplanarity intelligent leveling process in a high-density QFN package, but also to temporarily store data that has been output or will be output.

[0224] In some embodiments, the processor 10 may be composed of integrated circuits, such as a single-packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control unit of the electronic device, connecting various components of the entire electronic device through various interfaces and lines. It executes programs or modules stored in the memory 11 (e.g., intelligent leveling process programs for pin coplanarity in high-density QFN packages) and calls data stored in the memory 11 to perform various functions of the electronic device 1 and process data.

[0225] The bus 12 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 12 can be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to realize the connection and communication between the memory 11 and at least one processor 10, etc.

[0226] Figure 3 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 3The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0227] For example, although not shown, the electronic device 1 may also include a power supply (such as a battery) to power the various components. Preferably, the power supply can be logically connected to the at least one processor 10 through a power management system, thereby enabling functions such as charging management, discharging management, and power consumption management through the power management system. The power supply may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0228] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish communication connections between the electronic device 1 and other electronic devices.

[0229] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), and optionally, a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the electronic device 1 and to display a visual user interface.

[0230] The high-density QFN package pin coplanarity intelligent leveling process program stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When run in the processor 10, it can achieve the following:

[0231] Acquire QFN components, component design data, precision fixtures, multi-dimensional scanners, intelligent models for calibration decisions, and piezoelectric actuators. The multi-dimensional scanners include a 3D scanner and a high-definition camera.

[0232] Based on the pre-built orientation marks and precision fixtures in the QFN element, the QFN element is fixed to obtain a positioned element, wherein the positioned element includes multiple element pins;

[0233] Based on component design data and pre-built simulation software, a theoretical model is obtained. The theoretical model includes multiple theoretical model pins, and each theoretical model pin corresponds one-to-one with a component pin.

[0234] A multi-dimensional scanner is used to perform multi-dimensional scanning of the positioned component to obtain actual scan data.

[0235] Based on actual scanning data and simulation software, a physical simulation model is constructed, and based on the theoretical model and the physical simulation model, the initial difference sequence is obtained;

[0236] Based on the initial difference sequence and the preset allowable deviation parameters, a coplanarity judgment operation is performed to obtain the coplanarity judgment result;

[0237] If the coplanarity judgment result is a preset coplanarity anomaly, the located element is identified as the element to be corrected, and the initial correction instruction is obtained based on the initial difference sequence and the intelligent correction decision model.

[0238] Based on the initial calibration command, a simulation calibration operation is performed on the physical simulation model to obtain a simulation calibration difference sequence. Based on the simulation calibration difference sequence and the calibration decision intelligent model, the final calibration command is obtained.

[0239] Based on the piezoelectric actuator and the final calibration command, the pin calibration operation is performed on the element to be calibrated to obtain the leveled element. The leveled element is then used as the positioned element, and the process returns to the step of performing a multi-dimensional scanning operation on the positioned element using a multi-dimensional scanner.

[0240] If the coplanarity judgment result is that the pre-set coplanarity is normal, the positioned component is confirmed as a qualified component, and the intelligent leveling process of pin coplanarity in high-density QFN package is completed based on the qualified component.

[0241] Specifically, the processor 10's implementation method for the above instructions can be found in [reference needed]. Figures 1 to 3 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0242] Furthermore, if the modules / units integrated in the electronic device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or system capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0243] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can perform the following:

[0244] Acquire QFN components, component design data, precision fixtures, multi-dimensional scanners, intelligent models for calibration decisions, and piezoelectric actuators. The multi-dimensional scanners include a 3D scanner and a high-definition camera.

[0245] Based on the pre-built orientation marks and precision fixtures in the QFN element, the QFN element is fixed to obtain a positioned element, wherein the positioned element includes multiple element pins;

[0246] Based on component design data and pre-built simulation software, a theoretical model is obtained. The theoretical model includes multiple theoretical model pins, and each theoretical model pin corresponds one-to-one with a component pin.

[0247] A multi-dimensional scanner is used to perform multi-dimensional scanning of the positioned component to obtain actual scan data.

[0248] Based on actual scanning data and simulation software, a physical simulation model is constructed, and based on the theoretical model and the physical simulation model, the initial difference sequence is obtained;

[0249] Based on the initial difference sequence and the preset allowable deviation parameters, a coplanarity judgment operation is performed to obtain the coplanarity judgment result;

[0250] If the coplanarity judgment result is a preset coplanarity anomaly, the located element is identified as the element to be corrected, and the initial correction instruction is obtained based on the initial difference sequence and the intelligent correction decision model.

