Chip welding method and system and storage device

Abstract

The invention provides a chip welding method and system and a storage device, and the method comprises the steps: obtaining a welding pressure data set, dividing the welding pressure data into a plurality of discrete pressure data segments through the discrete characteristics of the welding pressure data set, determining a pressure deviation coefficient of a target chip, obtaining a pre-welding image of the target chip, and obtaining a pre-welding image of the target chip; the method comprises the following steps: pre-welding a target chip pin, extracting a to-be-welded spot distribution vector of the target chip pin from a pre-welding image, and determining a welding gun offset during welding of the target chip through a pressure deviation coefficient and the to-be-welded spot distribution vector, thereby determining an argon flow domain of each to-be-welded spot of the target chip pin, and further determining a flow compensation value of each pin of the target chip. Therefore, according to all the flow compensation values and the welding gun offset, the variable welding flux amounts of the target chip pins at different positions are determined, all the pins of the target chip are heated and welded through all the variable welding flux amounts, and dynamic adjustment of the welding flux amounts in the chip welding process can be achieved.

Description

Technical Field

[0001] The present application relates to the field of welding technology, and more specifically, to a chip welding method, system, and storage device. Background Art

[0002] Welding technology is a process of heating metal or non-metal materials to high temperatures to partially or completely melt them, and using molten materials to fill or not fill the gaps, so that the materials form a dense connection after cooling. It is an important manufacturing process and is widely used in engineering fields such as construction, aerospace, automobile manufacturing, electronic equipment, etc. There are different types of welding methods including arc welding, gas shielded welding, laser welding, plasma welding, etc. Each method has its applicable occasions and characteristics. The application of welding technology can not only realize the connection of materials, but also repair, improve and process them, providing important technical support for all walks of life.

[0003] Chip welding is a key process connection technology in electronic manufacturing. It connects tiny chips to circuit boards to achieve signal transmission. This technology involves accurately fusing the chip pins to the pads on the circuit board. This fusion method usually requires the welding process to be completed in a high-temperature environment. However, during the chip welding process, the amount of solder input affects the chip welding quality. Too much solder will cause a short circuit between the welding positions, and too little solder will cause cold welding (the cold welding phenomenon is that the solder does not fully wet the welding surface, resulting in poor welding contact). Therefore, how to achieve dynamic adjustment of the solder amount during chip welding, thereby improving the chip welding quality has become a difficult problem faced by the industry. Summary of the invention

[0004] The present application provides a chip welding method, system, and storage device, which can realize dynamic adjustment of the solder amount during the chip welding process.

[0005] In a first aspect, the present application provides a chip welding method, comprising the following steps: Start chip welding and obtain the welding pressure data set when welding the chip in history; Dividing the welding pressure data into a plurality of discrete pressure data segments according to the discrete characteristics of the welding pressure data set, and determining the pressure deviation coefficient of the target chip according to all the discrete pressure data segments; Acquire a pre-welding image of a target chip, and extract a distribution vector of solder joints to be soldered of the pins of the target chip from the pre-welding image based on a pre-trained solder joint feature model; Determine the welding gun offset when welding the target chip by using the pressure deviation coefficient and the distribution vector of the points to be welded; Determine the argon flow domain of each to-be-welded point of the target chip pin based on the argon flow data set during the chip welding history and the distribution vector of the to-be-welded points, and then determine the flow compensation value of each pin of the target chip according to all the argon flow domains; Therefore, the variable solder amounts of the pins of the target chip at different positions are determined according to all the flow compensation values ​​and the welding gun offset, and the various pins of the target chip are heated and welded by all the variable solder amounts.

[0006] In some embodiments, dividing the welding pressure data into a plurality of discrete pressure data segments according to the discrete features of the welding pressure data set specifically includes: determining a plurality of welding pressure change rate sequences according to the welding pressure data set; Determining discrete features of the welding pressure data set through all welding pressure change rate sequences; Determining a feature partition coefficient of the welding pressure data set according to the discrete feature; The welding pressure data set is divided into a plurality of discrete pressure data segments by the characteristic division coefficient.

[0007] In some embodiments, determining the pressure deviation coefficient of the target chip according to all discrete pressure data segments specifically includes: Determine the pressure extreme value difference of each discrete pressure data segment; Determining a welding pressure equilibrium value of a target chip according to the welding pressure data set; The pressure deviation coefficient of the target chip is determined by all the pressure extreme value differences and the welding pressure equilibrium value.

[0008] In some embodiments, extracting the distribution vector of the solder joints to be soldered of the pins of the target chip from the pre-soldering image based on the pre-trained solder joint feature model specifically includes: Acquire a historical pre-welding image set of chips of the same type as the target chip; Training a feature model of the spot to be welded based on the historical pre-welding image set; The feature model of the points to be soldered is used to extract features from the pre-soldering image of the target chip to obtain a distribution vector of the points to be soldered of the pins of the target chip.

[0009] In some embodiments, determining the argon flow domain of each unwelded point of the target chip pin based on the argon flow data set during the historical chip welding and the unwelded point distribution vector specifically includes: Obtain the argon gas flow data set during chip welding in history; extracting an argon flow sequence from the argon flow data set; Selecting a first argon gas flow rate from the argon gas flow rate sequence; Determine the upper limit and the lower limit of the argon gas flow rate of the target chip pin to be welded corresponding to the argon gas flow rate through the distribution vector of the to-be-welded points; The range formed by the upper limit of the argon gas flow rate and the lower limit of the argon gas flow rate is used as the argon gas flow rate domain of the target chip pin to be welded corresponding to the argon gas flow rate.

