Packaging method and system for realizing pressure sensor

Abstract

The invention relates to the field of semiconductor packaging, and discloses a packaging method and system for realizing a pressure sensor, and the method comprises the steps: obtaining a sensing chip image of a to-be-packaged pressure sensor, collecting region data, carrying out the optimization processing and feature marking, obtaining chip feature data, carrying out the matching with preset packaging equipment historical data, querying potential risk data, and carrying out the packaging of the to-be-packaged pressure sensor. The method comprises the following steps: screening dangerous packaging parameters, monitoring real-time parameters, calculating an influence degree value, analyzing a packaging fault type, positioning abnormal component detection performance and finding defect points, formulating a packaging planning target according to the defect points, analyzing a packaging measure query standard to generate an operation instruction, adjusting the packaging parameters according to the operation instruction to obtain final parameters, evaluating a packaging quality grade and extracting key quality factors. And generating a packaging scheme. The comprehensive measurement performance of the pressure sensor can be improved.

Description

Technical Field

[0001] The present invention relates to a packaging method and system for realizing a pressure sensor, belonging 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 reveal 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, severe temperature changes, etc.), resulting in the sensitive components 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 thinner, lighter, and more multifunctional. Moreover, in the process of signal transmission, 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 image of the sensing chip corresponding to the pressure sensor to be packaged, collect the chip area data corresponding to the image of the sensing chip, optimize the image of the sensing chip 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] Based on the influence degree value, analyze the packaging fault type corresponding to the preset packaging equipment. Based on the packaging fault type, locate the abnormal component in the preset packaging equipment, perform a performance test on the abnormal component to obtain the actual performance of the component, and query the specific defect points in the actual performance of the component;

[0009] Based on the specific defect points, formulate the packaging planning goal corresponding to the pressure sensor to be packaged, analyze the packaging measures in the packaging planning goal, 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;

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

[0011] Optionally, the optimizing the sensing chip image 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] Based on the regional texture feature surface, perform texture partitioning on the sensing chip picture, with a texture partitioning unit;

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

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

[0016] Based on the color adjustment coefficient, optimize the sensing chip image 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] Use the following formula to calculate the color adjustment coefficient corresponding to the color distribution data:

[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 a color dimension indicating the measured value of the actual color distribution corresponding to the color dimension indicating the variance value corresponding to the color dimension indicating the adjustment limit value corresponding to the color dimension

[0021] Optionally, querying the potential risk data in the encapsulated matching data includes:

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

[0023] Querying the abnormal environmental factors in the encapsulation environment conditions;

[0024] Sorting out the factor risk categories corresponding to the abnormal environmental factors;

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

[0026] Based on the specific risk cases, querying the potential risk data in the encapsulated matching data.

[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 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 indicating the main dangerous threshold corresponding to the indicating the importance coefficient corresponding to the represents the number of parameters corresponding to the secondary dangerous encapsulation parameters, represents the number index corresponding to the secondary dangerous encapsulation parameters, indicating the real-time encapsulation parameter value corresponding to the 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 hazardous encapsulation parameter in the influence degree value;

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

[0034] Querying the past fault modes associated with the core hazardous 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 dimensions;

[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 target 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 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 target 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 annotation module, configured to obtain the sensing chip image corresponding to the pressure sensor to be encapsulated, collect the 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 annotation 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 the historical encapsulation data of a preset encapsulation device to obtain encapsulation matching data, query the potential risk data in the encapsulation matching data, filter the dangerous encapsulation parameters corresponding to the potential risk data, and real-time monitor the real-time encapsulation parameters of the preset encapsulation device, and calculate the influence degree value of the dangerous encapsulation parameters on the real-time encapsulation parameters;

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

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

[0061] Compared with the problems in the background art, the present invention can accurately master the physical characteristics and structural details of the sensing chip 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, 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 the 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 judge 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 targeted, 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 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. 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 an encapsulation system for implementing 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 in conjunction with the embodiments with reference to the accompanying drawings. Detailed implementation manners

[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 encapsulation of a pressure sensor. The execution subject of the method for implementing the encapsulation 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 encapsulation 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] Referring to Figure 1 As shown, it is a flowchart of a method for implementing the encapsulation of a pressure sensor provided by an embodiment of the present invention. In this embodiment, the method for implementing the encapsulation of a pressure sensor includes:

[0069] S1. Obtain the sensing chip image corresponding to the pressure sensor to be encapsulated, collect the chip area data corresponding to the sensing chip image, based on the chip area data, perform optimization processing on the sensing chip image 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 encapsulated 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 encapsulation 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, due to the lack of encapsulation, its sensitive element is directly exposed and is easily affected by external factors such as humidity, corrosive gases, and temperature changes, making it unable to 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 image of the sensing chip 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 information such as the shape, size, positional relationship of each component, the circuit pattern on the chip surface, and the layout of micro-components of the sensing chip; the chip area data refers to the data set extracted from the image of the sensing chip that can reflect the physical properties 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. 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 image of the sensing chip 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 image of the sensing chip 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, separating the chip from the background to obtain accurate chip area data.

[0072] Furthermore, based on the chip area data, the present invention optimizes the image of the sensing chip 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 fine structure and potential defects of the chip.

[0073] Among them, the optimized chip image refers to the final image obtained after optimizing the image of the sensing chip 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 the saturation, contrast, and brightness of the color, enabling 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 plane, 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 status of attributes such as the brightness and saturation of colors. The color adjustment coefficient refers to a set of parameters calculated based on 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 analyzed chip area data can be implemented through texture feature extraction algorithms. 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 achieved 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 achieved 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 the 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] Calculating the color adjustment coefficient corresponding to the color distribution data by using the following formula:

[0079]

[0080] 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 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 a 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 belonging to a specific value range of a certain color dimension in the image, 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 to which the human eye is more sensitive, 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 obtains chip feature data by performing feature annotation on the optimized image of the chip, 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, and help improve the overall quality and performance of the pressure sensor packaging.

[0083] Among them, the chip feature data refers to a 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 functional information of each module. It is a 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 implemented 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 incompatibility; 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 final packaging completion, etc.; the packaging matching data refers to a set of corresponding data 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, equipment 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 data 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 rate of key features of the chip in the image can be enhanced, which is helpful for subsequent feature labeling and analysis work.

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

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

[0090] Among them, the encapsulation environmental conditions refer to various conditions related to the physical, chemical, etc. environments during the encapsulation of the pressure sensor, including but not limited to environmental parameters such as temperature, humidity, cleanliness, air pressure, etc., and factors such as the electromagnetic environment and vibration conditions of the encapsulation site; the abnormal environmental factors refer to those factors in the encapsulation environmental conditions that deviate from the normal range or standard. For example, the temperature exceeds a specific interval suitable for chip encapsulation, the high humidity causes the chip to be affected by moisture, the non-compliance of cleanliness allows dust and other impurities to mix into the encapsulation process, and the abnormal air pressure affects the normal operation of the encapsulation equipment. These are all abnormal environmental factors; the factor risk categories refer to the classification of the risks that may be caused by the abnormal environmental factors according to their natures and types. For example, the abnormal temperature can be classified into the heat-related risk category, which may cause changes in chip performance or deformation of the encapsulation material; the abnormal humidity can be classified into the moisture risk category, with risks such as causing chip short circuits or corrosion; the abnormal cleanliness belongs to the impurity contamination risk category, which is likely to cause problems such as encapsulation defects; the specific risk cases refer to the actual risk events caused by specific abnormal environmental factors in the past encapsulation history. For example, during a certain period, due to the failure of the air conditioner in the encapsulation workshop, the temperature exceeded the normal range, resulting in the phenomenon of virtual soldering of the solder joints of the encapsulated chip; or on a certain occasion, due to the failure of humidity control in the encapsulation environment, the internal circuit of the chip was corroded, and the final product was unqualified and other specific event records.

[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 the abnormal environment 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 the 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 the specific risk cases in the factor risk category can be achieved through the text matching algorithm, such as the edit distance algorithm, and when the factor risk category keyword is input, the most matching specific risk case is found by calculating the edit distance between the keyword and the case category description in the database; the query of the 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 cluster in the clustering result.