[0251] Based on the initial calibration command, a simulation calibration operation is performed on the physical simulation model to obtain a simulation calibration difference sequence. Based on the simulation calibration difference sequence and the calibration decision intelligent model, the final calibration command is obtained.

[0252] Based on the piezoelectric actuator and the final calibration command, the pin calibration operation is performed on the element to be calibrated to obtain the leveled element. The leveled element is then used as the positioned element, and the process returns to the step of performing a multi-dimensional scanning operation on the positioned element using a multi-dimensional scanner.

[0253] If the coplanarity judgment result is that the pre-set coplanarity is normal, the positioned component is confirmed as a qualified component, and the intelligent leveling process of pin coplanarity in high-density QFN package is completed based on the qualified component.

[0254] In the embodiments provided by this invention, it should be understood that the disclosed devices, systems, and processes can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and actual implementations may have other classification methods.

[0255] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0256] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0257] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0258] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A smart leveling process for pin coplanarity in a high-density QFN package, characterized in that, The process includes: Acquire QFN components, component design data, precision fixtures, multi-dimensional scanners, intelligent models for calibration decisions, and piezoelectric actuators. The multi-dimensional scanners include a 3D scanner and a high-definition camera. Based on the pre-built orientation marks and precision fixtures in the QFN element, the QFN element is fixed to obtain a positioned element, wherein the positioned element includes multiple element pins; Based on component design data and pre-built simulation software, a theoretical model is obtained. The theoretical model includes multiple theoretical model pins, and each theoretical model pin corresponds one-to-one with a component pin. A multi-dimensional scanner is used to perform multi-dimensional scanning of the positioned component to obtain actual scan data. Based on actual scanning data and simulation software, a physical simulation model is constructed, and based on the theoretical model and the physical simulation model, the initial difference sequence is obtained; Based on the initial difference sequence and the preset allowable deviation parameters, a coplanarity judgment operation is performed to obtain the coplanarity judgment result; If the coplanarity judgment result is a preset coplanarity anomaly, the located element is identified as the element to be corrected, and the initial correction instruction is obtained based on the initial difference sequence and the intelligent correction decision model. Based on the initial calibration command, a simulation calibration operation is performed on the physical simulation model to obtain a simulation calibration difference sequence. Based on the simulation calibration difference sequence and the calibration decision intelligent model, the final calibration command is obtained. Based on the piezoelectric actuator and the final calibration command, the pin calibration operation is performed on the element to be calibrated to obtain the leveled element. The leveled element is then used as the positioned element, and the process returns to the step of performing a multi-dimensional scanning operation on the positioned element using a multi-dimensional scanner. If the coplanarity judgment result is that the pre-set coplanarity is normal, the positioned component is confirmed as a qualified component, and the intelligent leveling process of pin coplanarity in high-density QFN package is completed based on the qualified component.

2. The intelligent leveling process for pin coplanarity in the high-density QFN package as described in claim 1, characterized in that, The process of using a multi-dimensional scanner to perform a multi-dimensional scanning operation on the located component to obtain actual scan data includes: The original point cloud data is obtained by performing a 3D scanning operation on the located components using a 3D scanner. Using a pre-built point cloud denoising algorithm, point cloud denoising is performed on the original point cloud data to obtain three-dimensional point cloud data; High-definition images of the positioned components are captured using a high-definition camera; A pre-built grayscale conversion algorithm is used to perform grayscale conversion on high-definition images to obtain grayscale images; A pre-built contrast enhancement algorithm is used to enhance the contrast of a grayscale image, resulting in an enhanced image. A pre-built image denoising algorithm is used to perform image denoising on the enhanced image to obtain a denoised image; Based on a pre-constructed segmentation threshold algorithm and a pre-constructed edge detection algorithm, edge detection is performed on the denoised image to obtain an edge-detected image; By combining 3D point cloud data and edge detection images, the actual scan data is obtained.

3. The intelligent leveling process for pin coplanarity in the high-density QFN package as described in claim 2, characterized in that, The construction of a physical simulation model based on actual scanning data and simulation software includes: An initial 3D model is constructed based on 3D point cloud data from simulation software and actual scanning data. Obtain the mapping relationship matrix, and based on the mapping relationship matrix, map the edge detection images in the actual scan data to the initial 3D model to obtain the textured 3D model; Material physical properties are extracted from component design data. Based on the material physical properties and textured 3D model, a physical simulation model is constructed. The physical simulation model includes multiple physical simulation model pins, and each physical simulation model pin corresponds one-to-one with a component pin.

