A packaging method and system for implementing a pressure sensor

Abstract

The present invention relates to the field of semiconductor packaging, and discloses a packaging method and system for realizing a pressure sensor, including: acquiring an image of a sensing chip of a pressure sensor to be packaged and collecting area data, obtaining chip feature data through optimization processing and feature annotation, matching with historical data of a preset packaging device, querying potential risk data, screening dangerous packaging parameters and monitoring real-time parameters, calculating an influence degree value, analyzing the type of packaging failure, locating an abnormal component to detect performance and find defect points, formulating a packaging planning goal accordingly, analyzing packaging measures and querying standards to generate operation instructions, adjusting packaging parameters according to this to obtain final parameters, evaluating the packaging quality grade and extracting key quality factors, and further generating a packaging scheme. The present invention can improve the comprehensive measurement performance of the pressure sensor.

Description

Technical Field

[0001] The present invention relates to a packaging method and system for realizing a pressure sensor, and belongs to the field of semiconductor packaging. Background Art

[0002] In many fields such as modern industry, automobile manufacturing, medical equipment, and consumer electronics, pressure sensors play a crucial role. With the continuous progress of technology and the increasingly stringent requirements for sensor performance in various industries, the application scenarios of pressure sensors are becoming more and more extensive, and higher standards are also put forward for their reliability, stability, and accuracy.

[0003] At present, traditional pressure sensor packaging methods gradually show a series of problems when dealing with complex and changeable usage environments and continuously improving performance requirements. On the one hand, traditional packaging materials and processes cannot provide sufficient reliable protection for pressure sensors in the face of harsh working environments (such as high humidity, strongly corrosive gases, drastic temperature changes, etc.), resulting in the sensitive elements of the sensors being easily eroded by external factors, thereby reducing the sensitivity and service life of the sensors. On the other hand, traditional packaging methods encounter bottlenecks in the process of realizing miniaturization and integration, and it is difficult to meet the development trend of current electronic products being thin, light, and multi-functional. Moreover, during the signal transmission process, interference and loss are easily introduced, affecting the measurement accuracy of pressure sensors. Therefore, a packaging method for realizing pressure sensors is needed to improve the comprehensive measurement performance of pressure sensors. Summary of the Invention

[0004] The present invention provides a packaging method and system for realizing a pressure sensor, and its main purpose is to improve the comprehensive measurement performance of the pressure sensor.

[0005] To achieve the above object, a packaging method for realizing a pressure sensor provided by the present invention includes:

[0006] Obtain the sensing chip image corresponding to the pressure sensor to be packaged, collect the chip area data corresponding to the sensing chip image, optimize the sensing chip image based on the chip area data to obtain an optimized chip image, and perform feature annotation on the optimized chip image to obtain chip feature data;

[0007] Match the chip feature data with the historical packaging data of a preset packaging device to obtain packaging matching data, query the potential risk data in the packaging matching data, screen the dangerous packaging parameters corresponding to the potential risk data, and real-time monitor the real-time packaging parameters of the preset packaging device, and calculate the influence degree value of the dangerous packaging parameters on the real-time packaging parameters;

[0008] Analyze the encapsulation fault type corresponding to the preset encapsulation device based on the influence degree value, locate the abnormal components in the preset encapsulation device based on the encapsulation fault type, perform performance detection on the abnormal components to obtain the actual performance of the components, and query the specific defect points in the actual performance of the components;

[0009] Based on the specific defect points, formulate the encapsulation planning target corresponding to the pressure sensor to be encapsulated, analyze the encapsulation measures in the encapsulation planning target, and query the encapsulation standards corresponding to the encapsulation measures. Based on the encapsulation standards, generate the encapsulation operation instructions corresponding to the pressure sensor to be encapsulated;

[0010] Based on the encapsulation operation instructions, adjust the encapsulation parameters of the pressure sensor to be encapsulated to obtain the final encapsulation parameters. Based on the final encapsulation parameters, analyze the encapsulation quality level corresponding to the pressure sensor to be encapsulated, and extract the key quality factors in the encapsulation quality level. Based on the key quality factors, generate the encapsulation scheme corresponding to the pressure sensor to be encapsulated.

[0011] Optionally, the optimizing the image of the sensing chip based on the chip region data to obtain an optimized chip image includes:

[0012] Analyze the regional texture features in the chip region data;

[0013] Perform texture partitioning on the sensing chip picture based on the regional texture feature surface, with texture partitioning units;

[0014] Identify the color distribution data in the texture partitioning units;

[0015] Calculate the color adjustment coefficient corresponding to the color distribution data;

[0016] Optimize the image of the sensing chip based on the color adjustment coefficient to obtain an optimized chip image.

[0017] Optionally, as an embodiment of the present invention, the calculating the color adjustment coefficient corresponding to the color distribution data includes:

[0018] Calculate the color adjustment coefficient corresponding to the color distribution data using the following formula:

[0019]

[0020] Wherein, represents the color adjustment coefficient corresponding to the color distribution data, represents the number of dimensions of the color dimension corresponding to the color distribution data, represents the quantity index corresponding to the color dimension, represents the The sensitivity weight corresponding to each color dimension, indicating the measured value of the actual color distribution corresponding to the th color dimension, the variance value corresponding to the th color dimension, and the adjustment limit value corresponding to the

[0021] th color dimension. Optionally, querying the potential risk data in the encapsulated matching data includes:

[0022] analyzing the encapsulated environment conditions corresponding to the encapsulated matching data;

[0023] querying the abnormal environment factors in the encapsulated environment conditions;

[0024] sorting out the factor risk categories corresponding to the abnormal environment factors;

[0025] extracting the specific risk cases in the factor risk categories;

[0026] and querying the potential risk data in the encapsulated matching data based on the specific risk cases.

[0027] Optionally, calculating the influence degree value of the dangerous encapsulation parameters on the real-time encapsulation parameters includes:

[0028] calculating the influence degree value of the dangerous encapsulation parameters on the real-time encapsulation parameters by using the following formula:

[0029]

[0030] where, represents the influence degree value of the dangerous encapsulation parameters on the real-time encapsulation parameters, represents the number of parameters corresponding to the main dangerous encapsulation parameters, represents the number index corresponding to the main dangerous encapsulation parameters, indicating the real-time encapsulation parameter value corresponding to the th main dangerous encapsulation parameter, represents the main danger threshold corresponding to the th main dangerous encapsulation parameter, represents the importance coefficient corresponding to the th main dangerous encapsulation parameter, represents the number of parameters corresponding to the secondary dangerous encapsulation parameters, indicating the real-time encapsulation parameter value corresponding to the th secondary dangerous encapsulation parameter, indicating the The secondary hazard threshold corresponding to a secondary hazard encapsulation parameter.

[0031] Optionally, analyzing the encapsulation fault type corresponding to a preset encapsulation device based on the influence degree value includes:

[0032] Analyzing the contribution ratio of each hazard encapsulation parameter in the influence degree value;

[0033] Identifying the core hazard parameter corresponding to the influence degree value based on the contribution ratio;

[0034] Querying the past fault modes associated with the core hazard parameter in the historical fault database;

[0035] Extracting the encapsulation process characteristics in the past fault modes;

[0036] Analyzing the encapsulation fault type corresponding to a preset encapsulation device based on the encapsulation process characteristics.

[0037] Optionally, querying the specific defect points in the actual performance of the component includes:

[0038] Sorting out the performance dimensions corresponding to the actual performance of the component;

[0039] Analyzing the parameter fluctuation in each dimension of the performance dimension;

[0040] Drawing the performance change curve corresponding to the actual performance of the component based on the parameter fluctuation;

[0041] Identifying the abnormal fluctuation points in the performance change curve;

[0042] Querying the specific defect points in the actual performance of the component based on the abnormal fluctuation points.

[0043] Optionally, formulating the encapsulation planning goal for the pressure sensor to be encapsulated based on the specific defect points includes:

[0044] Querying the cause of generation corresponding to the specific defect point;

[0045] Analyzing the associated components corresponding to the cause of generation;

[0046] Generating the improvement measures corresponding to the associated components;

[0047] Determining the encapsulation improvement conditions for the pressure sensor to be encapsulated based on the improvement measures;

[0048] Formulating the encapsulation planning goal for the pressure sensor to be encapsulated based on the encapsulation improvement conditions.

[0049] Optionally, generating the encapsulation operation instruction corresponding to the pressure sensor to be encapsulated based on the encapsulation standard includes:

[0050] Analyze the standard detail entries in the encapsulation standard;

[0051] Extract the key entry elements from the standard detail entries;

[0052] Based on the key entry elements, construct the specific action process corresponding to the pressure sensor to be encapsulated;

[0053] Analyze the key action descriptions in the specific action process;

[0054] Based on the key action descriptions, generate the encapsulation operation instruction corresponding to the pressure sensor to be encapsulated.