[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 encapsulation parameters refer to those parameters that are very likely to cause serious problems such as encapsulation failure and product quality decline once they appear during the encapsulation process. These parameters can be derived from the analysis of potential risk data, such as too high or too low temperature, pressure beyond the safe range, too fast or too slow encapsulation speed, etc. These data deviate greatly from the normal encapsulation parameter range, directly affecting the bonding effect between the chip and the encapsulation 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 encapsulation parameters refer to the operating parameters generated and fed back in real time by the preset encapsulation equipment during the actual encapsulation operation. By monitoring these parameters in real time, operators can intuitively understand whether the encapsulation process is proceeding normally, such as whether the equipment temperature is within the normal operating range, whether the pressure is stable near the specified value, etc. Optionally, the screening of the dangerous encapsulation 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 encapsulation parameter data as input, the model is allowed to learn the feature representation of normal data. The parameters corresponding to the data points with large reconstruction errors are regarded as dangerous encapsulation parameters. The real-time monitoring of the real-time encapsulation parameters of the preset encapsulation 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. 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 encapsulation parameters in real time.

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

[0095] Among them, the influence degree value refers to the value that comprehensively reflects the influence of the dangerous encapsulation parameters on the real-time encapsulation parameters, intuitively understanding the degree to which the current real-time encapsulation parameters are affected by the dangerous encapsulation parameters, and then judging the severity of potential risks during the encapsulation 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 encapsulation parameters on the real-time encapsulation parameters includes:

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

[0098]

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

[0100] Specifically, the main danger 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 the main danger 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 danger threshold refers to the critical value set for each main danger 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 danger encapsulation parameter of encapsulation temperature, its main danger 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 k-th main danger 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 danger encapsulation parameter on the encapsulation result; the secondary danger 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 danger encapsulation parameters. For example, parameters such as humidity and air pressure in the encapsulation environment are usually regarded as secondary danger encapsulation parameters; the secondary danger threshold refers to the critical value set for each secondary danger 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 danger threshold, the secondary danger threshold is also determined based on factors such as the characteristics of the encapsulation process and 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 encapsulation fault types corresponding to the preset encapsulation 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 encapsulation 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 the encapsulation production.

[0103] Among them, the encapsulation fault type refers to the fault category that will occur in the preset encapsulation equipment finally determined according to the analysis of the encapsulation process characteristics. Common encapsulation fault types include mechanical faults (such as encapsulation 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 encapsulation defects caused by aging and deterioration of encapsulation materials), etc.

[0104] As an embodiment of the present invention, the analysis of the encapsulation fault types corresponding to the preset encapsulation equipment based on the influence degree value includes: analyzing the contribution ratio of each dangerous encapsulation parameter in the influence degree value; identifying the core dangerous parameters corresponding to the influence degree value based on the contribution ratio; querying the past fault modes associated with the core dangerous parameters in the historical fault database; extracting the encapsulation process characteristics in the past fault modes; and analyzing the encapsulation fault types corresponding to the preset encapsulation equipment based on the encapsulation 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 is 40% in the calculation of the influence degree value, it means 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 found through calculation 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 historical failure database and are associated with the core dangerous parameters. It includes various information at the time of failure, such as the time of 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 materials, the heating and cooling rates during the packaging process, etc. all belong to the packaging process characteristics.

[0106] Further, the contribution ratio corresponding to each dangerous encapsulation parameter in the parsed influence degree value can be realized through a contribution degree analysis algorithm based on Monte Carlo simulation. For example, by randomly changing the values of each dangerous encapsulation 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 through a threshold screening algorithm. For example, a threshold of the contribution ratio is set, and when the contribution ratio of a certain dangerous encapsulation 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 through 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 encapsulation process features in the past failure modes can be realized through a topic model algorithm. For example, algorithms such as LDA, taking the text data of the past failure modes as input, and through the LDA algorithm, mining the potential topics therein, and these topics often correspond to different encapsulation process features. What method, tool or algorithm can be used to analyze the encapsulation failure types corresponding to the preset encapsulation equipment? For example, according to the corresponding rules between the encapsulation process features and the known failure types, logical reasoning is carried out to judge the encapsulation failure types corresponding to the preset encapsulation equipment.

[0107] Based on the encapsulation failure types, the present invention locates the abnormal components in the preset encapsulation 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 the encapsulation production.