4. The intelligent leveling process for pin coplanarity in the high-density QFN package as described in claim 3, characterized in that, The process of obtaining the mapping matrix includes: Based on a pre-built corner detection algorithm and 3D point cloud data, obtain a set of 3D corner coordinates; Based on corner detection algorithms and edge detection images, a two-dimensional set of corner coordinates is obtained; Extract the preliminary matching relationship matrix from the multidimensional scanner, and map the two-dimensional corner coordinate set to the initial three-dimensional model based on the preliminary matching relationship matrix to obtain the mapped corner coordinate set; Matrix adjustment parameters are obtained based on the 3D corner coordinate set, the mapped corner coordinate set, and the pre-constructed nearest point registration algorithm; Based on the initial matching relationship matrix and matrix adjustment parameters, the mapping relationship matrix is ​​obtained.

5. The intelligent leveling process for pin coplanarity in the high-density QFN package as described in claim 4, characterized in that, The process of obtaining the initial difference sequence based on theoretical and physical simulation models includes: Spatial registration is performed on the theoretical model and the physical simulation model to obtain a merged model, which includes a theoretical model layer and a physical simulation model layer. The theoretical model pin point cloud sequence is obtained based on the theoretical model pins in the theoretical model layer, and the physical simulation model pin point cloud sequence is obtained based on the physical simulation model pins in the physical simulation model layer. The theoretical model pin point cloud sequence includes multiple theoretical model pin point clouds, and each theoretical model pin point cloud corresponds to a theoretical model pin. The physical simulation model pin point cloud sequence includes multiple physical simulation model pin point clouds, and each physical simulation model pin point cloud corresponds to a physical simulation model pin. Based on the pin point cloud sequence of the theoretical model and a pre-built plane fitting algorithm, the pin plane equation sequence of the theoretical model is obtained. Based on the pin point cloud sequence of the physical simulation model and the plane fitting algorithm, the pin plane equation sequence of the physical simulation model is obtained. Based on the pin plane equation sequence of the theoretical model and the pin plane equation sequence of the physical simulation model, calculate the pin plane normal vector sequence of the theoretical model and the pin plane normal vector sequence of the physical simulation model. The pin plane angle difference sequence is obtained based on the pin plane normal vector sequence of the theoretical model and the pin plane normal vector sequence of the physical simulation model. The pin plane angle difference sequence includes multiple pin plane angle differences, and the pin plane angle differences correspond one-to-one with the component pins. Based on the pin point cloud sequence of the theoretical model, obtain the coordinate sequence of the pin center point of the theoretical model. Based on the pin point cloud sequence of the physical simulation model, obtain the coordinate sequence of the pin center point of the physical simulation model; Based on the pin center point coordinate sequence of the theoretical model and the pin center point coordinate sequence of the physical simulation model, the pin distance difference sequence is obtained. The pin distance difference sequence includes multiple pin distance differences, and each pin distance difference corresponds one-to-one with a component pin. The initial difference sequence is obtained by summing the pin plane angle difference sequence and the pin distance difference sequence.

6. The intelligent leveling process for pin coplanarity in the high-density QFN package as described in claim 5, characterized in that, The process of performing a coplanarity determination operation based on the initial difference sequence and preset allowable deviation parameters to obtain the coplanarity determination result includes: Perform the following operation on each pin of the located components: Extract the pin plane angle difference and pin distance difference corresponding to the component pins from the initial difference sequence; Extract the distance tolerance range from the tolerance parameters. If the pin distance difference is within the distance tolerance range, mark the component pin as having a normal distance difference; otherwise, mark the component pin as having an abnormal distance difference. Extract the angle tolerance value from the tolerance parameters. If the angle difference between the pin planes is less than or equal to the angle tolerance value, mark the component pin as having a normal angle difference; otherwise, mark the component pin as having an abnormal angle difference. If the distance difference and angle difference of the component pin are normal, mark the component pin as a normal pin; if the distance difference or angle difference of the component pin is abnormal, mark the component pin as an abnormal pin. If every component pin in the located components is a normal pin, the coplanarity judgment result is confirmed as normal. If there is one component pin in the located components that is an abnormal pin, the coplanarity judgment result is confirmed as abnormal.

7. The intelligent leveling process for pin coplanarity in the high-density QFN package as described in claim 6, characterized in that, The intelligent model based on the initial difference sequence and correction decision obtains the initial correction instruction, including: Extract the pin data of all abnormal pins in the initial difference sequence to obtain the pin dataset to be corrected. The pin data includes: pin distance difference, pin plane normal vector of physical simulation model and pin plane angle difference. The material resilience rate is obtained based on the intelligent model of corrective decision-making and the physical properties of the material. For each pin data in the pin dataset to be calibrated, perform the following operation: The target correction amount is obtained based on the pin distance difference, the pin plane normal vector of the physical simulation model, and the pin plane angle difference. The actual correction amount is obtained based on the target correction amount and the material resilience. Based on the actual calibration amount and the intelligent model for calibration decision, a single pin calibration command is obtained, which includes the calibration point coordinates and calibration pressure. Summarize the individual pin calibration instructions to obtain a single pin calibration instruction set, and obtain the initial calibration instruction based on the single pin calibration instruction set.