[0055] To solve the above problems, the present invention also provides an encapsulation system for implementing a pressure sensor, and the system includes:

[0056] A feature marking module, configured to obtain a sensing chip image corresponding to a pressure sensor to be encapsulated, collect chip area data corresponding to the sensing chip image, optimize the sensing chip image based on the chip area data to obtain a chip-optimized image, and perform feature marking on the chip-optimized image to obtain chip feature data;

[0057] A degree value calculation module, configured to perform data matching on the chip feature data and historical encapsulation data of a preset encapsulation device to obtain encapsulation matching data, query potential risk data in the encapsulation matching data, screen out dangerous encapsulation parameters corresponding to the potential risk data, and real-time monitor real-time encapsulation parameters of the preset encapsulation device, and calculate an influence degree value of the dangerous encapsulation parameters on the real-time encapsulation parameters;

[0058] A defect point query module, configured to analyze an encapsulation failure type corresponding to a preset encapsulation device based on the influence degree value, locate an abnormal component in the preset encapsulation device based on the encapsulation failure type, perform performance detection on the abnormal component to obtain actual component performance, and query specific defect points in the actual component performance;

[0059] An instruction generation module, configured to formulate an encapsulation planning target corresponding to the pressure sensor to be encapsulated based on the specific defect points, analyze encapsulation measures in the encapsulation planning target, query an encapsulation standard corresponding to the encapsulation measures, and generate an encapsulation operation instruction corresponding to the pressure sensor to be encapsulated based on the encapsulation standard;

[0060] A solution generation module, which is used to adjust the packaging parameters of the pressure sensor to be packaged based on the packaging operation instruction to obtain the final packaging parameters, analyze the corresponding packaging quality level of the pressure sensor to be packaged based on the final packaging parameters, extract the key quality factors in the packaging quality level, and generate a packaging solution corresponding to the pressure sensor to be packaged based on the key quality factors.

[0061] Compared with the problems described in the background art, the present invention can accurately grasp the physical characteristics and structural details of the sensing chip by acquiring the image of the sensing chip corresponding to the pressure sensor to be packaged and collecting the chip area data corresponding to the sensing chip image, which helps to identify potential problems of the chip and also lays a solid foundation for a series of operations such as packaging equipment matching and risk assessment based on image features. By matching the chip feature data with the historical packaging data of the preset packaging equipment, the present invention obtains packaging matching data, which can quickly compare the compatibility between the chip and the packaging equipment and avoid packaging failures caused by incompatibility. It can accurately screen out the best packaging parameters and solutions suitable for the current chip based on successful cases and experience in historical data, improving packaging efficiency and quality. Further, based on the influence degree value, the present invention analyzes the packaging failure types corresponding to the preset packaging equipment, can accurately locate the root cause of the failure, and by numerically quantifying the influence degree, can quickly determine which dangerous packaging parameters have a large interference on the real-time parameters, and then infer the possible failure types, improving the continuity and stability of packaging production. Further, based on the specific defect points, the present invention formulates the packaging planning goal corresponding to the pressure sensor to be packaged, can effectively avoid potential risks, can adjust the packaging process and optimize the material selection in a targeted manner, ensure that the new packaging process avoids existing problems, can improve the finished product quality of the pressure sensor to be packaged, reduce the unqualified rate in subsequent inspections, improve production efficiency, and reduce production costs. Finally, based on the packaging operation instruction, the present invention adjusts the packaging parameters of the pressure sensor to be packaged to obtain the final packaging parameters, which can ensure that the packaging parameters are highly consistent with the operation instruction, make the packaging process strictly carried out according to the predetermined standard, effectively improve the packaging quality, and reduce packaging defects caused by unreasonable parameters. On the other hand, through adjustment, the packaging parameters can be more adapted to the actual packaging environment and equipment characteristics, improving the efficiency and success rate of the packaging work. Therefore, a packaging method and system for realizing a pressure sensor provided by an embodiment of the present invention can improve the comprehensive measurement performance of the pressure sensor. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a schematic flowchart of a packaging method for realizing a pressure sensor provided by an embodiment of the present invention;

[0063] Figure 2A schematic diagram of a module for implementing a packaging system of a pressure sensor provided by an embodiment of the present invention.

[0064] The implementation, functional features, and advantages of the present invention will be further described with reference to the embodiments and the accompanying drawings. Specific embodiments

[0065] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0066] An embodiment of the present application provides a method for implementing the packaging of a pressure sensor. The execution subject of the method for implementing the packaging of a pressure sensor includes, but is not limited to, at least one of electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for implementing the packaging of a pressure sensor can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to: a single server, a server cluster, a cloud server, or a cloud server cluster, etc.

[0067] Embodiment 1:

[0068] Refer to Figure 1 As shown, it is a flowchart of a method for implementing the packaging of a pressure sensor provided by an embodiment of the present invention. In this embodiment, the method for implementing the packaging of a pressure sensor includes:

[0069] S1. Obtain the sensing chip image corresponding to the pressure sensor to be packaged, collect the chip area data corresponding to the sensing chip image, optimize the sensing chip image based on the chip area data to obtain an optimized chip image, and perform feature annotation on the optimized chip image to obtain chip feature data.

[0070] By obtaining the sensing chip image corresponding to the pressure sensor to be packaged and collecting the chip area data corresponding to the sensing chip image, the present invention can accurately grasp the physical characteristics and structural details of the sensing chip, which helps to identify potential problems of the chip and also lays a solid foundation for a series of operations such as matching packaging equipment and risk assessment based on image features.

[0071] Among them, the pressure sensor to be encapsulated refers to the pressure sensor element that has been manufactured but not yet encapsulated. It is the core component of the entire pressure sensing system. Although it already has the basic pressure sensing function, since it is not encapsulated, its sensitive element is directly exposed and is easily affected by external factors such as humidity, corrosive gases, and temperature changes, and cannot work stably and reliably. It is urgent to protect its internal structure and improve its performance through the encapsulation process to adapt to different application scenarios; the sensing chip image refers to the visual image obtained by using a special image acquisition device to photograph the sensing chip inside the pressure sensor to be encapsulated. This image accurately records the shape, size, positional relationship of each component part of the sensing chip, as well as information such as the circuit pattern on the chip surface and the layout of micro-components; the chip area data refers to the data set extracted from the sensing chip image that can reflect the physical attributes and geometric features of each area of the sensing chip. It covers the size parameters of different areas of the chip, the width and spacing of circuit lines, the type and specifications of components, and the electrical performance parameters of each area, etc. These data provide key quantitative bases for the optimization processing, feature annotation of the sensing chip image, and the design and analysis of subsequent encapsulation processes. Optionally, the acquisition of the sensing chip image corresponding to the pressure sensor to be encapsulated can be achieved through a scanning electron microscope. For example, it can obtain nanoscale resolution. For a tiny sensing chip, it can clearly show the distribution of its internal circuits and components and obtain an accurate chip image; the acquisition of the chip area data corresponding to the sensing chip image can be achieved through the GrabCut algorithm in the OpenCV library. For example, the approximate contour of the chip is marked as the initial segmentation area, and then the GrabCut algorithm will automatically and iteratively optimize the segmentation result according to information such as the texture and color of the image, separate the chip from the background, and obtain accurate chip area data.

[0072] Furthermore, based on the chip area data, the present invention optimizes the sensing chip image to obtain an optimized chip image, which can effectively improve the image quality, remove noise interference, make details such as the contour and circuit of the chip in the image clearer, and provide a good basis for subsequent accurate analysis. Enhance the image contrast, highlight key features, and help identify the subtle structure and potential defects of the chip.

[0073] Among them, the optimized chip image refers to the final image obtained after optimizing the sensing chip image based on the color adjustment coefficient. Compared with the original image, the optimized image is more coordinated and clear in color, can more accurately present the true features and details of the chip, and has achieved a better balance in aspects such as color saturation, contrast, and brightness, making information such as the texture and structure of the chip.

[0074] As an embodiment of the present invention, the optimization process of the sensing chip image based on the chip region data to obtain a chip-optimized image includes: analyzing the regional texture features in the chip region data; performing texture partitioning on the sensing chip picture based on the regional texture feature surface, with a texture partitioning unit; identifying the color distribution data in the texture partitioning unit; calculating the color adjustment coefficient corresponding to the color distribution data; and performing an optimization process on the sensing chip image based on the color adjustment coefficient to obtain a chip-optimized image.

[0075] Among them, the regional texture feature refers to the characteristics extracted from the chip region data that can reflect the texture conditions of different regions of the sensing chip, which includes the thickness of the texture. For example, the texture of some regions of the chip is rough, while some are delicate; the regularity of the texture, such as whether it is arranged randomly or regularly; and the repetition pattern of the texture, etc. The texture partitioning unit refers to each sub-region formed after partitioning the sensing chip picture according to the regional texture feature. The texture within each partitioning unit has high similarity, that is, the various characteristic parameters of its texture are relatively close. Through this partitioning, the chip image can be divided into different parts according to the texture characteristics. The color distribution data refers to the distribution information about colors identified in the texture partitioning unit, specifically including the proportion of various colors in the partitioning unit, such as the proportion of pixels of each color such as red, green, and blue in the partitioning; the spatial distribution law of colors, that is, the position distribution of different colors in the partitioning; and the distribution of attributes such as the brightness and saturation of colors. The color adjustment coefficient refers to a set of parameters calculated according to the color distribution data and used to adjust the color of the sensing chip image, which determines the amplitude and direction of color adjustment required for each texture partitioning unit, such as enhancing the intensity of a certain color, changing the saturation of the color, or adjusting the contrast of the color, etc.