[0010] Continue to determine the argon gas flow rate domain of the target chip pin's to-be-welded point corresponding to the remaining argon gas flow rate in the argon gas flow rate sequence.

[0011] In some embodiments, determining the flow compensation value of each pin of the target chip according to all argon gas flow domains specifically includes: Determine the flow compensation margin of the target chip according to all argon gas flow domains; The flow compensation value of each pin of the target chip is determined by the flow compensation margin.

[0012] In some embodiments, determining the variable solder amount at different positions of the target chip pin according to all flow compensation values ​​and the welding gun offset specifically includes: Determine the solder adjustment coefficient of the target chip pin at different positions according to all flow compensation values ​​and the welding gun offset; The variable solder amounts of the target chip pins at different positions are determined by all solder adjustment coefficients and the preset solder balance amount.

[0013] In a second aspect, the present application provides a chip bonding system, comprising: An acquisition module is used to acquire a welding pressure data set during chip welding in history after chip welding is started; A processing module, used to divide the welding pressure data into a plurality of discrete pressure data segments according to the discrete features of the welding pressure data set, and determine the pressure deviation coefficient of the target chip according to all the discrete pressure data segments; The processing module is further used to obtain a pre-welding image of the target chip, and extract a distribution vector of the to-be-welded points of the pins of the target chip from the pre-welding image based on a pre-trained solder point feature model; The processing module is further used to determine the welding gun offset when welding the target chip through the pressure deviation coefficient and the distribution vector of the points to be welded; The processing module is further used to determine the argon flow domain of each to-be-welded point of the target chip pin based on the argon flow data set when welding the chip in the history and the distribution vector of the to-be-welded point, and then determine the flow compensation value of each pin of the target chip according to all the argon flow domains; The execution module is used to determine the variable solder amount of the target chip pins at different positions according to all the flow compensation values ​​and the welding gun offset, and heat and weld each pin of the target chip through all the variable solder amounts.

[0014] In a third aspect, the present application provides a computer device, comprising a memory and a processor, wherein the memory stores codes, and the processor is configured to obtain the codes and execute the above-mentioned chip welding method.

[0015] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned chip welding method is implemented.

[0016] The technical solution provided by the embodiments disclosed in this application has the following beneficial effects: In the chip welding method, system, and storage device provided by the present application, chip welding is first started, a welding pressure data set when welding chips in history is obtained, the welding pressure data is divided into multiple discrete pressure data segments according to the discrete features of the welding pressure data set, the pressure deviation coefficient of the target chip is determined according to all the discrete pressure data segments, a pre-welding image of the target chip is obtained, and a distribution vector of the waiting points of the target chip pins is extracted from the pre-welding image based on a pre-trained solder point feature model, a welding gun offset when welding the target chip is determined according to the pressure deviation coefficient and the distribution vector of the waiting points, an argon flow domain of each waiting point of the target chip pin is determined based on the argon flow data set when welding chips in history and the distribution vector of the waiting points, and then a flow compensation value of each pin of the target chip is determined according to all the argon flow domains, thereby determining the variable solder amount of the target chip pins at different positions according to all the flow compensation values ​​and the welding gun offset, and heating and welding each pin of the target chip by all the variable solder amounts.

[0017] It can be seen that, firstly, the pressure deviation coefficient of the target chip is determined according to all discrete pressure data segments, wherein the pressure deviation coefficient reflects the fluctuation of pressure during chip welding, and then the pressure deviation coefficient is used to measure the welding gun offset during welding of the target chip; secondly, the maximum and minimum values ​​of the argon flow rate on each to-be-welded point of the target chip pin are determined through the argon flow data set, so as to limit the argon flow range (i.e., argon flow domain) of the to-be-welded point on each pin of the target chip, and then each argon flow domain is quantified to obtain a flow compensation value that can be used to compensate for the argon flow rate of the target chip pin; then, the welding gun offset and the flow compensation value of the target chip pin are combined to quantitatively limit the amount of solder required for each to-be-welded point of the target chip pin, thereby avoiding the adverse effects caused by excessive or insufficient use of solder, and then realizing dynamic adjustment of the solder amount during chip welding, thereby improving the chip welding quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is an exemplary flow chart of a chip bonding method according to some embodiments of the present application; Figure 2 is a schematic diagram of a process for determining a distribution vector of points to be welded in some embodiments of the present application; Figure 3 is a schematic diagram of a process for determining a variable solder amount in some embodiments of the present application; Figure 4 is a structural block diagram of a chip bonding system in some embodiments of the present application; Figure 5 It is a structural schematic diagram of a computer device for implementing a chip welding method according to some embodiments of the present application. DETAILED DESCRIPTION

[0019] The core of the present application is to start chip welding, obtain a welding pressure data set when welding chips in history, divide the welding pressure data into multiple discrete pressure data segments according to the discrete features of the welding pressure data set, determine the pressure deviation coefficient of the target chip according to all the discrete pressure data segments, obtain a pre-welding image of the target chip, extract the distribution vector of the waiting points of the target chip pin from the pre-welding image based on a pre-trained solder point feature model, determine the welding gun offset when welding the target chip according to the pressure deviation coefficient and the distribution vector of the waiting points, determine the argon flow domain of each waiting point of the target chip pin based on the argon flow data set when welding chips in history and the distribution vector of the waiting points, and then determine the flow compensation value of each pin of the target chip according to all the argon flow domains, thereby determining the variable solder amount of the target chip pin at different positions according to all the flow compensation values ​​and the welding gun offset, and heat and weld each pin of the target chip by all the variable solder amounts, so as to realize dynamic adjustment of the solder amount during chip welding.