[0108] Among them, the abnormal components refer to the components in the preset encapsulation equipment that cause the equipment to have encapsulation failures 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 deviation of the encapsulation action 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 functional modules such as heating and refrigeration, which will affect the temperature environment required for encapsulation when they fail. Optionally, the location of the abnormal components in the preset encapsulation equipment can be realized through 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, by performing performance detection on the abnormal component, the actual performance of the component is obtained, which can accurately determine the specific performance deviation of the abnormal component, provide accurate data support for subsequent repair or replacement, avoid blind operation, help to deeply analyze the root cause of the fault, and provide a basis for optimizing the equipment design and improving the process.

[0110] Among them, the actual performance of the component refers to the current actual working characteristics and ability performance of the component obtained after performing performance detection on the abnormal component. This covers multiple key dimensions, such as the actual rotation speed, torque, wear degree, and vibration amplitude of mechanical components; the actual resistance, capacitance, inductance values, as well as current and voltage parameters of electrical components, and the accuracy and stability of signal transmission. These performance data intuitively reflect the current operating state of the abnormal component. Compared with the performance indicators in the normal working state, the performance deviation of the component can be clearly presented. Optionally, the performance detection of the abnormal component can be achieved through the simulated working condition test method. For example, under the actual working environment and working condition conditions of the simulated preset packaging equipment, the abnormal component is tested, and its operating performance and performance indicators under the simulated working conditions are observed, so as to obtain the actual performance of the component.

[0111] Furthermore, by querying the specific defect points in the actual performance of the component, the root cause of the problem can be accurately located, providing a clear direction for subsequent repair, replacement or adjustment, helping to deeply analyze the cause of the defect, providing a valuable basis for optimizing the component design and improving the production process, and thus improving the component quality and the overall performance of the equipment.

[0112] Among them, the specific defect point refers to the fault location or damaged part that actually exists on the component and causes performance abnormality. For example, through the analysis of the vibration performance change curve, it is found that the vibration is extremely abnormal at a certain moment, and 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 querying of the specific defect points in the actual performance of the component includes: sorting out the performance dimensions corresponding to the actual performance of the component; analyzing the parameter fluctuation conditions in each dimension of the performance dimension; based on the parameter fluctuation conditions, drawing the performance change curve corresponding to the actual performance of the component; identifying the abnormal fluctuation points in the performance change curve; and based on the abnormal fluctuation points, querying the specific defect points 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, which are 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.; for electrical components, they 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 changes such as increases, decreases, or sudden fluctuations in speed 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 values 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, with time as the abscissa and recording the temperature parameter values of a component during operation, the curve of temperature change 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, which 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] Furthermore, 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 tools like Origin and Excel. 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 objectives refer 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 objectives corresponding to the pressure sensor to be packaged based on the specific defect points includes: querying the causes corresponding to the specific defect points; analyzing the associated components corresponding to the causes; 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 objectives corresponding to the pressure sensor to be packaged based on the packaging improvement conditions.

[0120] Among them, the cause of generation 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, 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 of generation and the associated components. If the cause of generation is insufficient equipment precision, the improvement measures can be to calibrate, repair, or replace key components of the equipment; if there are problems with 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, querying the cause corresponding to the specific defect point can be achieved through a fault tree reverse reasoning algorithm. For example, in a fault tree model, through logical gate relationships, starting from the defect point of "abnormal output signal of the pressure sensor", it can be traced back to the basic events such as "loose soldering of the welding point" and "damage to the internal circuit of the chip", which are the causes; analyzing the associated components corresponding to the causes can be achieved through an association mining algorithm based on a knowledge graph. For example, constructing a graph containing knowledge such as pressure sensor components, processes, and faults, taking the causes as nodes, and using the association relationship mining algorithm of the graph to find the associated component nodes connected to them, and these components are the associated components; generating improvement measures corresponding to the associated components can be achieved through a case-based reasoning (CBR) algorithm. For example, when encountering a new problem of associated components, through a similarity matching algorithm, retrieving similar cases 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; determining the packaging improvement conditions corresponding to the pressure sensor to be packaged can be achieved through a 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, finding the parameter combination that optimizes the performance of the pressure sensor; formulating the packaging planning goal corresponding to the pressure sensor to be packaged can be achieved through a goal programming algorithm. For example, among multiple goals such as cost, quality, and efficiency, determining the weights of each goal according to the enterprise strategy, and solving the optimal packaging planning goal under certain resource constraints through the goal programming algorithm.