[0076] Further, the regional texture features in the data of the chip area can be realized through a texture feature extraction algorithm. For example, by comparing the central pixel with its neighboring pixels to generate binary codes, and then counting the occurrence frequencies of different coding patterns to describe the texture features of the image. The texture partitioning of the sensing chip picture can be realized through texture partitioning based on the watershed algorithm. For example, by simulating the process of water flooding, the image is segmented into different regions, and each region corresponds to a texture partitioning unit. The recognition of the color distribution data in the texture partitioning unit can be realized through the OpenCV library. For example, in OpenCV, the calcHist function can be used to calculate the color histogram of the image to obtain the color distribution data. The calculation of the color adjustment coefficient corresponding to the color distribution data can be realized through the following calculation formula. The optimization processing of the sensing chip image can be realized through image optimization processing based on non-local means denoising (NLM). For example, by using the information of similar blocks in the image, the noise is estimated and removed, while the details and edge information of the image are retained, and finally an optimized chip image is obtained.

[0077] Specifically, as an embodiment of the present invention, the calculation of the color adjustment coefficient corresponding to the color distribution data includes:

[0078] The color adjustment coefficient corresponding to the color distribution data is calculated using the following formula:

[0079]

[0080] Where, represents the color adjustment coefficient corresponding to the color distribution data, represents the number of dimensions of the color dimension corresponding to the color distribution data, represents the number index corresponding to the color dimension, represents the th sensitivity weight corresponding to the color dimension, represents the th actual color distribution measurement value corresponding to the color dimension, represents the th variance value corresponding to the color dimension, represents the th adjustment limit value corresponding to the color dimension.

[0081] Specifically, the color adjustment coefficient refers to a coefficient used to quantitatively measure the degree and direction of color adjustment required for given color distribution data, which integrates relevant information of each color dimension in the color distribution data; the color dimension refers to different attributes or aspects for describing colors. There are multiple dimensions in common color models. For example, in the RGB (Red - Green - Blue) color model, red, green, and blue are three color dimensions; in the HSV (Hue - Saturation - Value) color model, hue, saturation, and value are different color dimensions respectively; the sensitivity weight refers to the relative importance of the j-th color dimension in color adjustment. The importance of different color dimensions varies for specific application scenarios. For example, in some images, the hue dimension has a greater impact on the overall style of the image, and its corresponding sensitivity weight will be set relatively high; in other cases, the saturation dimension is more critical, and the sensitivity weight of the saturation dimension will be relatively high at this time; the actual color distribution measurement value refers to the data obtained by measuring the actual distribution of the j-th color dimension in the image, which can be obtained through various methods, such as counting the number of pixels in the image belonging to a specific value range of a certain color dimension, or calculating the average value of pixel values under this color dimension, etc.; the variance value refers to a value used to measure the degree of dispersion of the data of the j-th color dimension. In the context of color dimensions, it reflects the degree of uniformity of the pixel values of this color dimension in the image. A smaller variance value indicates that the pixel values of this color dimension are relatively concentrated and the color distribution is relatively uniform; a larger variance value means that the pixel values are more dispersed and the color difference is larger; the adjustment limit value refers to the limit set for the adjustment range of the j-th color dimension. Different color dimensions may set different adjustment limit values according to their characteristics and application requirements. For example, for the hue dimension that is more sensitive to the human eye, a relatively small adjustment limit value may be set to ensure that the adjusted color is still within an acceptable visual range.

[0082] The present invention performs feature annotation on the optimized image of the chip to obtain chip feature data, marking the positions, sizes, shapes and other features of each component of the chip, making the chip characteristics clear at a glance. These feature data provide accurate basis for the design of subsequent packaging processes and the matching of packaging equipment, helping to improve the overall quality and performance of the pressure sensor packaging.

[0083] Among them, the chip feature data refers to the data set obtained by performing feature annotation on the optimized image of the chip, covering physical features such as the geometric size, shape, and component positions of the chip, electrical characteristic data such as resistance and capacitance, and the function information of each module. It is the key basis for comprehensively reflecting the chip characteristics and supporting the design of subsequent packaging processes, equipment matching, and performance evaluation. Optionally, the feature annotation of the optimized image of the chip can be realized through semantic segmentation algorithms, such as algorithms like U - Net, DeepLabv3+.

[0084] S2. Match the chip feature data with the historical packaging data of a preset packaging device to obtain packaging matching data, query the potential risk data in the packaging matching data, screen the dangerous packaging parameters corresponding to the potential risk data, and monitor the real-time packaging parameters of the preset packaging device in real time, and calculate the influence degree value of the dangerous packaging parameters on the real-time packaging parameters.

[0085] By matching the chip feature data with the historical packaging data of the preset packaging device to obtain packaging matching data, the present invention can quickly compare the compatibility between the chip and the packaging device, avoiding packaging failures caused by mismatches; based on successful cases and experiences in historical data, it can accurately screen out the best packaging parameters and solutions suitable for the current chip, improving packaging efficiency and quality.

[0086] Among them, the preset packaging device refers to a professional device selected in advance for completing the packaging task of the pressure sensor. Such devices have specific packaging process capabilities, covering various operations such as wire bonding, potting, and molding, and can adjust corresponding parameters according to the size, structure, and functional characteristics of different chips to ensure the accuracy and efficiency of the packaging process. For example, a common automatic wire bonder can precisely control the connection of metal wires between the chip and the packaging substrate to build a stable electrical connection for the pressure sensor; the historical packaging data refers to a series of data generated and accumulated by the preset packaging device during past pressure sensor packaging operations. These data detail all key information of each packaging task, including but not limited to the type of chip used during packaging, corresponding packaging process parameters (such as temperature, pressure, time, etc.), abnormal situations and handling methods during packaging, and the product performance test results after the final packaging is completed, etc.; the packaging matching data refers to a set of data with corresponding relationships obtained after comparing and matching the chip feature data with the historical packaging data of the preset packaging device, which clearly indicates the most suitable packaging process, device parameters, possible problems and coping strategies, etc. in the historical packaging cases of the preset packaging device for the current pressure sensor chip to be packaged. Optionally, the matching of the chip feature data with the historical packaging data of the preset packaging device can be achieved through a deep learning framework, such as TensorFlow, PyTorch, etc.

[0087] Furthermore, by querying the potential risk data in the packaging matching data, the present invention can accurately correct the color deviation of the chip image, making the color performance more in line with the true color characteristics of the chip, improving the visual accuracy and authenticity of the image. By reasonably adjusting the color, the recognition of key features of the chip in the image can be enhanced, which is helpful for subsequent feature annotation and analysis work.

[0088] Among them, the potential risk data refers to the data related to the risks that the current packaging task will face, which are identified in the packaging matching data after analyzing and querying the packaging environment conditions, abnormal environmental factors, factor risk categories and specific risk cases. These data include information such as the types of risks that may occur, the causes, the probability of occurrence of similar risks in the past, and the impacts that are likely to be caused.

[0089] As an embodiment of the present invention, the query of potential risk data in the packaging matching data includes: analyzing the packaging environment conditions corresponding to the packaging matching data; querying the abnormal environmental factors in the packaging environment conditions; sorting out the factor risk categories corresponding to the abnormal environmental factors; extracting specific risk cases in the factor risk categories; and based on the specific risk cases, querying the potential risk data in the packaging matching data.

[0090] Among them, the packaging environment conditions refer to various conditions related to the physical, chemical and other environments in the pressure sensor packaging process, including but not limited to environmental parameters such as temperature, humidity, cleanliness, air pressure, and electromagnetic environment, vibration conditions and other factors at the packaging site; the abnormal environmental factors refer to those factors in the packaging environment conditions that deviate from the normal range or standard, for example, the temperature exceeds the specific range suitable for chip packaging, the humidity is too high, causing the chip to get damp, the cleanliness does not meet the standard, causing dust and other impurities to mix into the packaging process, and the abnormal air pressure affects the normal operation of the packaging equipment. These are all abnormal environmental factors; the factor risk category refers to the risks that may be caused by abnormal environmental factors according to their nature and type. For example, temperature anomalies can be classified as heat-related risk categories, which may cause changes in chip performance or deformation of packaging materials; humidity anomalies can be classified as moisture risk categories, which may cause risks such as chip short circuits or corrosion; cleanliness anomalies belong to impurity contamination risk categories, which may easily cause packaging defects and other problems; the specific risk cases refer to actual risk events caused by specific abnormal environmental factors in the past packaging history, for example, due to a failure of the air conditioner in the packaging workshop at a certain time period, the temperature exceeded the normal range, resulting in cold solder joints on the packaged chip; or a specific event record such as the failure of the humidity control in the packaging environment, which caused corrosion of the chip's internal circuit and unqualified final products.