[0020] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. Figure 1 , which is an exemplary flow chart of a chip bonding method according to some embodiments of the present application, the chip bonding method 100 mainly includes the following steps: In step 101, chip welding is started, and a welding pressure data set during chip welding in history is obtained.

[0021] In a specific implementation, after chip welding is started, a welding pressure data set during historical chip welding is obtained from a database of a welding system of a target chip.

[0022] It should be noted that the welding pressure data set described in the present application represents a collection of multiple welding pressure data, wherein the welding pressure data in the welding pressure data set represents a sequence of pressure values ​​collected by the pressure sensor on the welding gun when welding each chip pin, wherein during the collection, one chip pin is fixed and the pressure value applied by the welding gun to the chip pin is collected in a counterclockwise order, and each welding pressure data in the welding pressure data set is the welding pressure data obtained when welding a standard chip.

[0023] In step 102, the welding pressure data is divided into a plurality of discrete pressure data segments according to the discrete features of the welding pressure data set, and the pressure deviation coefficient of the target chip is determined according to all the discrete pressure data segments.

[0024] In some embodiments, dividing the welding pressure data into a plurality of discrete pressure data segments according to the discrete features of the welding pressure data set may be achieved by the following steps: determining a plurality of welding pressure change rate sequences according to the welding pressure data set; Determining discrete features of the welding pressure data set through all welding pressure change rate sequences; Determining a feature partition coefficient of the welding pressure data set according to the discrete feature; The welding pressure data set is divided into a plurality of discrete pressure data segments by the characteristic division coefficient.

[0025] In a specific implementation, a plurality of welding pressure change rate sequences are determined according to the welding pressure data set, that is, a welding pressure data set is selected from the welding pressure data set, a curve fitting is performed on the welding pressure data set, and the obtained curve is used as the pressure fitting curve of the welding pressure data set, and then the derivative is calculated for the position in the pressure fitting curve corresponding to each welding pressure value in the welding pressure data set, all the obtained derivatives are used as the welding pressure change rate, all the obtained welding pressure change rates are arranged in the order of arrangement of the welding pressure values ​​in the welding pressure data set, and the sequence obtained after the arrangement is used as the welding pressure change rate sequence of the welding pressure data set, and the welding pressure change rate sequence of the remaining welding pressure data in the welding pressure data set is continued to be determined.

[0026] It should be noted that the curve fitting described in this application can be implemented by different algorithms, such as polynomial fitting, least squares method, spline interpolation, etc. The purpose of curve fitting is to find the changing trend between the welding pressure values ​​in the welding pressure data.

[0027] In addition, it should be noted that the welding pressure change rate sequence in the present application represents a sequence composed of welding pressure change rates between adjacent welding pressure values ​​in the welding pressure data, wherein each welding pressure change rate represents the pressure change rate between adjacent chip pins during the chip welding process. The welding pressure gradient sequence can be used to measure the stability of the welding pressure during chip welding. The greater the welding pressure change rate, the worse the stability of the welding pressure.

[0028] In specific implementation, the discrete characteristics of the welding pressure data set are determined by all welding pressure change rate sequences, that is, the discrete characteristics represent the normalized numerical value of the overall discreteness of the welding pressure change rate sequence of each welding pressure data in the welding pressure data set. In some embodiments, the mean of the variance of each welding pressure gradient sequence can be used as the normalized numerical value of the discreteness of the welding pressure data set.

[0029] In a specific implementation, the characteristic division coefficient of the welding pressure data set is determined according to the discrete feature, that is, the discrete feature is raised to a negative exponential power with e as the base, and the value obtained after the negative exponential power is taken as the characteristic division coefficient of the welding pressure data set, that is, the characteristic division coefficient is , is a discrete feature.

[0030] It should be noted that the feature partition coefficients described in the present application represent coefficients for partitioning the welding pressure data set, and welding pressure data with different discrete features can be distinguished by the feature partition coefficients.

[0031] In a specific implementation, the welding pressure data set is divided into a plurality of discrete pressure data segments by the feature partition coefficient, that is: first, the feature partition coefficient is multiplied by 0.5, and the multiplied value is used as the feature partition center coefficient, and then the discrete feature is obtained for each welding pressure data in the welding pressure data set and then the negative exponential power is taken, and the value obtained by taking the negative exponential power is used as the discrete feature coefficient of the corresponding welding pressure data, and all welding pressure data whose discrete feature coefficients are less than or equal to the feature partition center coefficient are divided into the same part, and the welding pressure data constituting the part are used as a discrete pressure data segment, and all welding pressure data whose discrete feature coefficients are greater than the feature partition center coefficient and less than or equal to the feature partition coefficient are divided into the same part, and the welding pressure data constituting the part are used as a discrete pressure data segment, and all welding pressure data whose discrete feature coefficients are greater than the feature partition coefficient are divided into the same part, and the welding pressure data constituting the part are used as a discrete pressure data segment.

[0032] It should be noted that the discrete pressure data segments described in the present application represent welding pressure data with the same abnormal pressure fluctuations. By combining the welding pressure data with the same abnormal pressure fluctuations together, it is easier to detect changes in abnormal welding pressure values ​​on chip pins.