[0122] By analyzing the packaging measures in the packaging planning goal and querying the packaging standards corresponding to the packaging measures, the present invention can ensure the scientificity and rationality of the packaging measures, make them comply with industry specifications and requirements, and avoid 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 goals 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, encapsulation or sealing; and applying specific encapsulation processes, such as welding, bonding and other processes to effectively encapsulate the various components 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 goals 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 deteriorates due to the aging of the encapsulation materials. Through FMEA analysis, encapsulation measures such as strengthening the anti-aging performance of the encapsulation materials can be identified; 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 standard", 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 to ensure that the operator can accurately complete the encapsulation work according to the instructions and ensure 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 detailed rules entries in the encapsulation standard; extracting the key entry elements from the standard detailed rules 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 detailed rules entries refer to the specific contents of the detailed regulations on all aspects of the encapsulation process in the encapsulation standard, which cover from the selection specifications of encapsulation materials, such as the chemical composition of the materials and the requirements for physical performance indicators; to the specific parameter settings of the encapsulation process, like the numerical ranges 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 detailed rules 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 about the sequence of each operation link, etc.; the specific action process refers to disassembling 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 preparatory work, such as cleaning the pins of the pressure sensor to be encapsulated, to the intermediate core encapsulation actions, such as soldering the chip and filling the encapsulation materials, 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 entries in the encapsulation standard can be implemented through a syntactic analysis algorithm, such as: dependency syntactic analysis, which can analyze the dependency relationships between words in the text to help understand the semantic structure of the standard sentences, thereby accurately parsing out the detail entries; the extraction of the key entry elements in the standard detail entries can be implemented through the TF-IDF algorithm tool, such as: the sklearn library in Python, which can calculate the importance of each word in the detail entries and screen out important keywords as the key entry elements; the construction of the specific action process corresponding to the pressure sensor to be encapsulated can be implemented through the Petri net algorithm, such as: the operation steps, conditions, and events in the encapsulation process can be represented by the elements of the Petri net, thereby constructing the specific action process; the analysis of the key action descriptions in the specific action process can be implemented through the association rule mining algorithm, such as: analyzing the association relationships between actions, finding the action combinations closely related to the encapsulation quality and performance, and determining the key action descriptions from them. For another example, it is found that there is a strong association between "accurate chip installation position" and "firm welding", and the descriptions of these actions are the key action descriptions; the generation of the encapsulation operation instructions corresponding to the pressure sensor to be encapsulated can be implemented through the natural language generation algorithm, such as: a series of models such as GPT based on the Transformer architecture. By inputting the key action descriptions and relevant information, the model can generate a natural and fluent encapsulation operation instruction text. For another example, inputting the key action description "accurately place the pressure sensor chip on the encapsulation base and gently fix it with tweezers" into the model generates a detailed operation instruction statement.

[0129] S5. 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.

[0130] Based on the encapsulation operation instructions, the present invention adjusts the encapsulation parameters of the pressure sensor to be encapsulated to obtain the final encapsulation parameters, which can ensure a high degree of fit between the encapsulation parameters and the operation instructions, make the encapsulation process strictly follow the predetermined standards, effectively improve the encapsulation quality, and reduce the encapsulation defects caused by unreasonable parameters. On the other hand, through adjustment, the encapsulation parameters can be better adapted to the actual encapsulation environment and equipment characteristics, improving the efficiency and success rate of the encapsulation work.