[0091] Further, the analysis of the encapsulation environment conditions corresponding to the encapsulation matching data can be achieved through data analysis tools, such as tools like MATLAB, to organize, analyze, and visualize the collected data to obtain the encapsulation environment conditions; the query of abnormal environmental factors in the encapsulation environment conditions can be achieved through the threshold comparison method, such as setting a reasonable threshold range for each environmental factor according to the known normal encapsulation environment conditions and industry standards, comparing the actual encapsulation environment conditions with the threshold range, and the factors exceeding the threshold range are abnormal environmental factors; the sorting of the factor risk categories corresponding to the abnormal environmental factors can be achieved through the decision tree algorithm, such as training a decision tree model based on the characteristics and historical data of the abnormal environmental factors, and when a new abnormal environmental factor is input, the decision tree model outputs the corresponding factor risk category; the extraction of specific risk cases in the factor risk categories can be achieved through the text matching algorithm, such as the edit distance algorithm, and when the factor risk category keywords are input, the most matching specific risk cases are found by calculating the edit distance between the keywords and the case category descriptions in the database; the query of potential risk data in the encapsulation matching data can be achieved through the clustering algorithm, such as the DBSCAN clustering algorithm, clustering the encapsulation matching data, and determining the potential risk data according to the data points deviating from the normal data clusters in the clustering results.

[0092] By screening the dangerous encapsulation parameters corresponding to the potential risk data and real-time monitoring the real-time encapsulation parameters of the preset encapsulation equipment, the present invention can accurately locate the key parameters that may cause encapsulation problems, avoid risks in advance, and effectively reduce the error rate during the encapsulation process.

[0093] Among them, the dangerous packaging parameters refer to those parameters that are very likely to cause serious problems such as packaging failure and product quality decline once they appear during the packaging process. These parameters can be derived from the analysis of potential risk data, such as too high or too low temperature, pressure exceeding the safe range, too fast or too slow packaging speed, etc. These data deviate greatly from the normal packaging parameter range, directly affecting the bonding effect between the chip and the packaging material, the stability of electrical connection, etc., and then leading to risks such as solder joint voiding, chip damage, and poor sealing; the real-time packaging parameters refer to the operating parameters generated and fed back in real time by the preset packaging equipment during the actual packaging operation. By monitoring these parameters in real time, operators can intuitively understand whether the packaging process is proceeding normally, such as whether the equipment temperature is within the normal working range, whether the pressure is stable near the specified value, etc. Optionally, the screening of the dangerous packaging parameters corresponding to the potential risk data can be achieved through an anomaly detection algorithm based on deep learning, such as an autoencoder. Taking the packaging parameter data as input, the model is allowed to learn the feature representation of normal data, and the parameters corresponding to the data points with large reconstruction errors are regarded as dangerous packaging parameters; the real-time monitoring of the real-time packaging parameters of the preset packaging equipment can be achieved through a monitoring algorithm based on time series analysis, such as using an ARIMA model to model and predict the parameter data collected in real time, and by comparing the difference between the predicted value and the actual collected value, it is judged whether the parameter has abnormal fluctuations, so as to monitor the change of the packaging parameters in real time.

[0094] Furthermore, the present invention calculates the influence degree value of the dangerous packaging parameters on the real-time packaging parameters, providing clear data support for the risk assessment of the packaging process, enabling technicians to clearly understand the severity of potential risks; and helping to reasonably allocate resources according to the influence degree value and prioritize the treatment of risk factors with greater influence.

[0095] Among them, the influence degree value refers to the value that comprehensively reflects the influence of the dangerous packaging parameters on the real-time packaging parameters, intuitively understanding the degree to which the current real-time packaging parameters are affected by the dangerous packaging parameters, and then judging the severity of potential risks during the packaging process based on this, so as to take corresponding measures in a timely manner.

[0096] As an embodiment of the present invention, the calculation of the influence degree value of the dangerous packaging parameters on the real-time packaging parameters includes:

[0097] Using the following formula to calculate the influence degree value of the dangerous packaging parameters on the real-time packaging parameters:

[0098]

[0099] Among them, represents the influence degree value of the dangerous packaging parameters on the real-time packaging parameters, Indicates the number of parameters corresponding to the main hazard encapsulation parameters, Indicates the quantity index corresponding to the main hazard encapsulation parameters, Indicates the Real-time encapsulation parameter value corresponding to the Indicates the Main hazard threshold corresponding to the Indicates the Importance coefficient corresponding to the Indicates the number of parameters corresponding to the secondary hazard encapsulation parameters, Indicates the quantity index of the secondary hazard encapsulation parameters, Indicates the Real-time encapsulation parameter value corresponding to the Indicates the Secondary hazard threshold corresponding to the

[0100] Specifically, the main hazard encapsulation parameters refer to the parameters that have a key impact on the encapsulation quality and product performance during the encapsulation process. For example, in chip encapsulation, parameters such as temperature and pressure are usually regarded as main hazard encapsulation parameters, which directly affect key performance indicators such as the bonding effect between the chip and the encapsulation material and the stability of electrical connections; the main hazard threshold refers to the critical value set for each main hazard encapsulation parameter, which is determined comprehensively based on factors such as encapsulation process requirements, material characteristics, and past production experience. For example, for the main hazard encapsulation parameter of encapsulation temperature, its main hazard threshold can be set as an upper limit temperature and a lower limit temperature, and exceeding this range may affect the encapsulation quality; the importance coefficient refers to the relative importance of the kth main hazard encapsulation parameter in the entire encapsulation process, and its value range is usually between 0 and 1. The larger the value, the more important the impact of the main hazard encapsulation parameter on the encapsulation result; the secondary hazard encapsulation parameters refer to those parameters that also have a certain impact on the encapsulation quality and product performance, but have a smaller impact compared to the main hazard encapsulation parameters. For example, parameters such as humidity and air pressure in the encapsulation environment are usually regarded as secondary hazard encapsulation parameters; the secondary hazard threshold refers to the critical value set for each secondary hazard encapsulation parameter. When the real-time encapsulation parameter approaches or exceeds this threshold, the possibility of risks occurring during the encapsulation process will increase. Similar to the main hazard threshold, the secondary hazard threshold is also determined based on factors such as the characteristics of the encapsulation process and the material tolerance.

[0101] S3. Based on the influence degree value, analyze the encapsulation fault types corresponding to the preset encapsulation equipment. Based on the encapsulation fault types, locate the abnormal components in the preset encapsulation equipment, perform performance detection on the abnormal components to obtain the actual performance of the components, and query the specific defect points in the actual performance of the components.

[0102] Based on the influence degree value, the present invention analyzes the packaging fault types corresponding to the preset packaging equipment, can accurately locate the root cause of the fault, and by quantifying the influence degree with numerical values, it can quickly judge which dangerous packaging parameters have a large interference on the real-time parameters, and then infer the possible fault types that may be caused, improving the continuity and stability of packaging production.

[0103] Among them, the packaging fault type refers to the fault category that will occur in the preset packaging equipment finally determined according to the analysis of the packaging process characteristics. Common packaging fault types include mechanical faults (such as packaging deviation caused by wear and looseness of equipment components), electrical faults (such as abnormal electrical performance caused by poor connection of chip pins), material faults (such as packaging defects caused by aging and deterioration of packaging materials), etc.

[0104] As an embodiment of the present invention, the analyzing the packaging fault types corresponding to the preset packaging equipment based on the influence degree value includes: analyzing the contribution ratio of each dangerous packaging parameter in the influence degree value; identifying the core dangerous parameter corresponding to the influence degree value based on the contribution ratio; querying the past fault modes associated with the core dangerous parameter in the historical fault database; extracting the packaging process characteristics in the past fault modes; and analyzing the packaging fault types corresponding to the preset packaging equipment based on the packaging process characteristics.

[0105] Among them, the contribution ratio refers to the proportion of the contribution of each dangerous packaging parameter to this value, which reflects the relative importance of the role played by each dangerous packaging parameter in the process of affecting the real-time packaging parameter. For example, if the contribution ratio of the dangerous packaging parameter of temperature in the calculation of the influence degree value is 40%, it indicates that in the overall influence degree, the influence of the temperature parameter is relatively large, and its change needs to be focused on; the core dangerous parameter refers to the dangerous packaging parameter that is identified based on the contribution ratio and plays a key role in the influence degree value. These parameters are usually factors that have a significant impact on product quality and packaging effect during the packaging process. For example, in chip packaging, if it is calculated that the contribution ratio of the pressure parameter is relatively high among all dangerous packaging parameters, then the pressure parameter can be identified as a core dangerous parameter; the past failure mode refers to the packaging failure situations that have occurred in the past and are associated with the core dangerous parameters recorded in the historical failure database. It includes various information at the time of the failure, such as the time of the failure, the operating state of the packaging equipment at that time, the abnormal packaging parameter values, etc., as well as the specific manifestation forms of the failure, such as chip solder joint voids, package body cracking, etc.; the packaging process characteristics refer to various characteristics and elements related to the packaging process reflected in the past failure mode. This includes the characteristics of the packaging materials used, the operation steps and parameter settings in the packaging process, the connection situation between different process links, etc. For example, the thermal expansion coefficient of the packaging material, the heating and cooling rates during the packaging process, etc. all belong to the packaging process characteristics.