[0033] In some embodiments, determining the pressure deviation coefficient of the target chip according to all discrete pressure data segments can be achieved by using the following steps: Determine the pressure extreme value difference of each discrete pressure data segment; Determining a welding pressure equilibrium value of a target chip according to the welding pressure data set; The pressure deviation coefficient of the target chip is determined by all the pressure extreme value differences and the welding pressure equilibrium value.

[0034] In specific implementation, the pressure extreme value difference of each discrete pressure data segment is determined, that is, the maximum welding pressure value and the minimum welding pressure value are selected from each discrete pressure data segment, the maximum welding pressure value and the minimum welding pressure value are subtracted, and the value obtained by the difference is used as the pressure extreme value difference of the corresponding discrete pressure data segment; the welding pressure equilibrium value of the target chip is determined according to the welding pressure data set, that is, the average value of all welding pressure values ​​in the welding pressure data set is calculated, and the average value is used as the welding pressure equilibrium value of the target chip.

[0035] The pressure extreme value difference described in the present application can be used to measure the data volatility of a discrete pressure data segment. The larger the pressure extreme value difference is, the greater the data volatility of the discrete pressure data segment is.

[0036] In specific implementation, the pressure deviation coefficient of the target chip is determined by all the pressure extreme value differences and the welding pressure equilibrium value, that is, the average of all the pressure extreme value differences is calculated and divided by the welding pressure equilibrium value, and the value obtained by the division is used as the pressure deviation coefficient of the target chip.

[0037] It should be noted that the pressure deviation coefficient described in this application reflects the fluctuation of pressure during the chip welding process. If the pressure deviation coefficient is smaller, it means that the pressure change during the chip welding process is small, and the chip welding quality is relatively stable. If the pressure deviation coefficient is larger, it means that there is a large pressure fluctuation during the chip welding process, and the chip welding quality is unstable.

[0038] In step 103, a pre-welding image of the target chip is acquired, and a distribution vector of the to-be-welded points of the pins of the target chip is extracted from the pre-welding image based on a pre-trained solder point feature model.

[0039] In a specific implementation, the target chip is photographed by a microscope camera, and the photographed image is used as a pre-welding image of the target chip.

[0040] It should be noted that the pre-welding image in the present application is an image of the target chip being precisely placed on the pad of the printed circuit board (PCB) for pre-welding preparation. In addition, the type of the target chip belongs to the surface mount technology chip type, specifically including: integrated circuit chips, capacitor chips, crystal oscillator chips and memory chips, etc.

[0041] In some embodiments, reference Figure 2 As shown, this figure is a schematic diagram of the process of determining the distribution vector of the solder joints to be welded in some embodiments of the present application. In this embodiment, the distribution vector of the solder joints to be welded of the target chip pins is extracted from the pre-welding image based on the pre-trained solder joint feature model, which can be implemented by the following steps: First, in step 1031, a historical pre-welding image set of chips of the same type as the target chip is obtained; Secondly, in step 1032, a feature model of the spot to be welded is trained based on the historical pre-welding image set; Then, in step 1033, the feature model of the points to be soldered is used to extract features from the pre-soldering image of the target chip to obtain a distribution vector of the points to be soldered of the pins of the target chip.

[0042] In a specific implementation, a historical pre-welding image set of chips of the same type as the target chip is obtained, that is, multiple historical pre-welding images of chips of the same type as the target chip are obtained from a database of a chip welding system, and a collection of multiple historical pre-welding images is used as a historical pre-welding image set.

[0043] It should be noted that the historical pre-welding image described in this application is an image with detailed information marked, wherein the detailed information includes: the position coordinates, area, etc. of each soldering point of the chip pin, and the detailed information is organized into a vector, that is: each historical pre-welding image corresponds to a vector.

[0044] In a specific implementation, the feature model of the unwelded joint is trained based on the historical pre-welding image set, that is, first, the historical pre-welding image set is pre-processed, and the pre-processing includes: Gaussian filtering, perspective transformation, histogram equalization, etc. Gaussian filtering is mainly used to remove the influence of noise, perspective transformation is mainly used to correct the perspective distortion caused by the position deviation between the chip placement position and the camera, and histogram equalization is mainly used to improve the contrast between the target area of ​​the pre-welding image and the surrounding area. Then, the historical pre-welding image set is divided into a pre-welding image training set and a pre-welding image verification set according to the ratio of 70% and 30%, respectively, to ensure that the pre-welding image samples in each data set are evenly distributed, and the pre-welding image training set is used to train the residual neural network, and the parameters of the residual neural network are optimized by the back propagation algorithm. The pre-welding image verification set is used to monitor the training process, and the hyperparameters of the residual neural network are adjusted according to the pre-welding image verification set, such as: learning rate, batch size, etc. Finally, the residual neural network model after training is used as the feature model of the unwelded joint.

[0045] It should be noted that the feature model of the solder joints described in the present application refers to a model for extracting feature information of the solder joints of the chip pins in the pre-soldering image of the surface mount technology type chip.

[0046] In a specific implementation, the feature model of the points to be soldered is used to extract features from the pre-welding image of the target chip to obtain a distribution vector of the points to be soldered of the pins of the target chip, that is, the pre-welding image of the target chip is preprocessed in the same manner as the above operation, the preprocessed image is used as the input image, the feature model of the points to be soldered is used to forward propagate the input image, and the vector obtained by the forward propagation is used as the distribution vector of the points to be soldered of the pins of the target chip.

[0047] It should be noted that the solder point distribution vector described in the present application is composed of multiple solder point components, and the solder point components in the solder point distribution vector represent the position coordinates, area and other information of the chip pin solder points, wherein each solder point component corresponds to a chip pin solder point.

