A method for detecting and analyzing the consistency of bosses on a PCB board

Abstract

This invention belongs to the field of printed circuit board (PCB) manufacturing and inspection technology, and relates to a method for detecting and analyzing the consistency of PCB bosses. This invention acquires boss height data, divides the board surface into regions and grid cells, analyzes the variation trends of height characteristic values ​​in the X and Y directions and the distribution patterns of adjacent cells, identifies systematic spatial gradients and regional clustering anomalies, and constructs a consistency judgment framework for anomaly detection. When an anomaly is detected, based on the identified gradient and anomaly pattern, combined with reverse compensation rules and coordinate mapping relationships, a process parameter compensation matrix is ​​generated and converted into equipment instructions. This method solves the problem of difficulty in systematically identifying spatial patterns of height deviations and real-time feedback control in traditional inspection, achieving rapid and accurate diagnosis of boss height consistency and adaptive process compensation, thus improving PCB manufacturing yield and process stability.

Description

Technical Field

[0001] This invention belongs to the field of printed circuit board manufacturing and testing technology, and relates to a method for testing and analyzing the consistency of PCB board bosses. Background Technology

[0002] In the field of high-density interconnection and advanced packaging of printed circuit boards (PCBs), the height consistency of bosses such as pad bumps and pressure bosses is a key quality indicator that determines the reliability of subsequent surface mount, chip packaging, or electromechanical interconnection. Excessive fluctuations in boss height can lead to a series of process problems such as poor soldering, stress concentration, and unstable signal transmission.

[0003] Currently, the industry commonly uses point cloud or point-by-point measurement data to inspect PCB boss consistency. This method has the following drawbacks: Existing methods typically provide an overall statistical description of all boss heights. While this provides a macroscopic assessment of whether the entire board meets tolerance specifications, it ignores the intrinsic relationship between individual measurement data and their spatial location on the board. In actual production, many process deviations exhibit a systematic trend of continuous variation along a specific direction or region. However, this trend is easily masked by random noise in the overall statistics, making it impossible to effectively identify and locate, thus hindering accurate detection of height consistency.

[0004] Secondly, regarding the identification of localized and clustered defects. For example, localized stencil blockage, localized dispensing head abnormalities, or uneven material distribution may cause multiple adjacent bosses within a rectangular area to be collectively higher or lower than normal. Such regional clustered defects pose a greater threat to assembly quality, but if their average value remains within the overall tolerance, they are easily overlooked by the overall pass rate index, resulting in missed defects.

[0005] More importantly, existing testing and analysis results typically only output pass / fail judgments or simple statistical reports. Even when out-of-tolerance conditions are identified, the process mainly relies on engineers' experience to make qualitative adjustments to process parameters, lacking a quantitative, automated, spatial model-based compensation decision-making mechanism. The absence of a precise mathematical mapping between process adjustments and detected spatial anomalies leads to low adjustment efficiency, high trial-and-error costs, and difficulty in achieving rapid and accurate closed-loop control of the production process. Summary of the Invention

[0006] In view of this, in order to solve the problems mentioned in the background technology, a PCB board boss consistency detection and analysis method is proposed.

[0007] The objective of this invention can be achieved through the following technical solution: a PCB board boss consistency detection and analysis method, comprising: S1, calculating the overall height statistics based on the height dataset of all bosses on the PCB board.

[0008] S2. Based on the process layout, divide the board surface into multiple analysis areas and obtain their average height. Divide the board surface into grid-like spatial units by rows and columns, and calculate the height characteristic value of each unit.

[0009] S3. Analyze the variation trend of the height characteristic values ​​of each unit in the X and Y directions, identify the systematic spatial gradient of height deviation, analyze the distribution pattern of the height characteristic values ​​of adjacent units, and identify regional clustering anomalies.

[0010] S4. Based on the overall height statistics, systematic spatial gradient, and regional clustering anomalies, an anomaly judgment is made by constructing a consistency judgment framework.