[0106] Furthermore, the contribution ratio corresponding to each dangerous packaging parameter in the parsed influence degree value can be realized by a contribution degree analysis algorithm based on Monte Carlo simulation. For example, by randomly changing the values of each dangerous packaging parameter in large quantities and observing the changes in the influence degree value, the average value of the changes in the influence degree value caused by each parameter is statistically calculated, and this is used as the contribution ratio of the parameter. The identification of the core dangerous parameter corresponding to the influence degree value can be realized by a threshold screening algorithm. For example, a threshold of the contribution ratio is set, and when the contribution ratio of a certain dangerous packaging parameter exceeds this threshold, it is identified as a core dangerous parameter. The query of the past failure modes associated with the core dangerous parameter in the historical failure database can be realized by an association rule mining algorithm. For example, the Apriori algorithm, by analyzing the association relationship between each parameter and the failure mode in the historical failure database, finds the failure modes that frequently co-occur with the core dangerous parameter. The extraction of the packaging process features in the past failure modes can be realized by a topic model algorithm. For example, algorithms such as LDA, taking the text data of the past failure modes as input, and mining the potential topics therein through the LDA algorithm, and these topics often correspond to different packaging process features. What method, tool or algorithm can be used to analyze the packaging failure types corresponding to the preset packaging equipment? For example, according to the corresponding rules between the packaging process features and the known failure types, logical reasoning is carried out to judge the packaging failure types corresponding to the preset packaging equipment.

[0107] Based on the packaging failure types, the present invention locates the abnormal components in the preset packaging equipment, making the equipment maintenance more targeted, quickly narrowing down the troubleshooting scope, accurately locking the components that may have problems, avoiding blind detection, reducing unnecessary disassembly losses, quickly restoring the normal operation of the equipment, and ensuring the efficient progress of packaging production.

[0108] Among them, the abnormal components refer to the components in the preset packaging equipment that cause packaging failures of the equipment due to reasons such as their own damage, performance degradation, and parameter deviation from the normal range. These components can be the mechanical transmission components of the equipment, such as gears and chains, which will cause deviations in packaging actions due to wear or looseness; they can also be electrical components, such as sensors and controllers, whose abnormal performance will cause the equipment operation parameters to get out of control; they can also be function modules such as heating and refrigeration, which will affect the temperature environment required for packaging when they fail. Optionally, the location of the abnormal components in the preset packaging equipment can be realized by a fault tree analysis method. For example, by constructing a fault tree, taking the failure type as the top event, and gradually analyzing the intermediate events and bottom events that can lead to this failure, and the bottom events often correspond to the specific abnormal components in the equipment.

[0109] Furthermore, the present invention performs performance testing on the abnormal components to obtain the actual performance of the components, and can accurately determine the specific performance deviation of the abnormal components, provide accurate data support for subsequent maintenance or replacement, avoid blind operations, and help to deeply analyze the root causes of the faults, providing a basis for optimizing equipment design and improving processes.

[0110] Among them, the actual performance of the component refers to the current actual working characteristics and capability performance of the component obtained after the performance test of the abnormal component, which covers multiple key dimensions, such as the actual speed, torque, wear degree, and vibration amplitude of mechanical parts; the actual resistance, capacitance, inductance values, and current and voltage parameters of electrical parts, as well as the accuracy and stability of signal transmission. These performance data intuitively reflect the current operating status of the abnormal component. Compared with the performance indicators under normal working conditions, the performance deviation of the component can be clearly presented. Optionally, the performance test of the abnormal component can be achieved through a simulated working condition test method, such as: testing the abnormal component under the actual working environment and working conditions of the simulated preset packaging equipment, observing its operating performance and performance indicators under the simulated working conditions, so as to obtain the actual performance of the component.

[0111] Furthermore, the present invention can accurately locate the root cause of the problem by querying the specific defects in the actual performance of the component, provide a clear direction for subsequent maintenance, replacement or adjustment, help to deeply analyze the causes of the defects, and provide valuable basis for optimizing component design and improving production processes, thereby improving component quality and overall equipment performance.

[0112] The specific defect point refers to the actual fault location or damaged part on the component that causes abnormal performance. For example, through the analysis of the vibration performance change curve, it is found that the vibration is abnormally severe at a certain moment. Further inspection determines that it is due to severe wear of a certain bearing of the component. This worn bearing is the specific defect point.

[0113] As an embodiment of the present invention, the query of specific defects in the actual performance of the component includes: sorting out the performance dimensions corresponding to the actual performance of the component; analyzing the parameter fluctuations in each dimension of the performance dimensions; based on the parameter fluctuations, drawing a performance change curve corresponding to the actual performance of the component; identifying abnormal fluctuation points in the performance change curve; based on the abnormal fluctuation points, querying the specific defects in the actual performance of the component.

[0114] Among them, the performance dimension refers to the categories obtained by dividing the actual performance of a component from different perspectives, and is used to comprehensively describe the performance characteristics of the component. For example, the performance dimensions of mechanical components may include motion performance, mechanical properties, wear performance, etc.; electrical components cover electrical parameter performance, signal transmission performance, etc. The parameter fluctuation situation refers to the numerical change state of each specific parameter under the performance dimension of the component during the actual operation process. For example, the rotational speed of a motor is a parameter under the motion performance dimension, and the changes such as increase, decrease, or sudden speed changes during operation are the parameter fluctuation situations. The performance change curve refers to a curve plotted with time or other relevant variables as the horizontal axis and the parameter value under the performance dimension of the component as the vertical axis, which can intuitively reflect the change of the component's performance with the variable. For example, taking time as the abscissa and recording the temperature parameter value of a component during operation, the curve of temperature changing with time is the performance change curve. The abnormal fluctuation point refers to a special point on the performance change curve that deviates from the normal change trend. The parameter values corresponding to these points have a large difference from the normal range, and can be manifested as sudden increases, decreases, or sharp fluctuations. For example, on the pressure performance change curve, the point where the pressure value suddenly exceeds the normal working pressure range by a large margin is an abnormal fluctuation point.

[0115] Further, the performance dimensions corresponding to the actual performance of the component can be sorted out through functional analysis methods. For example, starting from the design function of the component, its overall function is decomposed into multiple sub-functions, and each sub-function corresponds to one or more performance dimensions. The analysis of the parameter fluctuation situations in each dimension of the performance dimension can be achieved through the moving average algorithm. For example, the parameter data is smoothed to highlight the fluctuation trend of the data, facilitating the analysis of the fluctuation situation. The plotting of the performance change curve corresponding to the actual performance of the component can be achieved through plotting tools such as Origin, Excel, etc. The identification of the abnormal fluctuation points in the performance change curve can be achieved through the threshold method. For example, a reasonable threshold range is set, and when the data points in the curve exceed this range, they are determined as abnormal fluctuation points. The query of the specific defect points in the actual performance of the component can be achieved through the fault tree analysis method. For example, starting from the abnormal fluctuation points, various factors that may cause the abnormality are analyzed reversely, and a fault tree is constructed to find the specific defect points.

[0116] S4. Based on the specific defect points, formulate the packaging planning objectives corresponding to the pressure sensor to be packaged, analyze the packaging measures in the packaging planning objectives, and query the packaging standards corresponding to the packaging measures. Based on the packaging standards, generate the packaging operation instructions corresponding to the pressure sensor to be packaged.

[0117] Based on the specific defect points, the present invention formulates the packaging planning objectives corresponding to the pressure sensor to be packaged, which can effectively avoid potential risks, can specifically adjust the packaging process and optimize the material selection, ensure that the new packaging process avoids existing problems, can improve the finished product quality of the pressure sensor to be packaged, can also reduce the unqualified rate in subsequent inspections, improve production efficiency, and reduce production costs.

[0118] Among them, the packaging planning objective refers to a comprehensive and specific objective system formulated after comprehensively considering factors such as the performance requirements of the pressure sensor to be packaged and the packaging improvement conditions, covering performance indicators such as the accuracy, stability, and sensitivity of pressure measurement; production efficiency indicators, specifying the number of packages completed within a certain time; quality control indicators, setting the allowable defective rate; cost control objectives, determining the cost budget in the packaging process, etc.

[0119] As an embodiment of the present invention, formulating the packaging planning objective corresponding to the pressure sensor to be packaged based on the specific defect points includes: querying the cause corresponding to the specific defect point; analyzing the associated components corresponding to the cause; generating improvement measures corresponding to the associated components; determining the packaging improvement conditions corresponding to the pressure sensor to be packaged based on the improvement measures; and formulating the packaging planning objective corresponding to the pressure sensor to be packaged based on the packaging improvement conditions.

[0120] Among them, the cause refers to the root factors that lead to the appearance of specific defect points. These factors can cover multiple aspects. For example, material quality problems, such as insufficient toughness and poor temperature resistance of the selected encapsulation materials, are likely to cause defects in subsequent processes; improper process operations, such as inaccurate temperature control and uneven pressure application during the encapsulation process; equipment precision defects, where the equipment used for encapsulation is aged and worn, resulting in some key operations not meeting the design requirements. The associated components refer to the components of the pressure sensor to be encapsulated that are directly related to the cause of the defect or the components involved in the encapsulation process. For example, if the defect is caused by a loose solder joint at the welding point, the associated components may be the pins of the pressure sensor, the welding material, and the soldering iron tip of the welding equipment; if the problem is caused by a mismatch in the thermal expansion coefficient of the encapsulation material, the associated components are the pressure sensor chip, the encapsulation housing, and the filling material between them, etc. The improvement measures refer to a series of solutions formulated for the cause and the associated components. If the cause is insufficient equipment precision, the improvement measures can be to calibrate, repair, or replace key components of the equipment; if there are problems with the material quality, the improvement measures are to select higher-quality and more suitable materials; for improper process operations, detailed operation specifications and training plans will be formulated to improve the skills of the operators. The encapsulation improvement conditions refer to a series of conditions determined based on the improvement measures to achieve a good encapsulation effect, including new requirements for the encapsulation materials, such as higher purity and more appropriate physical and chemical properties; adjustment of the encapsulation process parameters, such as precise setting of parameters like temperature, time, and pressure; control of the production environment, such as requirements for cleanliness, humidity, and temperature range; and standards for improving the skills of the operators, etc.