[0048] In step 104, the welding gun offset during welding of the target chip is determined by the pressure deviation coefficient and the distribution vector of the points to be welded.

[0049] In some embodiments, the welding gun offset when welding the target chip can be determined by the pressure deviation coefficient and the distribution vector of the welding points to be welded by the following steps: Determine the adjacent distances between chip pin solder points corresponding to each solder point component in the solder point distribution vector; The welding gun offset during welding of the target chip is determined according to the pressure deviation coefficient and the adjacent distances of all chip pins to be welded.

[0050] In specific implementation, the adjacent distances of the chip pins to be soldered corresponding to each of the points to be soldered components in the points to be soldered distribution vector are determined, that is, a point to be soldered component is selected from the points to be soldered distribution vector, and the distance is calculated according to the position coordinates of the chip pins to be soldered corresponding to the point to be soldered component and the position coordinates of the chip pins to be soldered corresponding to the remaining points to be soldered components, and the sum of the distances between the two pins to be soldered adjacent to the chip pins to be soldered corresponding to the point to be soldered component is taken as the adjacent distance of the chip pins to be soldered corresponding to the point to be soldered component, and the adjacent distances of the chip pins to be soldered corresponding to the remaining points to be soldered components in the points to be soldered distribution vector are continued to be determined.

[0051] In a specific implementation, the welding gun offset when welding the target chip is determined according to the pressure deviation coefficient and the adjacent distances of all the chip pins to be welded, that is, the average of the adjacent distances of all the chip pins to be welded is calculated, the pressure deviation coefficient is divided by the obtained average, and the value obtained by the division is used as the welding gun offset when welding the target chip.

[0052] It should be noted that the welding gun offset described in this application reflects the accuracy of the welding gun position during the chip welding process. The more accurate the welding gun position is, the more sufficient and uniform the contact between the welding point and the chip pin can be ensured, thereby improving the chip welding quality.

[0053] In step 105, the argon flow domains of each unwelded point of the target chip pin are determined based on the argon flow data set during the historical chip welding and the distribution vector of the unwelded points, and then the flow compensation value of each pin of the target chip is determined according to all the argon flow domains.

[0054] In some embodiments, the following steps may be used to determine the argon flow domain of each unwelded point of the target chip pin based on the argon flow data set during the historical chip welding and the unwelded point distribution vector: Obtain the argon gas flow data set during chip welding in history; extracting an argon flow sequence from the argon flow data set; Selecting a first argon gas flow rate from the argon gas flow rate sequence; Determine the upper limit and the lower limit of the argon gas flow rate of the target chip pin to be welded corresponding to the argon gas flow rate through the distribution vector of the to-be-welded points; The range formed by the upper limit of the argon gas flow rate and the lower limit of the argon gas flow rate is used as the argon gas flow rate domain of the target chip pin to be welded corresponding to the argon gas flow rate.

[0055] Continue to determine the argon gas flow rate domain of the target chip pin's to-be-welded point corresponding to the remaining argon gas flow rate in the argon gas flow rate sequence.

[0056] In the specific implementation, the argon gas flow data set during the historical chip welding is obtained from the database of the chip welding system.

[0057] It should be noted that the argon gas flow data set described in the present application represents a collection of multiple argon gas flow data, wherein the argon gas flow data in the argon gas flow data set represents a sequence of argon gas flow rates collected by the gas flow meter on the welding gun when welding each chip pin, wherein during the collection, a chip pin is fixed and the argon gas flow rate released on each chip pin is collected in a counterclockwise order. In addition, argon gas is generally used as a shielding gas to protect the welding area from the influence of oxygen and water vapor in the air. By adjusting and controlling the flow rate of argon gas, it can be ensured that the welding area is adequately protected.

[0058] In some embodiments, extracting the argon flow sequence from the argon flow data set may be achieved by using the following steps: determining an argon flow distribution entropy of the argon flow data set; The argon flow rate data set is feature extracted using the argon flow rate distribution entropy to obtain an argon flow rate sequence.

[0059] In a specific implementation, the argon flow distribution entropy of the argon flow data set is determined, that is: first, the numerical value in each argon flow data in the argon flow data set is taken as a random variable, and then the information entropy is used to calculate the entropy value of each argon flow data, and then the mean of the entropy values ​​of all the argon flow data is calculated, and the obtained mean is used as the argon flow distribution entropy of the argon flow data set. In some embodiments, the information entropy can be calculated in the form of cross entropy, and in other embodiments, other existing entropy calculation methods can also be used for calculation, which is not limited here.

[0060] It should be noted that the argon flow distribution entropy described in the present application represents a measure of the uncertainty of the argon flow distribution in the argon flow data set. The higher the flow distribution entropy, the more random the distribution of the argon flow between different values, and the greater the difference between the flow values. The lower the flow distribution entropy, the more concentrated the argon flow is near certain specific values, and the distribution is more concentrated and more certain.

[0061] In a specific implementation, the argon flow data set is feature extracted by using the argon flow distribution entropy to obtain an argon flow sequence, that is, argon flow data having an entropy value greater than the argon flow distribution entropy are selected from the argon flow data set, and then the median of each argon flow data at the same position is calculated for the selected argon flow data, all the medians are arranged in the order of collection, and the arranged order is used as the argon flow sequence, for example: S1=[2,2,3], S2=[3,2,4], S3=[4,2,5] are three selected argon flow data, the first position of S1, S2, and S3 is 2, 3, and 4, so the median is 3, the second position of S1, S2, and S3 is 2, 2, and the median is 2, the third position of S1, S2, and S3 is 3, 4, and 5, so the median is 4, so the sequence obtained by arranging in the order of collection is [3,2,4].