[0011] S5. When an anomaly is detected, based on the systematic spatial gradient and regional clustered anomalies, combined with the reverse compensation rules and the coordinate mapping relationship between the control unit and the spatial unit, a process parameter compensation matrix is ​​generated and converted into equipment instructions for issuance.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) By dividing the PCB board surface into grid-like spatial units and analyzing the variation trend of height feature values ​​in the X and Y directions, the present invention can effectively identify the systematic gradient of height deviation as the spatial position changes continuously. This method solves the problem that the existing overall statistical description masks the spatial trend, so that the progressive process deviation caused by equipment tilt, uneven temperature field, etc. can be accurately located and quantified, thereby improving the pertinence and interpretability of the detection.

[0013] (2) By analyzing the distribution pattern of height characteristic values ​​of adjacent spatial units, this invention can group consecutively exceeding the standard areas into sets for identification. This method solves the problem of regional cluster defects being missed because the average value is within acceptable limits, and enables the effective detection of clustered height anomalies caused by local process abnormalities, thereby providing a more comprehensive assessment of the board surface quality.

[0014] (3) This invention constructs a consistency judgment framework and automatically generates a process parameter compensation matrix based on the identified spatial anomaly patterns, combined with reverse compensation rules and coordinate mapping relationships. This method solves the problems of relying on manual experience for adjustment and lacking a quantitative decision-making mechanism, and realizes closed-loop automation from detection to compensation, significantly improving the accuracy and efficiency of process adjustment and reducing trial and error costs. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating the steps of a PCB board boss consistency detection and analysis method according to the present invention.

[0017] Figure 2 This is a flowchart illustrating the process of identifying regional clustering anomalies in this invention.

[0018] Figure 3 This is a flowchart illustrating the specific method S4 in this invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1 As shown, the present invention provides a PCB board boss consistency detection and analysis method, including: S1, calculating the overall height statistics based on the height dataset of all bosses on the PCB board.

[0021] Considering that the height consistency of PCB board bosses directly affects the assembly accuracy and electrical performance stability of electronic components, it is necessary to first evaluate the distribution of boss heights from an overall perspective.

[0022] Therefore, in one specific embodiment, the actual height data of each boss is collected by a height detection device such as a laser altimeter to form a structured height dataset, and noise suppression processing is performed, such as moving average filtering or median filtering and other digital signal processing algorithms.

[0023] The calculation of the overall height statistic specifically includes: taking the difference between the maximum and minimum values ​​of the height dataset as the overall coplanarity error, which is used to characterize the overall fluctuation range of the boss height.

[0024] Its standard deviation is used as the height dispersion to reflect the degree of dispersion of boss height; the proportion of bosses whose height values ​​fall within the preset tolerance zone is used as the tolerance compliance rate to reflect the proportion of bosses that meet the process requirements.

[0025] Specifically, determining the range of the preset tolerance zone involves: determining the initial tolerance range based on the nominal value of the boss height specified in the PCB design documents and the industry standard IPC-6012; measuring the actual height distribution data of the bosses through trial production of 30-50 samples; and statistically analyzing the collected data in conjunction with the reliability requirements of the product application scenario to calculate the upper and lower limits of the 99.5% confidence interval as the benchmark for tolerance compliance rate calculation.

[0026] S2. Based on the process layout, divide the board surface into multiple analysis areas and obtain their average height. Divide the board surface into grid-like spatial units by rows and columns, and calculate the height characteristic value of each unit.

[0027] Because the production process layout of regional PCB boards, such as the lamination area, exposure area, and adhesive spraying path, varies, the height deviation of the bosses has regional characteristics. If only the overall statistics are used for evaluation, it will mask the abnormal fluctuations within different process areas, thus failing to effectively capture defects caused by local equipment status or material unevenness. Furthermore, by regularly discretizing the entire board surface in the X and Y directions and dividing it into grid-like spatial units, a refined spatial representation of height characteristics on a two-dimensional plane can be achieved, providing structured data for S3 to identify systematic spatial gradients and regional clustering anomalies.