8. The intelligent leveling process for pin coplanarity in the high-density QFN package as described in claim 7, characterized in that, The process involves performing simulated calibration operations on the physical simulation model based on the initial calibration command to obtain a simulated calibration difference sequence. Based on this sequence and the intelligent calibration decision model, the final calibration command is then obtained, including: A simulated piezoelectric actuator was constructed based on simulation software. By using a simulated piezoelectric actuator and initial calibration instructions, a simulated pin calibration operation is performed on each abnormal pin in the physical simulation model to obtain a simulated calibration element model. Based on the theoretical model and the simulated correction element model, the simulated correction difference sequence is obtained; Based on the simulated correction difference sequence and the intelligent correction decision model, the command correction parameters are obtained; Based on the initial calibration command and command correction parameters, the final calibration command is obtained.

9. The intelligent leveling process for pin coplanarity in the high-density QFN package as described in claim 8, characterized in that, The step of performing pin calibration on the component to be calibrated based on the piezoelectric actuator and the final calibration command to obtain a leveled component includes: For each faulty pin in the component to be calibrated, perform the following operation: Extract the final single-pin calibration instruction corresponding to the abnormal pin from the final calibration instruction. The final single-pin calibration instruction includes: the final calibration point coordinates and the final calibration pressure. Using a pre-built multi-axis motion platform, the piezoelectric actuator is moved to the final calibration point coordinates to obtain the piezoelectric actuator to be started; Based on the final calibration pressure and the preset voltage-pressure conversion equation, the initial driving voltage is obtained. The initial driving voltage is used to perform a single-pin calibration operation on the piezoelectric actuator to be started, resulting in a calibrated piezoelectric actuator. The pressure of the calibrated piezoelectric actuator is then monitored to obtain the actual pressure. The pressure difference ratio is obtained based on the actual pressure and the corrected pressure. If the pressure difference ratio is not within the preset allowable pressure difference ratio range, the initial drive voltage is adjusted based on the pressure difference ratio to obtain a new drive voltage. The new drive voltage is then used as the initial drive voltage, and the process returns to the step of performing a single pin calibration operation on the piezoelectric actuator to be started using the initial drive voltage. If the pressure difference ratio is within the preset allowable pressure difference ratio range, the single pin calibration operation is completed based on the preset calibration pressure action time, resulting in a leveled component.

10. A smart leveling system for pin coplanarity in a high-density QFN package, characterized in that, The system includes: The basic data acquisition module is used to acquire QFN components, component design data, precision fixtures, multi-dimensional scanners, intelligent calibration decision models, and piezoelectric actuators. The multi-dimensional scanner includes a 3D scanner and a high-definition camera. Based on the pre-built orientation marks in the QFN component and the precision fixture, the QFN component is fixed to obtain a positioned component. The positioned component includes multiple component pins. Based on the component design data and pre-built simulation software, a theoretical model is obtained. The theoretical model includes multiple theoretical model pins, which correspond one-to-one with the component pins. The multi-dimensional scanner is used to perform multi-dimensional scanning operations on the positioned component to obtain actual scan data. The pin coplanarity analysis module is used to construct a physical simulation model based on actual scanning data and simulation software, obtain an initial difference sequence based on the theoretical model and the physical simulation model, and perform a coplanarity judgment operation based on the initial difference sequence and preset allowable deviation parameters to obtain the coplanarity judgment result. The intelligent decision-making and simulation correction module is used to identify the located component as the component to be corrected if the coplanarity judgment result is a preset coplanarity anomaly. Based on the initial difference sequence and the intelligent correction decision model, it obtains the initial correction instruction, performs a simulation correction operation on the physical simulation model based on the initial correction instruction, obtains the simulation correction difference sequence, and obtains the final correction instruction based on the simulation correction difference sequence and the intelligent correction decision model. The pin calibration and output module is used to perform pin calibration operations on the component to be calibrated based on the piezoelectric actuator and the final calibration command, to obtain a leveled component. The leveled component is then used as the positioned component. The module returns to the step of performing multi-dimensional scanning operation on the positioned component using a multi-dimensional scanner. If the coplanarity judgment result is the preset coplanarity normal, the positioned component is confirmed as a qualified component. Based on the qualified component, the intelligent pin coplanarity leveling process in the high-density QFN package is completed.

Images