[0062] It should be noted that the argon flow distribution sequence is a sequence consisting of argon flow data selected according to the argon flow distribution entropy characteristics in the argon flow data set.

[0063] In specific implementation, the upper limit and lower limit of the argon flow rate of the target chip pin to be welded corresponding to the argon flow rate are determined by the distribution vector of the points to be welded, that is: the average of all the argon flow rates in the argon flow rate sequence is calculated, and the obtained value is used as the argon flow rate balance value; if the argon flow rate is greater than the argon flow rate balance value, the argon flow rate is multiplied by the area of ​​the chip pin corresponding to the point to be welded component in the distribution vector of the points to be welded, and the multiplied value is used as the upper limit of the argon flow rate of the chip pin to be welded corresponding to the point to be welded component; if the argon flow rate is less than or equal to the argon flow rate balance value, the argon flow rate balance value is multiplied by the area of ​​the chip pin corresponding to the point to be welded component in the distribution vector of the points to be welded, and the multiplied value is used as the lower limit of the argon flow rate of the chip pin to be welded corresponding to the point to be welded component.

[0064] It should be noted that the argon flow rate domain described in the present application represents the range consisting of the lower limit and upper limit of the argon flow rate during the chip welding process, wherein the lower limit of the argon flow rate represents the minimum allowable value of the argon flow rate on the chip pins during welding, and the upper limit of the argon flow rate represents the maximum allowable value of the argon flow rate on the chip pins during welding. If the argon flow rate is lower than the lower limit of the argon flow rate, the welding area will not be fully protected and problems such as oxidation and hydrogen intrusion will occur. If the argon flow rate exceeds the upper limit of the argon flow rate, it may cause problems such as airflow disturbance and increased sputtering, and will also increase argon consumption.

[0065] In some embodiments, determining the flow compensation value of each pin of the target chip according to all argon gas flow domains can be implemented by the following steps: Determine the flow compensation margin of the target chip according to all argon gas flow domains; The flow compensation value of each pin of the target chip is determined by the flow compensation margin.

[0066] In specific implementation, the flow compensation margin of the target chip is determined based on all argon flow domains, that is: the largest argon flow upper limit is selected from all argon flow domains, the smallest argon flow lower limit is selected from all argon flow domains, the largest argon flow upper limit is subtracted from the smallest argon flow lower limit, and the subtracted value is used as the flow compensation margin of the target chip.

[0067] It should be noted that the flow compensation margin described in the present application represents the additional margin reserved when controlling the argon flow rate during the chip welding process, and is used to measure the tolerance for argon flow fluctuations during the chip welding process. The larger the flow compensation margin, the higher the tolerance for argon flow fluctuations during the chip welding process.

[0068] In specific implementation, the flow compensation value of each pin of the target chip is determined by the flow compensation margin, that is: an argon flow domain is selected, the flow compensation margin is subtracted from the argon flow upper limit of the argon flow domain, and the value obtained by subtraction is used as the flow compensation value of the pin of the target chip corresponding to the argon flow domain, and the flow compensation value of the pin of the target chip corresponding to the remaining argon flow domains is continued to be determined.

[0069] It should be noted that the flow compensation value described in this application represents the parameter value for compensating the argon gas flow on the chip pin. By adjusting the flow compensation value, it can be ensured that the solder is evenly distributed during the chip pin welding process to avoid too much or too little solder.

[0070] In step 106, variable solder amounts at different positions of the pins of the target chip are determined according to all flow compensation values ​​and the welding gun offset, and each pin of the target chip is heated and welded by all variable solder amounts.

[0071] In some embodiments, reference Figure 3 As shown, this figure is a schematic diagram of the process of determining the variable solder amount in some embodiments of the present application. In this embodiment, the variable solder amount of the target chip pin at different positions can be determined according to all flow compensation values ​​and the welding gun offset, which can be achieved by the following steps: First, in step 1061, the solder adjustment coefficients of the target chip pins at different positions are determined according to all flow compensation values ​​and the welding gun offset; Next, in step 1062, the variable solder amounts of the target chip pins at different positions are determined by using all solder adjustment coefficients and the preset solder balance amount.

[0072] In specific implementation, the solder adjustment coefficients of the target chip pins at different positions are determined according to all the flow compensation values ​​and the welding gun offset, that is, a flow compensation value is selected, the flow compensation value is divided by the welding gun offset, and the value obtained by the division is used as the solder adjustment coefficient of the chip pin corresponding to the flow compensation value, and the solder adjustment coefficients of the chip pins corresponding to the remaining flow compensation values ​​are further determined.

[0073] It should be noted that the solder adjustment coefficient described in this application represents a parameter for adjusting the solder quantity and distribution at different positions of the target chip pins.

[0074] In specific implementation, the variable solder amount of the target chip pin at different positions is determined through all the solder adjustment coefficients and the preset solder balancing amount, that is: a solder adjustment coefficient is selected, and if the solder adjustment coefficient is greater than 0, the value obtained by adding the solder adjustment coefficient to the solder balancing amount is used as the variable solder amount of the chip pin corresponding to the solder adjustment coefficient; if the solder adjustment coefficient is less than or equal to 0, the value obtained by subtracting the solder balancing amount from the solder adjustment coefficient is used as the variable solder amount of the chip pin corresponding to the solder adjustment coefficient, and the variable solder amounts of the chip pins corresponding to the remaining solder adjustment coefficients are continued to be determined.