[0028] In one specific embodiment, the specific method of S2 is as follows: the PCB board surface is divided into multiple analysis areas such as edge area, center area, and pin-dense area according to the production process layout, and the average height of each analysis area is obtained.

[0029] The PCB board surface is divided into columns in the X direction and rows in the Y direction to form a grid-like spatial unit. The method for determining the number of grid rows and columns includes: calculating the effective analysis length in the X and Y directions based on the external dimensions of the PCB board, and determining the required spatial resolution based on the minimum control accuracy of the process equipment.

[0030] By comparing the number of anomalies detected and the data processing time under different grid densities through comparative experiments, the ratio of the increase in anomaly detection rate to the increase in computation time was calculated, and finally the number of grid rows and columns that maximized this ratio was selected.

[0031] For each spatial cell, a height characteristic value is calculated based on the average height of the analysis region associated with that cell.

[0032] The method for determining the association relationship is as follows: a spatial unit is assigned to the analysis region where its geometric center point is located; when the geometric center point of a spatial unit falls exactly on the common boundary of two or more analysis regions, the overlapping area of ​​the spatial unit with each adjacent analysis region is calculated, and it is assigned to the analysis region with the largest overlapping area.

[0033] When a spatial unit spans two or more analysis regions due to its size or location, the overlapping area between the spatial unit and each analysis region is directly calculated, and the unit is assigned to the analysis region with the largest overlapping area.

[0034] Furthermore, the method for obtaining the height feature value is as follows: for spatial units belonging to the same analysis region, calculate the relative deviation between the height values ​​of all protrusions in the unit and the average height of the associated analysis region, and use the relative deviation as the height feature value of the spatial unit.

[0035] When there are no protrusions in a spatial unit, its height characteristic value is determined by interpolation. It is calculated by bilinear interpolation based on the height characteristic values ​​of four adjacent spatial units with protrusions. If it is located at the edge and there are fewer than four adjacent units, the arithmetic mean of the height characteristic values ​​of the available adjacent units is taken.

[0036] This height feature value reflects the overall height deviation characteristics of each spatial unit relative to its analysis region, providing basic data for subsequent identification of systematic spatial gradients and regional clustering anomalies.

[0037] S3. Analyze the variation trend of the height characteristic values ​​of each unit in the X and Y directions, identify the systematic spatial gradient of height deviation, analyze the distribution pattern of the height characteristic values ​​of adjacent units, and identify regional clustering anomalies.

[0038] Considering that if only the overall height statistics are used for consistency assessment, although it can reflect the central tendency and dispersion of the height of the board bosses, it will ignore the overall spatial distribution information of the PCB board, resulting in the inability to identify the specific distribution pattern of the height deviation on the two-dimensional plane.

[0039] Specifically, there are two typical spatial anomaly patterns: one is a systematic gradual change along the X or Y direction. For example, due to the temperature gradient of the heating plate of the pressing equipment, the height of the boss shows a gradual increase or decrease from one side of the plate to the other. The other is a concentrated exceedance in local adjacent areas. For example, due to an abnormality in the solder paste application system in a certain area, the amount of solder paste applied is too much, causing the height of multiple adjacent bosses in that area to exceed the process limit.

[0040] These two anomalous patterns, which have clear directional or regional characteristics, cannot be identified effectively and independently through overall statistics.

[0041] In one specific embodiment, the systematic spatial gradient for identifying height deviation specifically involves: comparing the change trend of height characteristic values ​​of each column of spatial units in the X direction of the PCB board, and comparing the change trend of height characteristic values ​​of each row of spatial units in the Y direction.

[0042] Traverse all columns in the X direction and compare the height feature values ​​of adjacent columns in turn according to the column number. If three or more consecutive columns show an increasing or decreasing trend, that is, the height feature value of the next column is always greater than or always less than the previous column, then it is determined that there is a systematic spatial gradient with height deviation in the X direction.