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

[0132] Furthermore, based on the final encapsulation parameters, the present invention analyzes the corresponding encapsulation quality level of the pressure sensor to be encapsulated, extracts the key quality factors in the encapsulation quality level, can accurately evaluate the encapsulation quality, and clarifies whether the encapsulation meets the expected standard according to the correspondence between the final encapsulation parameters and the quality level, 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. It 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, taking performance indicators such as the insulation resistance and response time of the sensor as inputs, and after calculation by the SVM model, outputting its quality grade. 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 plan refers to the plan for planning and standardizing the encapsulation process of the pressure sensor to be encapsulated. This plan is centered around key quality factors and covers all aspects from material selection, process design, equipment selection to quality inspection. In material selection, for the material performance requirements involved in key quality factors, such as strength, insulation, temperature resistance, etc., the most suitable encapsulation materials are selected to ensure the stable operation of the sensor in various environments. In the process design link, based on key quality factors, each operation in the encapsulation process is precisely planned, and various process parameters are clarified, such as welding temperature, pressure, time, and the curing conditions of encapsulation materials, to ensure that each operation can meet the quality requirements. When selecting equipment, referring to the requirements of key quality factors for aspects such as accuracy and efficiency, equipment capable of achieving high-quality encapsulation is selected, such as high-precision dispensing machines, advanced curing furnaces, etc. The quality inspection part also focuses on key quality factors, formulates strict inspection standards and procedures, 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, generating the encapsulation plan corresponding to the pressure sensor to be encapsulated can be achieved through plan generation tools, such as tools like Project and Jira.

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

[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 the 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 in Figure 1 a method for implementing the encapsulation of a pressure sensor described above, and can produce the same technical effects, which will not be elaborated here.

[0147] It is obvious to those skilled in the art 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 realizing a pressure sensor, characterized in that: The method comprises: Acquire a sensor chip image corresponding to the pressure sensor to be packaged, and collect chip area data corresponding to the sensor chip image, optimize the sensor chip image based on the chip area data to obtain a chip optimized image, and perform feature annotation on the chip optimized image to obtain chip feature data; Match the chip feature data with 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, 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; Based on the impact degree value, analyzing the packaging fault type corresponding to the preset packaging equipment, locating the abnormal component in the preset packaging equipment based on the packaging fault type, performing performance detection on the abnormal component to obtain the actual performance of the component, and querying the specific defect point in the actual performance of the component; Based on the specific defect point, formulate a packaging planning target corresponding to the pressure sensor to be packaged, analyze the packaging measures in the packaging planning target, query the packaging standard corresponding to the packaging measure, and generate a packaging operation instruction corresponding to the pressure sensor to be packaged based on the packaging standard; Based on the packaging operation instructions, the packaging parameters of the pressure sensor to be packaged are adjusted to obtain final packaging parameters; based on the final packaging parameters, the packaging quality grade corresponding to the pressure sensor to be packaged is analyzed, and the key quality factors in the packaging quality grade are extracted; based on the key quality factors, a packaging solution corresponding to the pressure sensor to be packaged is generated.

2. A packaging method for implementing a pressure sensor according to claim 1, characterized in that: The step of optimizing the sensor chip image based on the chip area data to obtain a chip optimized image includes: Analyzing regional texture features in the chip regional data; Based on the regional texture feature surface, texture partitioning is performed on the sensor chip image, a texture partitioning unit; identifying color distribution data in the texture partition unit; Calculating a color adjustment coefficient corresponding to the color distribution data; Based on the color adjustment coefficient, the sensor chip image is optimized to obtain a chip optimized image.

3. A packaging method for implementing a pressure sensor as claimed in claim 2, characterized in that: As an embodiment of the present invention, the calculating of the color adjustment coefficient corresponding to the color distribution data includes: The color adjustment coefficient corresponding to the color distribution data is calculated using the following formula: ； in, 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, Indicates the quantity index corresponding to the color dimension, Indicates The sensitivity weight corresponding to the color dimension, Indicates The actual color distribution measurement value corresponding to the color dimension, Indicates The variance value corresponding to the color dimension is Indicates The adjustment limit value corresponding to each color dimension.

4. A packaging method for implementing a pressure sensor according to claim 1, characterized in that: The querying of the potential risk data in the package matching data includes: Analyzing the packaging environment conditions corresponding to the packaging matching data; Querying abnormal environmental factors in the packaging environmental conditions; Sort out the risk categories of the factors corresponding to the abnormal environmental factors; Extract specific risk cases within the risk category of the factor; Based on the specific risk case, the potential risk data in the package matching data is queried.