[0121] Further, the query of the cause corresponding to the specific defect point can be realized by the fault tree reverse reasoning algorithm. For example, in a fault tree model, through the logical gate relationship, starting from the defect point of "abnormal output signal of the pressure sensor", it is found that it is caused by basic events such as "loose soldering at the welding point" and "damage to the internal circuit of the chip". These are the causes. The analysis of the associated components corresponding to the causes can be realized by the association mining algorithm based on the knowledge graph. For example, a knowledge graph containing components, processes, faults, etc. of the pressure sensor is constructed. Taking the causes as nodes, through the association relationship mining algorithm of the graph, the component nodes connected to them are found. These components are the associated components. The generation of the improvement measures corresponding to the associated components can be realized by the case-based reasoning (CBR) algorithm. For example, when encountering a new problem of associated components, through the similarity matching algorithm, similar cases are retrieved from the case library, referring to their improvement measures and adjusting them in combination with the current actual situation to generate the final improvement measures. The determination of the packaging improvement conditions corresponding to the pressure sensor to be packaged can be realized by the genetic algorithm. For example, taking process parameters such as temperature and pressure as genes, through the selection, crossover, and mutation operations of the genetic algorithm, the parameter combination that optimizes the performance of the pressure sensor is found. The formulation of the packaging planning objective corresponding to the pressure sensor to be packaged can be realized by the goal programming algorithm. For example, among multiple goals such as cost, quality, and efficiency, according to the enterprise strategy, the weights of each goal are determined, and the optimal packaging planning objective under certain resource constraints is solved by the goal programming algorithm.

[0122] By analyzing the packaging measures in the packaging planning objective of the present invention and querying the packaging standards corresponding to the packaging measures, the scientificity and rationality of the packaging measures can be ensured, making them meet the industry specifications and requirements, and avoiding packaging quality problems caused by improper measures.

[0123] Among them, the encapsulation measures refer to a series of specific operations and methods taken to achieve the encapsulation planning objectives of the pressure sensor to be encapsulated, including but not limited to selecting appropriate encapsulation materials, such as using polymer materials with good insulation and corrosion resistance; determining the encapsulation structure, such as adopting different structural forms like potting, encapsulating or sealing; and applying specific encapsulation processes, such as welding, bonding and other processes to effectively encapsulate each component of the pressure sensor to protect the sensor from the influence of external environmental factors. The encapsulation standards refer to a series of specifications and requirements formulated to ensure the encapsulation quality and performance during the encapsulation process of the pressure sensor, covering the performance index standards of the encapsulation materials, such as the hardness, flexibility, thermal stability, etc. of the materials; the design standards of the encapsulation structure, including requirements in aspects such as dimensional accuracy, spatial layout, etc.; the operation standards of the encapsulation process, such as the specifications of parameters like welding temperature, time, pressure, etc., and the performance test standards after encapsulation, such as the specific numerical requirements of indicators like sealing performance, insulation performance, anti-interference performance, etc. Optionally, the analysis of the encapsulation measures in the encapsulation planning objectives can be achieved through the FMEA analysis method. For example, for the encapsulation of a pressure sensor, there is a failure mode that the performance of the sensor decreases due to the aging of the encapsulation materials. Through FMEA analysis, it can be determined that encapsulation measures such as strengthening the anti-aging performance of the encapsulation materials are required. The query of the encapsulation standards corresponding to the encapsulation measures can be achieved through a standard query platform. For example, the China National Standard Service Network, NSSN (National Standard System Network of the United States), etc. By inputting keywords such as "pressure sensor encapsulation standards", relevant encapsulation standard documents can be quickly retrieved, and specific standard contents such as the performance of encapsulation materials and encapsulation process parameters can be obtained from them.

[0124] Furthermore, based on the encapsulation standards, the present invention generates encapsulation operation instructions corresponding to the pressure sensor to be encapsulated, which can make the encapsulation operation more standardized and ensure that each link is strictly executed in accordance with the established standards, thereby improving the consistency and stability of the encapsulation quality.

[0125] Among them, the encapsulation operation instructions refer to the specific instructions finally formed to guide the operator to encapsulate the pressure sensor to be encapsulated, which are presented in a concise, clear and easy-to-execute language and format, including complete operation steps, execution requirements and precautions for each step. It can be an operation manual in text form or a visual operation guide, ensuring that the operator can accurately complete the encapsulation work according to the instructions and guarantee that the encapsulation quality meets the encapsulation standards.

[0126] As an embodiment of the present invention, generating the encapsulation operation instructions corresponding to the pressure sensor to be encapsulated based on the encapsulation standard includes: parsing the standard detail entries in the encapsulation standard; extracting the key entry elements from the standard detail entries; constructing the specific action process corresponding to the pressure sensor to be encapsulated based on the key entry elements; analyzing the key action descriptions in the specific action process; and generating the encapsulation operation instructions corresponding to the pressure sensor to be encapsulated based on the key action descriptions.

[0127] Among them, the standard detail entries refer to the specific contents of the detailed regulations on all aspects of the encapsulation process in the encapsulation standard, which cover the selection specifications of encapsulation materials, such as the chemical composition of the materials and the requirements for physical performance indicators; to the setting of specific parameters of the encapsulation process, such as the numerical range of welding temperature, time, and pressure; and then to the sequence and operation specifications of the encapsulation process, etc.; the key entry elements refer to the core points extracted from the standard detail entries, and these elements determine the key aspects of the encapsulation operation. For example, in the material element, the temperature resistance level and insulation characteristics of the encapsulation materials, etc.; in the process element, the operation parameters of the key process steps, such as the glue volume control and curing time during potting; the process element is related to the sequence of each operation link, etc.; the specific action process refers to decomposing the encapsulation process into a series of ordered actual operation steps based on the key entry elements, which clarifies the detailed sequence from the start of the preparation work, such as cleaning the pins of the pressure sensor to be encapsulated, to the core encapsulation actions in the middle, such as soldering the chip and filling the encapsulation material, and then to the final inspection and packaging links. Each step is closely connected to the key entry elements to ensure that the encapsulation operation meets the standard requirements; the key action description refers to the detailed description of the important operation steps in the specific action process, which includes the specific operation method, tool use, precautions, etc. For example, in the key action of soldering, describe the type of soldering tool (such as the power of the soldering iron), the soldering technique (such as the contact angle and time between the soldering iron tip and the solder joint), and the requirements for the temperature and humidity environment during the soldering process, etc.

[0128] Furthermore, the parsing of the standard detail items in the packaging standard can be achieved through a syntactic analysis algorithm, such as dependency syntactic analysis, which can analyze the dependency relationship between words in the text, help understand the semantic structure of the standard sentence, and thus accurately parse the detail items; the extraction of key item elements in the standard detail items can be achieved through a TF-IDF algorithm tool, such as the sklearn library in Python, which can calculate the importance of each word in the detail item and screen out important keywords as key item elements; the construction of the specific action flow corresponding to the pressure sensor to be packaged can be achieved through a Petri net algorithm, such as the operation steps, conditions and events in the packaging process can be represented by elements of the Petri net, thereby constructing a specific action flow; the analysis of the specific action The description of key actions in the operation process can be realized through an association rule mining algorithm, such as: analyzing the association relationship between actions, finding the action combination closely related to the packaging quality and performance, and determining the key action description therefrom. For another example, it is found that there is a strong correlation between "accurate chip installation position" and "firm welding", and the descriptions of these actions are key action descriptions; the generation of the packaging operation instructions corresponding to the pressure sensor to be packaged can be realized through a natural language generation algorithm, such as: a series of models such as GPT based on the Transformer architecture. By inputting key action descriptions and related information, the model can generate natural and smooth packaging operation instruction texts. For another example, the key action description "accurately place the pressure sensor chip on the packaging base and gently fix it with tweezers" is input into the model to generate detailed operation instruction statements.

[0129] S5. Based on the packaging operation instructions, adjust the packaging parameters of the pressure sensor to be packaged to obtain final packaging parameters; based on the final packaging parameters, analyze the packaging quality grade corresponding to the pressure sensor to be packaged, and extract the key quality factors in the packaging quality grade; based on the key quality factors, generate a packaging solution corresponding to the pressure sensor to be packaged.

[0130] The present invention adjusts the packaging parameters of the pressure sensor to be packaged based on the packaging operation instructions to obtain final packaging parameters, which can ensure that the packaging parameters are highly consistent with the operation instructions, so that the packaging process is strictly carried out in accordance with predetermined standards, effectively improve the packaging quality, and reduce packaging defects caused by unreasonable parameters. On the other hand, the packaging parameters can be made more adaptable to the actual packaging environment and equipment characteristics through adjustment, thereby improving the efficiency and success rate of the packaging work.