[0043] Using the same logic, traverse all rows in the Y direction. If three or more consecutive rows show an increasing or decreasing trend, then it is determined that there is a systematic spatial gradient in the Y direction.

[0044] Further, please refer to Figure 2 As shown, the identification of regional clustering anomalies specifically involves: traversing all spatial units and determining whether the height feature value of each unit is greater than a preset process upper limit or less than a preset process lower limit.

[0045] The method for determining the preset upper and lower limits of the process is as follows: read the standard value of the boss height in the PCB design file, then collect historical production data, calculate the production process capability index CPK according to the statistical process control standard method, and when the CPK value meets the set requirements, determine the upper and lower limit parameters of the process control based on the standard value and the CPK calculation results.

[0046] For a cell that meets the above conditions, find its directly adjacent cells in the X or Y direction. If the height characteristic values ​​of the adjacent cells satisfy the same size relationship, that is, both exceed the upper limit or both are below the lower limit, then these cells are grouped into a set. Each independent set is a regional clustering anomaly. Record the spatial coordinate range of the set and the number of cells it contains.

[0047] S4. Based on the overall height statistics, systematic spatial gradient, and regional clustering anomalies, an anomaly judgment is made by constructing a consistency judgment framework.

[0048] Considering that the consistency judgment of protrusions needs to take into account both global indicators and local abnormal features, relying solely on a single dimension can easily lead to misjudgment, such as the overall indicators being qualified but there being local cluster anomalies, or there being a systematic gradient but the overall indicators meeting the standards.

[0049] Therefore, a hierarchical decision-making logic is constructed. In one specific embodiment, please refer to [link to relevant documentation]. Figure 3 As shown, the specific method of S4 is: to construct a consistency determination framework, which includes a first-level determination condition and a second-level determination condition.

[0050] The first-level judgment criteria include: identifying the presence of a systematic spatial gradient on the board surface or the presence of regional clustering anomalies; when any of the first-level judgment criteria is met, the PCB board boss consistency is directly determined to be abnormal.

[0051] If none of the first-level judgment conditions are met, the second-level judgment is then executed.

[0052] The second-level judgment criteria include: the tolerance compliance rate meets the process requirements, the overall coplanarity error meets the process requirements, and the height dispersion meets the process requirements; when all the second-level judgment criteria are met, the PCB board boss consistency is judged to be qualified; when any second-level judgment criterion is not met, the PCB board boss consistency is judged to be abnormal.

[0053] The process for determining the tolerance compliance rate requirements involves obtaining the minimum acceptable standard for boss height compliance rate specified in the product quality agreement or customer standard for that PCB model. If there are no non-standard requirements, the obtained standard value is set as the compliance rate threshold for judgment, referring to the acceptable level in industry-standard data.

[0054] It should be explained that the core principle of the constructed hierarchical judgment framework is that systematic spatial gradients and regional clustering anomalies reflect structural process deviations related to spatial location. Such deviations pose a direct risk to the coplanarity and connection reliability of the final assembly, and their anomalies may be weakened in the overall height statistics due to cancellation with other random fluctuations, thus leading to missed detection. Therefore, identifying such spatial anomaly patterns is used as the first-level judgment condition. Once the first-level condition is triggered, it is judged as a consistency anomaly.

[0055] When none of the first-level judgment conditions are met, it indicates that there are no significant spatial structural anomalies on the PCB board surface, and the height fluctuation is more likely to be randomly distributed. In this case, the judgment proceeds to the second level, where three indicators are used to comprehensively evaluate the overall compliance and uniformity of the manufacturing results from three dimensions: the proportion of conforming specifications, the maximum fluctuation range, and the degree of dispersion.

[0056] The purpose of setting the second level of judgment is to ensure that the product meets the basic size specifications and process stability standards, based on the confirmation that there are no spatial abnormalities. Only when all the second level conditions are met will the PCB board boss consistency be finally judged to be qualified.