[0075] It should be noted that the variable solder amount in the present application refers to the amount of solder required to be used at different positions of the target chip pins.

[0076] In some embodiments, heating and soldering each pin of the target chip by using all variable solder amounts can be achieved by using the following steps: Obtaining the distribution vector of the points to be welded; Pulse heating welding is performed on each pin of the target chip according to the distribution vector of the points to be welded and all the variable solder amounts.

[0077] In specific implementation, pulse heating welding is performed on each pin of the target chip according to the distribution vector of the points to be soldered and all the variable solder amounts, that is: a point to be soldered component is selected from the distribution vector of the points to be soldered, and the solder is periodically pulse heated using a pulse heating device (such as a resistance heater, an induction heater, a microwave heater, etc.) until the solder amount reaches the variable solder amount of the chip pin corresponding to the point to be soldered component, and then the heating is stopped, and then a welding gun is used to align the heated solder amount with the position coordinates of the chip pin to be soldered corresponding to the point to be soldered component, and finally the heated solder amount is released to the chip pin to be soldered, and pulse heating welding is continued for the chip pin to be soldered corresponding to the remaining points to be soldered components in the distribution vector.

[0078] It should be noted that the solder described in this application may be solder, solder paste, solder glue or other welding materials.

[0079] In addition, in another aspect of the present application, in some embodiments, the present application provides a chip bonding system, referring to Figure 4 , which is a schematic diagram of exemplary hardware and / or software of a chip bonding system according to some embodiments of the present application, the chip bonding system 400 includes: an acquisition module 401, a processing module 402 and an execution module 403, which are described as follows: Acquisition module 401, in this application, acquisition module 401 is mainly used to acquire a welding pressure data set during chip welding in history after chip welding is started; Processing module 402, in the present application, the processing module 402 is used to divide the welding pressure data into a plurality of discrete pressure data segments according to the discrete features of the welding pressure data set, and determine the pressure deviation coefficient of the target chip according to all the discrete pressure data segments; It should be noted that the processing module 402 in the present application is also used to obtain a pre-welding image of the target chip, and extract the distribution vector of the to-be-welded points of the pins of the target chip from the pre-welding image based on a pre-trained solder point feature model; It should be noted that the processing module 402 in the present application is also used to determine the welding gun offset when welding the target chip through the pressure deviation coefficient and the distribution vector of the welding points to be welded; In addition, the processing module 402 in the present application is also used to determine the argon flow domain of each to-be-welded point of the target chip pin based on the argon flow data set during the historical chip welding and the distribution vector of the to-be-welded point, and then determine the flow compensation value of each pin of the target chip according to all the argon flow domains; Execution module 403, in this application, execution module 403 is mainly used to determine the variable solder amount of the target chip pins at different positions according to all flow compensation values ​​and the welding gun offset, and heat and weld each pin of the target chip through all variable solder amounts.

[0080] In addition, the present application also provides a computer device, which includes a memory and a processor, wherein the memory stores codes, and the processor is configured to obtain the codes and execute the above-mentioned chip welding method.

[0081] In some embodiments, reference Figure 5 , which is a schematic diagram of the structure of a computer device using a chip bonding method according to some embodiments of the present application. The chip bonding method in the above embodiment can be Figure 5 The computer device 500 shown in the figure is implemented, and the computer device 500 includes at least one processor 501, a communication bus 502, a memory 503 and at least one communication interface 504.

[0082] The processor 501 may be a general-purpose central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more processors for controlling the execution of the chip bonding method in the present application.

[0083] Communication bus 502 may include a pathway for transmitting information between the above-mentioned components.

[0084] The memory 503 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, an optical disc storage (including a compressed optical disc, a laser disc, an optical disc, a digital versatile disc, a Blu-ray disc, etc.), a magnetic disk or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of an instruction or data structure and can be accessed by a computer, but is not limited thereto. The memory 503 may exist independently and be connected to the processor 501 via the communication bus 502. The memory 503 may also be integrated with the processor 501.

[0085] The memory 503 is used to store the program code for executing the solution of the present application, and the execution is controlled by the processor 501. The processor 501 is used to execute the program code stored in the memory 503. The program code may include one or more software modules. The determination of the distribution vector of the welding points to be welded in the above embodiment can be implemented by the processor 501 and one or more software modules in the program code in the memory 503.

[0086] The communication interface 504 uses any transceiver or other device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0087] In a specific implementation, as an embodiment, a computer device may include multiple processors, each of which may be a single-CPU processor or a multi-CPU processor. The processor here may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0088] The above-mentioned computer device may be a general-purpose computer device or a special-purpose computer device. In a specific implementation, the computer device may be a desktop computer, a portable computer, a network server, a personal digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device or an embedded device. The embodiment of the present application does not limit the type of computer device.

[0089] In addition, the present application also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned chip welding method is implemented.

[0090] In summary, in the chip welding method, system, and storage device disclosed in the embodiments of the present application, chip welding is started, a welding pressure data set when welding chips in history is obtained, the welding pressure data is divided into multiple discrete pressure data segments according to the discrete features of the welding pressure data set, the pressure deviation coefficient of the target chip is determined according to all the discrete pressure data segments, a pre-welding image of the target chip is obtained, and a distribution vector of the to-be-welded points of the target chip pin is extracted from the pre-welding image based on a pre-trained solder point feature model, and a welding gun offset when welding the target chip is determined according to the pressure deviation coefficient and the distribution vector of the to-be-welded points. The argon flow domain of each to-be-welded point of the target chip pin is determined based on the argon flow data set when welding chips in history and the distribution vector of the to-be-welded points, and then the flow compensation value of each pin of the target chip is determined according to all the argon flow domains, so as to determine the variable solder amount of the target chip pin at different positions according to all the flow compensation values ​​and the welding gun offset, and each pin of the target chip is heated and welded by all the variable solder amounts, so as to realize dynamic adjustment of the solder amount during chip welding.