[0131] Among them, the final packaging parameters refer to a series of key values that have been precisely adjusted to ensure that the pressure sensor to be packaged reaches the best performance and quality standards. These parameters cover multiple aspects, such as packaging material-related parameters, like the thickness of the material, curing time, etc.; packaging process parameters, including welding temperature, pressure magnitude, operation duration, etc.; and parameters related to the packaging structure, such as the spacing between components, dimensional accuracy, etc. These parameters are obtained through repeated adjustment and optimization under the guidance of packaging operation instructions, combined with the characteristics of the pressure sensor, application scenarios, and packaging standards, and jointly determine the performance of the pressure sensor after packaging. Optionally, adjusting the packaging parameters of the pressure sensor to be packaged can be achieved by numerical simulation methods. For example: using computer simulation technology to establish a mathematical model of the pressure sensor packaging process, and through numerical calculation, simulate the packaging process and results under different packaging parameters, predict the impact of packaging parameters on the packaging effect, so as to optimize the packaging parameters.

[0132] Furthermore, based on the final packaging parameters, the present invention analyzes the corresponding packaging quality level of the pressure sensor to be packaged, extracts the key quality factors in the packaging quality level, can accurately evaluate the packaging quality, and based on the corresponding relationship between the final packaging parameters and the quality level, clarify whether the packaging meets the expected standards, providing a reliable basis for the final product quality control.

[0133] Among them, the encapsulation quality grade refers to the comprehensive evaluation result of the encapsulation quality of the pressure sensor based on the final encapsulation parameters and the preset quality evaluation criteria, which is usually presented in the form of grades, such as excellent, good, qualified, unqualified. This evaluation covers multiple aspects, including whether there are defects in the appearance of the encapsulated sensor, such as cracks and bubbles; whether the internal structure is stable and whether the connections of each component are tight; whether the electrical performance meets the standards, such as insulation resistance and signal transmission stability. The key quality factor refers to the key factor that plays a decisive role in the evaluation of the encapsulation quality grade. For example, the purity and stability of the encapsulation material. High-purity and stable materials can reduce signal interference and improve the accuracy and reliability of the sensor; the quality of the welding points. Firm and non-false-welded welding points can ensure stable electrical connections and affect the signal transmission quality; the temperature and pressure control accuracy in the encapsulation process. Precise temperature and pressure control can ensure the integrity and consistency of the encapsulation structure. Optionally, analyzing the encapsulation quality grade corresponding to the pressure sensor to be encapsulated can be achieved through the support vector machine (SVM) algorithm. For example, performance indicators such as the insulation resistance and response time of the sensor are used as inputs, and after calculation by the SVM model, its quality grade is output; extracting the key quality factors in the encapsulation quality grade can be achieved through the principal component analysis (PCA) tool. For example, for a data set containing multiple performance indicators and process parameters, after PCA analysis, several components with high variance contribution rates are extracted, and the indicators or parameters corresponding to these components (such as welding temperature, material hardness, etc.) are the key quality factors.

[0134] Furthermore, based on the key quality factors, the present invention generates an encapsulation scheme corresponding to the pressure sensor to be encapsulated, clarifies the key points affecting the encapsulation quality, can achieve precise encapsulation, enables the encapsulation process to closely focus on the key factors affecting the quality, effectively improves the encapsulation quality, and ensures the stable and reliable performance of the pressure sensor.

[0135] Among them, the encapsulation scheme refers to the encapsulation process used to plan and standardize the pressure sensor to be encapsulated. This scheme takes the key quality factors as the core basis and covers all aspects from material selection, process design, equipment selection to quality inspection. In terms of material selection, the most suitable encapsulation materials are selected according to the material performance requirements involved in the key quality factors, such as strength, insulation, and temperature resistance, to ensure the stable operation of the sensor in various environments; in the process design link, each step of the encapsulation process is precisely planned according to the key quality factors, and various process parameters are clarified, such as welding temperature, pressure, time, and the curing conditions of the encapsulation materials, to ensure that each operation can meet the quality requirements. When selecting equipment, refer to the requirements of the key quality factors for accuracy, efficiency, etc., and select equipment that can achieve high-quality encapsulation, such as high-precision dispensing machines, advanced curing furnaces, etc.; the quality inspection part also focuses on the key quality factors, formulates strict inspection standards and processes, conducts comprehensive inspections at different stages of encapsulation, discovers and solves possible quality problems in a timely manner, and ensures that the final product meets the high-quality encapsulation requirements. Optionally, the generation of the encapsulation scheme corresponding to the pressure sensor to be encapsulated can be realized through scheme generation tools, such as: tools like Project, Jira, etc.

[0136] Compared with the problems described in the background technology, the present invention can accurately grasp the physical characteristics and structural details of the sensor chip by obtaining the sensor chip image corresponding to the pressure sensor to be packaged and collecting the chip area data corresponding to the sensor chip image, which is helpful to identify potential problems of the chip and also lays a solid foundation for a series of operations such as packaging equipment matching and risk assessment based on image features. The present invention obtains packaging matching data by matching the chip feature data with the historical packaging data of the preset packaging equipment, which can quickly compare the compatibility of the chip and the packaging equipment to avoid packaging failure caused by mismatch; based on the successful cases and experience in the historical data, the best packaging parameters and solutions suitable for the current chip can be accurately screened to improve packaging efficiency and quality. Furthermore, based on the impact degree value, the present invention analyzes the packaging failure type corresponding to the preset packaging equipment, can accurately locate the root cause of the fault, and can quickly determine which dangers are caused by numerically quantifying the impact degree. The packaging parameters have a great interference with the real-time parameters, and then the possible fault types can be inferred, thereby improving the continuity and stability of packaging production. Furthermore, the present invention formulates the packaging planning target corresponding to the pressure sensor to be packaged based on the specific defect points, which can effectively avoid potential risks, and can adjust the packaging process and optimize the material selection in a targeted manner to ensure that the new packaging process avoids existing problems, which can improve the quality of the finished product of the pressure sensor to be packaged, and can also reduce the unqualified rate in subsequent inspections, improve production efficiency, and reduce production costs. Finally, the present invention adjusts the packaging parameters of the pressure sensor to be packaged based on the packaging operation instructions to obtain the final packaging parameters, which can ensure that the packaging parameters are highly consistent with the operation instructions, so that the packaging process is strictly carried out in accordance with the predetermined standards, effectively improve the packaging quality, and reduce packaging defects caused by unreasonable parameters. On the other hand, the packaging parameters can be adjusted to be more adapted to the actual packaging environment and equipment characteristics, thereby improving the efficiency and success rate of the packaging work. Therefore, a packaging method and system for realizing a pressure sensor provided by an embodiment of the present invention can improve the comprehensive measurement performance of the pressure sensor.

[0137] Embodiment 2:

[0138] like Figure 2 FIG. 1 is a functional module diagram of a packaging system for implementing a pressure sensor according to the present invention.

[0139] The packaging system 200 for implementing a pressure sensor of the present invention can be installed in an electronic device. According to the functions implemented, the packaging system for implementing a pressure sensor can include a feature annotation module 201, a degree value calculation module 202, a defect point query module 203, an instruction generation module 204, and a solution generation module 205. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0140] In the embodiments of the present invention, the functions of each module / unit are as follows:

[0141] The feature annotation module 201 is configured to obtain a sensing chip image corresponding to a pressure sensor to be encapsulated, collect chip region data corresponding to the sensing chip image, optimize the sensing chip image based on the chip region data to obtain an optimized chip image, and perform feature annotation on the optimized chip image to obtain chip feature data;

[0142] The degree value calculation module 202 is configured to match the chip feature data with historical encapsulation data of a preset encapsulation device to obtain encapsulation matching data, query potential risk data in the encapsulation matching data, screen out dangerous encapsulation parameters corresponding to the potential risk data, and monitor the real-time encapsulation parameters of the preset encapsulation device in real time, and calculate an influence degree value of the dangerous encapsulation parameters on the real-time encapsulation parameters;

[0143] The defect point query module 203 is configured to analyze an encapsulation failure type corresponding to a preset encapsulation device based on the influence degree value, locate an abnormal component in the preset encapsulation device based on the encapsulation failure type, perform performance detection on the abnormal component to obtain the actual performance of the component, and query specific defect points in the actual performance of the component;

[0144] The instruction generation module 204 is configured to formulate an encapsulation planning target corresponding to the pressure sensor to be encapsulated based on the specific defect points, analyze encapsulation measures in the encapsulation planning target, query encapsulation standards corresponding to the encapsulation measures, and generate an encapsulation operation instruction corresponding to the pressure sensor to be encapsulated based on the encapsulation standards;

[0145] The solution generation module 205 is configured to adjust the encapsulation parameters of the pressure sensor to be encapsulated based on the encapsulation operation instruction to obtain final encapsulation parameters, analyze an encapsulation quality level corresponding to the pressure sensor to be encapsulated based on the final encapsulation parameters, extract key quality factors in the encapsulation quality level, and generate an encapsulation solution corresponding to the pressure sensor to be encapsulated based on the key quality factors.

[0146] Specifically, each module in the encapsulation system 200 for implementing a pressure sensor in the embodiments of the present invention adopts the same technical means as those Figure 1 described in a method for implementing the encapsulation of a pressure sensor as described above, and can produce the same technical effects, which will not be elaborated here.