[0057] The process for determining the overall coplanarity error process requirements is as follows: Extract the maximum allowable boss height difference from the PCB design documents or product assembly technical requirements to ensure reliable electrical connections and smooth mechanical assembly, and directly set this allowable value as the upper limit of the overall coplanarity error process requirements.

[0058] The process for determining the required high dispersion is as follows: collect historical production data of multiple batches of the product under stable process conditions, calculate the standard deviation of the boss height of each batch, apply statistical process control methods, analyze the distribution of these standard deviations, and add 3 times the standard deviation of the distribution to the mean of the distribution as the upper limit of the process control for high dispersion.

[0059] S5. When an anomaly is detected, based on the systematic spatial gradient and regional clustered anomalies, combined with the reverse compensation rules and the coordinate mapping relationship between the control unit and the spatial unit, a process parameter compensation matrix is ​​generated and converted into equipment instructions for issuance.

[0060] When the height of a boss exhibits a systematic spatial gradient or regional clustering anomaly, if the direction of the compensation command generated accordingly is improperly set—for example, if it follows the same trend as the identified deviation—it will not only fail to correct the error but may even amplify the original systematic deviation. Therefore, reverse compensation logic must be employed to ensure the effectiveness of the correction action.

[0061] In addition, the physical control units of process equipment, such as pressure heads and nozzle arrays, often have a non-one-to-one correspondence with the spatial units on the PCB board. Directly compensating independently by unit will lead to fragmentation of control commands and conflicts with the actuators.

[0062] Therefore, by establishing reverse compensation rules and coordinate mapping relationships, a precise conversion from the detection space to the execution space was achieved.

[0063] In one specific embodiment, the method for constructing the reverse compensation rule and the coordinate mapping relationship between the control unit and the spatial unit is as follows: the compensation direction is set according to the identified systematic spatial gradient direction. When the height feature value in the X direction shows a monotonically increasing trend, the X direction process parameter compensation is set to decrease; when the height feature value in the X direction shows a monotonically decreasing trend, the X direction process parameter compensation is set to increase; the setting logic for the Y direction compensation direction is the same.

[0064] Obtain the spatial layout coordinate data of the process equipment control unit, match the center coordinates of each grid-like spatial unit with the center coordinates of the control unit, and establish a physical mapping relationship.

[0065] When a single control unit covers multiple spatial units, the adjustment amount of the control unit is calculated by weighting the deviation of the height characteristic value of each spatial unit it covers. The weighting coefficient is negatively correlated with the Euclidean distance from the center of the spatial unit to the center of the control unit.

[0066] For example, the specific method for setting the weight coefficient is as follows: the weight coefficient of each spatial unit is inversely proportional to the Euclidean distance from its center point to the center of the control unit, and then after normalization, the sum of the weight coefficients of all relevant spatial units is 1.

[0067] Furthermore, the generation of the process parameter compensation matrix specifically includes: for each control unit, determining the direction of the process parameter adjustment command based on the projection value of the planar coordinates of its corresponding spatial unit onto the identified systematic spatial gradient direction, wherein the direction of the adjustment command is strictly consistent with the compensation direction.

[0068] The deviation between the height characteristic value of the corresponding spatial unit of each control unit and the process target value specified in the PCB design document is calculated. The selection of the process target value depends on the type of systematic spatial gradient or regional clustering anomaly. Specifically, when the identified systematic spatial gradient trend is that the height characteristic value increases monotonically in a certain direction, or when the regional clustering anomaly is that the height characteristic value is collectively higher than the preset process upper limit, the lower limit of the process range is selected as the process target value to guide the compensation action to reduce the boss height.

[0069] When the identified systematic spatial gradient trend is that the height feature value decreases monotonically in a certain direction, or when the regional clustering anomaly is that the height feature value is collectively lower than the preset process lower limit, the upper limit of the process range is selected as the process target value to guide the compensation action to increase the height of the boss.