[0091] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0092] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A chip welding method, characterized in that: The steps include: Start chip welding and obtain the welding pressure data set when welding the chip in history; Dividing the welding pressure data into a plurality of discrete pressure data segments according to the discrete characteristics of the welding pressure data set, and determining the pressure deviation coefficient of the target chip according to all the discrete pressure data segments; Acquire a pre-welding image of a target chip, and extract a distribution vector of solder joints to be soldered of the pins of the target chip from the pre-welding image based on a pre-trained solder joint feature model; Determine the welding gun offset when welding the target chip by using the pressure deviation coefficient and the distribution vector of the points to be welded; Determine the argon flow domain of each to-be-welded point of the target chip pin based on the argon flow data set during the chip welding history and the distribution vector of the to-be-welded points, and then determine the flow compensation value of each pin of the target chip according to all the argon flow domains; Therefore, the variable solder amounts of the pins of the target chip at different positions are determined according to all the flow compensation values ​​and the welding gun offset, and the various pins of the target chip are heated and welded by all the variable solder amounts.

2. The method according to claim 1, characterized in that Dividing the welding pressure data into a plurality of discrete pressure data segments according to the discrete features of the welding pressure data set specifically includes: determining a plurality of welding pressure change rate sequences according to the welding pressure data set; Determining discrete features of the welding pressure data set through all welding pressure change rate sequences; Determining a feature partition coefficient of the welding pressure data set according to the discrete feature; The welding pressure data set is divided into a plurality of discrete pressure data segments by the characteristic division coefficient.

3. The method according to claim 1, characterized in that Determining the pressure deviation coefficient of the target chip based on all discrete pressure data segments specifically includes: Determine the pressure extreme value difference of each discrete pressure data segment; Determining a welding pressure equilibrium value of a target chip according to the welding pressure data set; The pressure deviation coefficient of the target chip is determined by all the pressure extreme value differences and the welding pressure equilibrium value.

4. The method according to claim 1, characterized in that Extracting the distribution vector of the solder joints of the target chip pins from the pre-soldering image based on the pre-trained solder joint feature model specifically includes: Acquire a historical pre-welding image set of chips of the same type as the target chip; Training a feature model of the spot to be welded based on the historical pre-welding image set; The feature model of the points to be soldered is used to extract features from the pre-soldering image of the target chip to obtain a distribution vector of the points to be soldered of the pins of the target chip.

5. The method according to claim 1, characterized in that Determining the argon flow domain of each to-be-welded point of the target chip pin based on the argon flow data set during the chip welding history and the distribution vector of the to-be-welded points specifically includes: Obtain the argon gas flow data set during chip welding in history; extracting an argon flow sequence from the argon flow data set; Selecting a first argon gas flow rate from the argon gas flow rate sequence; Determine the upper limit and the lower limit of the argon gas flow rate of the target chip pin to be welded corresponding to the argon gas flow rate through the distribution vector of the to-be-welded points; The range formed by the upper limit of the argon gas flow rate and the lower limit of the argon gas flow rate is used as the argon gas flow rate domain of the target chip pin to be welded corresponding to the argon gas flow rate; Continue to determine the argon gas flow rate domain of the target chip pin's to-be-welded point corresponding to the remaining argon gas flow rate in the argon gas flow rate sequence.

6. The method according to claim 1, characterized in that Determining the flow compensation value of each pin of the target chip based on all argon gas flow domains specifically includes: Determine the flow compensation margin of the target chip according to all argon gas flow domains; The flow compensation value of each pin of the target chip is determined by the flow compensation margin.

7. The method according to claim 1, characterized in that Determining the variable solder amount at different positions of the target chip pin according to all flow compensation values ​​and the welding gun offset specifically includes: Determine the solder adjustment coefficient of the target chip pin at different positions according to all flow compensation values ​​and the welding gun offset; The variable solder amounts of the target chip pins at different positions are determined by all solder adjustment coefficients and the preset solder balance amount.

8. A chip bonding system, characterized in that: include: An acquisition module is used to acquire a welding pressure data set during chip welding in history after chip welding is started; A processing module, used to divide the welding pressure data into a plurality of discrete pressure data segments according to the discrete features of the welding pressure data set, and determine the pressure deviation coefficient of the target chip according to all the discrete pressure data segments; The processing module is further used to obtain a pre-welding image of the target chip, and extract a distribution vector of the to-be-welded points of the pins of the target chip from the pre-welding image based on a pre-trained solder point feature model; The processing module is further used to determine the welding gun offset when welding the target chip through the pressure deviation coefficient and the distribution vector of the points to be welded; The processing module is further used to determine the argon flow domain of each to-be-welded point of the target chip pin based on the argon flow data set when welding the chip in the history and the distribution vector of the to-be-welded point, and then determine the flow compensation value of each pin of the target chip according to all the argon flow domains; The execution module is used to determine the variable solder amount of the target chip pins at different positions according to all the flow compensation values ​​and the welding gun offset, and heat and weld each pin of the target chip through all the variable solder amounts.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the chip bonding method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the chip bonding method according to any one of claims 1 to 7 are implemented.

Images