5. A packaging method for implementing a pressure sensor according to claim 1, characterized in that: The calculating the influence degree value of the dangerous packaging parameter on the real-time packaging parameter includes: The following formula is used to calculate the influence of the dangerous packaging parameters on the real-time packaging parameters: ； in, Indicates the influence degree of the dangerous packaging parameter on the real-time packaging parameter, Indicates the number of parameters corresponding to the main dangerous package parameters, Indicates the quantity index corresponding to the main dangerous package parameters, Indicates The real-time package parameter values ​​corresponding to the main dangerous package parameters, Indicates The main hazard threshold corresponding to the main hazard packaging parameters, Indicates The importance coefficients of the main hazardous packaging parameters are: Indicates the number of parameters corresponding to the secondary dangerous package parameters, Indicates the number of minor dangerous package parameters index, Indicates The real-time package parameter value corresponding to the minor dangerous package parameter, Indicates The secondary danger threshold corresponding to the secondary danger package parameter.

6. A packaging method for implementing a pressure sensor according to claim 1, characterized in that: The analyzing, based on the impact degree value, the packaging fault type corresponding to the preset packaging equipment includes: Analyze the contribution ratio of each dangerous packaging parameter in the impact degree value; Based on the contribution ratio, identifying the core risk parameter corresponding to the impact degree value; Querying a historical fault database for past failure modes associated with the core hazardous parameters; Extracting packaging process features in the past failure mode; Based on the packaging process characteristics, the packaging failure type corresponding to the preset packaging equipment is analyzed.

7. A packaging method for implementing a pressure sensor according to claim 1, characterized in that: The querying of specific defects in the actual performance of the component includes: Sort out the performance dimensions corresponding to the actual performance of the components; Analyzing parameter fluctuations in each dimension of the performance dimension; Based on the parameter fluctuation, a performance change curve corresponding to the actual performance of the component is drawn; Identifying abnormal fluctuation points in the performance change curve; Based on the abnormal fluctuation points, the specific defect points in the actual performance of the component are queried.

8. A packaging method for implementing a pressure sensor according to claim 1, characterized in that: The step of formulating a packaging planning target corresponding to the pressure sensor to be packaged based on the specific defect point includes: Query the cause of the specific defect; Analyze the associated components corresponding to the cause; Generating improvement measures corresponding to the associated components; Based on the improvement measures, determining packaging improvement conditions corresponding to the pressure sensor to be packaged; Based on the packaging improvement conditions, a packaging planning target corresponding to the pressure sensor to be packaged is formulated.

9. A packaging method for implementing a pressure sensor according to claim 1, characterized in that: The step of generating a packaging operation instruction corresponding to the pressure sensor to be packaged based on the packaging standard includes: Parsing the standard details in the packaging standard; Extracting key elements from the standard details; Based on the key entry elements, construct a specific action flow corresponding to the pressure sensor to be packaged; Analyze the key action descriptions in the specific action flow; Based on the key action description, a packaging operation instruction corresponding to the pressure sensor to be packaged is generated.

10. A packaging system for implementing a pressure sensor, characterized in that: The system comprises: A feature annotation module is used to obtain a sensor chip image corresponding to the pressure sensor to be packaged, and collect chip area data corresponding to the sensor chip image, optimize the sensor chip image based on the chip area data to obtain a chip optimized image, and perform feature annotation on the chip optimized image to obtain chip feature data; A degree value calculation module is used to match the chip feature data with the historical packaging data of the preset packaging equipment 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 equipment in real time to calculate the influence degree value of the dangerous packaging parameters on the real-time packaging parameters; A defect point query module is used to analyze the packaging fault type corresponding to the preset packaging equipment based on the impact degree value, locate the abnormal component in the preset packaging equipment based on the packaging fault type, perform performance detection on the abnormal component, obtain the actual performance of the component, and query the specific defect point in the actual performance of the component; An instruction generation module, used to formulate a packaging planning target corresponding to the pressure sensor to be packaged based on the specific defect point, analyze the packaging measures in the packaging planning target, query the packaging standard corresponding to the packaging measure, and generate a packaging operation instruction corresponding to the pressure sensor to be packaged based on the packaging standard; A scheme generating module is used to adjust the packaging parameters of the pressure sensor to be packaged based on the packaging operation instructions to obtain final packaging parameters, analyze the packaging quality grade corresponding to the pressure sensor to be packaged based on the final packaging parameters, extract the key quality factors in the packaging quality grade, and generate a packaging scheme corresponding to the pressure sensor to be packaged based on the key quality factors.

Images