[0070] When there is no significant systematic spatial gradient or regional clustering anomaly on the board surface, but the overall statistics do not meet the requirements and uniformity fine-tuning is necessary, the median value of the process range is selected as the process target value.

[0071] It should be noted that when both systematic spatial gradients and regional clustering anomalies exist on the same plate surface, the compensation direction and process target value should be determined first based on the direction of the systematic spatial gradient to ensure that the global trend deviation is corrected first.

[0072] All control units are arranged in descending order of the absolute value of the deviation. Based on the descending order, the magnitude of the process parameter adjustment commands is assigned. Control units with smaller sequence numbers are assigned higher process parameter adjustment levels.

[0073] Arrange the adjustment commands of all control units in spatial coordinate order to form an N-row, M-column matrix structure, and generate a process parameter compensation matrix, where N is the number of control units in the X direction and M is the number of control units in the Y direction.

[0074] Furthermore, the process parameter compensation matrix is ​​transformed into equipment instructions, including: encoding the adjustment direction and magnitude in the process parameter compensation matrix into serialized instructions that can be parsed and executed by the equipment controller according to the equipment control protocol.

[0075] In summary, this invention effectively establishes a mapping relationship between boss height data and spatial position by dividing the PCB board surface into grid-like spatial units and calculating the height feature value of each unit, thus solving the technical problem of ignoring spatial position correlation in the prior art; by analyzing the changing trend of height feature values ​​in the X and Y directions, it realizes the identification of systematic spatial gradients and overcomes the defect of overall statistical description masking spatial trends.

[0076] By constructing a consistency judgment framework that includes two levels of judgment conditions, the overall statistical indicators and spatial distribution characteristics are comprehensively considered, avoiding misjudgments caused by single-dimensional judgment. By establishing reverse compensation rules and coordinate mapping relationships between control units and spatial units, spatial anomaly patterns are transformed into quantitative process parameter adjustment instructions, realizing automatic closed-loop control from detection to compensation, which significantly improves the efficiency and accuracy of process adjustment.

[0077] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0078] Those skilled in the art will recognize that the algorithmic steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0079] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0080] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0081] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting and analyzing the consistency of bosses on a PCB board, characterized in that, include: S1. Calculate the overall height statistics based on the height dataset of all bosses on the PCB board; S2. Based on the process layout, divide the board surface into multiple analysis areas and obtain their average height. Divide the board surface into grid-like spatial units by rows and columns, and calculate the height characteristic value of each unit. S3. Analyze the variation trend of the height characteristic values ​​of each unit in the X and Y directions, identify the systematic spatial gradient of height deviation, analyze the distribution pattern of the height characteristic values ​​of adjacent units, and identify regional clustering anomalies. S4. Based on the overall height statistics, systematic spatial gradient, and regional clustering anomalies, an anomaly judgment is made by constructing a consistency judgment framework; S5. When an anomaly is detected, based on the systematic spatial gradient and regional clustered anomalies, combined with the reverse compensation rules and the coordinate mapping relationship between the control unit and the spatial unit, a process parameter compensation matrix is ​​generated and converted into equipment instructions for issuance. The method for constructing the reverse compensation rule and the coordinate mapping relationship between the control unit and the spatial unit is as follows: Based on the identified systematic spatial gradient direction, when the height characteristic value in the X direction shows a monotonically increasing trend, the X-direction process parameter compensation is set to decrease; when the height characteristic value in the X direction shows a monotonically decreasing trend, the X-direction process parameter compensation is set to increase; the setting logic for the Y-direction compensation direction is the same. Acquire the spatial layout coordinate data of the process equipment control unit, and establish a physical mapping relationship between each grid-like spatial unit on the PCB board and the spatial coordinates of the control unit; When a single control unit covers multiple spatial units, the adjustment amount of the control unit is calculated by weighting the deviation of the height characteristic value of each spatial unit it covers. The weighting coefficient is negatively correlated with the Euclidean distance from the center of the spatial unit to the center of the control unit. The generation of the process parameter compensation matrix specifically includes: For each control unit, the direction of the process parameter adjustment command is determined based on the projection value of the planar coordinates of its corresponding spatial unit onto the identified systematic spatial gradient direction. The direction of the adjustment command is strictly consistent with the compensation direction. Calculate the deviation between the height characteristic value of the corresponding spatial unit of each control unit and the process target value specified in the PCB design document, wherein the process target value is the upper limit, lower limit or median of the process range; Arrange all control units in descending order of the absolute value of the deviation; Based on the descending order of the results, the magnitude of the process parameter adjustment instructions is assigned, with the control unit having the smaller sequence number assigned the higher process parameter adjustment level. Arrange the adjustment commands of all control units in spatial coordinate order to form an N-row, M-column matrix structure, and generate a process parameter compensation matrix, where N is the number of control units in the X direction and M is the number of control units in the Y direction.