[0147] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0148] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A packaging method for implementing a pressure sensor, characterized in that, The method includes: Obtaining a sensing chip image corresponding to the pressure sensor to be encapsulated, collecting chip area data corresponding to the sensing chip image, optimizing the sensing chip image based on the chip area data to obtain an optimized chip image, and performing feature annotation on the optimized chip image to obtain chip feature data; Matching the chip feature data with historical encapsulation data of a preset encapsulation device to obtain encapsulation matching data, querying potential risk data in the encapsulation matching data, screening dangerous encapsulation parameters corresponding to the potential risk data, and real-time monitoring real-time encapsulation parameters of the preset encapsulation device, calculating an influence degree value of the dangerous encapsulation parameters on the real-time encapsulation parameters, where calculating the influence degree value of the dangerous encapsulation parameters on the real-time encapsulation parameters includes: Calculating the influence degree value of the dangerous encapsulation parameters on the real-time encapsulation parameters using the following formula: Among them, Ed represents the influence degree value of the dangerous encapsulation parameter on the real-time encapsulation parameter, s represents the number of parameters corresponding to the main dangerous encapsulation parameter, k represents the quantity index corresponding to the main dangerous encapsulation parameter, Tr k represents the real-time encapsulation parameter value corresponding to the k-th main dangerous encapsulation parameter, Td k represents the main danger threshold corresponding to the k-th main dangerous encapsulation parameter, C k represents the importance coefficient corresponding to the k-th main dangerous encapsulation parameter, t represents the number of parameters corresponding to the secondary dangerous encapsulation parameter, l represents the quantity index of the secondary dangerous encapsulation parameter, Sr l represents the real-time encapsulation parameter value corresponding to the l-th secondary dangerous encapsulation parameter, Sd l represents the secondary danger threshold corresponding to the l-th secondary dangerous encapsulation parameter; Analyzing an encapsulation failure type corresponding to the preset encapsulation device based on the influence degree value, locating an abnormal component in the preset encapsulation device based on the encapsulation failure type, performing performance detection on the abnormal component to obtain the actual performance of the component, and querying specific defect points in the actual performance of the component; Formulating an encapsulation planning goal corresponding to the pressure sensor to be encapsulated based on the specific defect points, analyzing encapsulation measures in the encapsulation planning goal, querying encapsulation standards corresponding to the encapsulation measures, and generating an encapsulation operation instruction corresponding to the pressure sensor to be encapsulated based on the encapsulation standards; Adjusting encapsulation parameters of the pressure sensor to be encapsulated based on the encapsulation operation instruction to obtain final encapsulation parameters, analyzing an encapsulation quality level corresponding to the pressure sensor to be encapsulated based on the final encapsulation parameters, extracting key quality factors in the encapsulation quality level, and generating an encapsulation scheme corresponding to the pressure sensor to be encapsulated based on the key quality factors.

2. The encapsulation method for implementing a pressure sensor according to claim 1, wherein, The optimizing the sensing chip image based on the chip area data to obtain an optimized chip image includes: Analyzing regional texture features in the chip area data; Performing texture partitioning on the sensing chip picture based on the regional texture features to obtain texture partitioning units; Identifying color distribution data in the texture partitioning units; Calculating a color adjustment coefficient corresponding to the color distribution data; Optimizing the sensing chip image based on the color adjustment coefficient to obtain an optimized chip image.

3. The encapsulation method for implementing a pressure sensor according to claim 2, characterized in that, The calculating the color adjustment coefficient corresponding to the color distribution data includes: Calculating the color adjustment coefficient corresponding to the color distribution data using the following formula: Among them, C adj represents the color adjustment coefficient corresponding to the color distribution data, m represents the number of dimensions of the color dimension corresponding to the color distribution data, j represents the number index corresponding to the color dimension, S j represents the sensitivity weight corresponding to the j-th color dimension, D j represents the measured value of the actual color distribution corresponding to the j-th color dimension, V j represents the variance value corresponding to the j-th color dimension, M j represents the adjustment limit value corresponding to the j-th color dimension.

4. A packaging method for implementing a pressure sensor according to claim 1, characterized in that, The querying potential risk data in the encapsulation matching data includes: Analyzing encapsulation environmental conditions corresponding to the encapsulation matching data; Querying abnormal environmental factors in the encapsulation environmental conditions; Sorting out factor risk categories corresponding to the abnormal environmental factors; Extracting specific risk cases in the factor risk categories; Querying potential risk data in the encapsulation matching data based on the specific risk cases.

5. A packaging method for implementing a pressure sensor as described in claim 1, characterized in that, The analyzing an encapsulation failure type corresponding to the preset encapsulation device based on the influence degree value includes: Analyze the contribution ratio of each dangerous packaging parameter corresponding to the influence degree value; Based on the contribution ratio, identify the core dangerous parameters corresponding to the influence degree value; Query the past failure modes associated with the core dangerous parameters in the historical failure database; Extract the packaging process characteristics in the past failure modes; Based on the packaging process characteristics, analyze the packaging failure types corresponding to the preset packaging equipment.

6. A packaging method for implementing a pressure sensor according to claim 1, characterized in that, The query of the specific defect points in the actual performance of the component includes: Sort out the performance dimensions corresponding to the actual performance of the component; Analyze the parameter fluctuation conditions in each dimension of the performance dimension; Based on the parameter fluctuation conditions, draw the performance change curve corresponding to the actual performance of the component; Identify the abnormal fluctuation points in the performance change curve; Based on the abnormal fluctuation points, query the specific defect points in the actual performance of the component.

7. A packaging method for implementing a pressure sensor as described in claim 1, characterized in that, Based on the specific defect points, formulate the packaging planning objectives for the pressure sensor to be packaged, including: Query the cause of the specific defect point; Analyze the associated components corresponding to the cause; Generate improvement measures corresponding to the associated components; Based on the improvement measures, determine the packaging improvement conditions for the pressure sensor to be packaged; Based on the packaging improvement conditions, formulate the packaging planning objectives for the pressure sensor to be packaged.

8. A packaging method for implementing a pressure sensor according to claim 1, characterized in that, Based on the packaging standard, generate the packaging operation instructions for the pressure sensor to be packaged, including: Analyze the standard detail items in the packaging standard; Extract the key item elements in the standard detail items; Based on the key item elements, construct the specific action process corresponding to the pressure sensor to be packaged; Analyze the key action descriptions in the specific action process; Based on the key action descriptions, generate the packaging operation instructions for the pressure sensor to be packaged.

9. A packaging system for implementing a pressure sensor, characterized in that, The system includes: A feature annotation module, configured to obtain a sensing chip image corresponding to a pressure sensor to be packaged, collect chip area data corresponding to the sensing chip image, optimize the sensing chip image based on the chip area data to obtain an optimized chip image, and perform feature annotation on the optimized chip image to obtain chip feature data; A degree value calculation module, configured to perform data matching on the chip feature data and the historical packaging data of a preset packaging device to obtain packaging matching data, query potential risk data in the packaging matching data, screen out dangerous packaging parameters corresponding to the potential risk data, and real-time monitor the real-time packaging parameters of the preset packaging device, and calculate the influence degree value of the dangerous packaging parameters on the real-time packaging parameters. Among them, the calculation of the influence degree value of the dangerous packaging parameters on the real-time packaging parameters includes: Use the following formula to calculate the influence degree value of the dangerous packaging parameters on the real-time packaging parameters: Among them, Ed represents the influence degree value of the dangerous encapsulation parameter on the real-time encapsulation parameter, s represents the number of parameters corresponding to the main dangerous encapsulation parameter, k represents the quantity index corresponding to the main dangerous encapsulation parameter, Tr k represents the real-time encapsulation parameter value corresponding to the k-th main dangerous encapsulation parameter, Td k represents the main danger threshold corresponding to the k-th main dangerous encapsulation parameter, C k represents the importance coefficient corresponding to the k-th main dangerous encapsulation parameter, t represents the number of parameters corresponding to the secondary dangerous encapsulation parameter, l represents the quantity index of the secondary dangerous encapsulation parameter, Sr l represents the real-time encapsulation parameter value corresponding to the l-th secondary dangerous encapsulation parameter, Sd l represents the secondary danger threshold corresponding to the l-th secondary dangerous encapsulation parameter; A defect point query module, configured to analyze the packaging failure types corresponding to a preset packaging device based on the influence degree value, locate abnormal components in the preset packaging device based on the packaging failure types, perform performance detection on the abnormal components to obtain the actual performance of the components, and query the specific defect points in the actual performance of the components; An instruction generation module, configured to formulate a packaging planning target corresponding to the pressure sensor to be packaged based on the specific defect points, analyze the packaging measures in the packaging planning target, query the packaging standards corresponding to the packaging measures, and generate a packaging operation instruction corresponding to the pressure sensor to be packaged based on the packaging standards; A solution generation module, configured to adjust the packaging parameters of the pressure sensor to be packaged based on the packaging operation instruction to obtain final packaging parameters, analyze the packaging quality level corresponding to the pressure sensor to be packaged based on the final packaging parameters, extract the key quality factors in the packaging quality level, and generate a packaging solution corresponding to the pressure sensor to be packaged based on the key quality factors.

Images