2. The PCB board boss consistency detection and analysis method as described in claim 1, characterized in that, The calculation of the overall height statistic specifically includes: The difference between the maximum and minimum values ​​of the height dataset is taken as the overall coplanarity error, its standard deviation is taken as the height dispersion, and the proportion of bosses whose height values ​​fall within the preset tolerance zone is taken as the tolerance compliance rate.

3. The PCB board boss consistency detection and analysis method as described in claim 1, characterized in that, The specific method of S2 is as follows: The PCB board surface is divided into multiple analysis areas based on the production process layout, and the average height of each analysis area is obtained. The PCB board surface is divided into columns in the X direction and rows in the Y direction to form a grid-like spatial unit. For each spatial cell, a height characteristic value is calculated based on the average height of the analysis region associated with that cell.

4. The PCB board boss consistency detection and analysis method as described in claim 1, characterized in that, S2 further includes: Before dividing the PCB board surface into grid-like spatial units, noise suppression processing is performed on the data sequence composed of the height data of each boss.

5. The PCB board boss consistency detection and analysis method as described in claim 1, characterized in that, The systematic spatial gradient for identifying height deviations is specifically: Compare the height characteristic value change trend of each column of spatial unit in the X direction of the PCB board, and compare the height characteristic value change trend of each row of spatial unit in the Y direction. If the height feature value shows a continuous increasing or decreasing trend in the X direction, or a continuous increasing or decreasing trend in the Y direction, then a systematic spatial gradient is identified in the height deviation of the corresponding direction.

6. The PCB board boss consistency detection and analysis method as described in claim 1, characterized in that, The specific meaning of identifying regional clustering anomalies is as follows: Spatial units that are directly adjacent in the X or Y direction and have the same height feature value are grouped into the same set. The same height feature value means that the height feature value of the adjacent spatial units is greater than the preset process upper limit or less than the preset process lower limit. Each merged set is treated as a regional clustering anomaly.

7. The PCB board boss consistency detection and analysis method as described in claim 2, characterized in that, The specific method of S4 is as follows: A consistency determination framework is constructed, which includes first-level determination conditions and second-level determination conditions. The first-level judgment criteria include: identifying the presence of a systematic spatial gradient on the plate surface, or the presence of regional clustering anomalies; If any of the first-level judgment conditions are met, the PCB board boss consistency is directly determined to be abnormal. If none of the first-level judgment conditions are met, the second-level judgment is then executed. The second-level judgment criteria include: the tolerance compliance rate meets the process requirements, the overall coplanarity error meets the process requirements, and the high dispersion meets the process requirements; When all the second-level judgment conditions are met, the PCB board boss consistency is deemed qualified. If any of the second-level judgment conditions is not met, the PCB board boss consistency is determined to be abnormal.

8. The PCB board boss consistency detection and analysis method as described in claim 1, characterized in that, The process parameter compensation matrix is ​​converted into equipment instructions, including: According to the equipment control protocol, the adjustment direction and magnitude in the process parameter compensation matrix are encoded into serialized instructions that can be parsed and executed by the equipment controller.

Images