Intelligent control method and system for printed circuit board resin plug hole manufacturing process

Abstract

The application discloses a printed circuit board resin hole filling manufacturing process intelligent control method and system, relates to the printed circuit board manufacturing technical field, and aims to solve the technical problem that the existing printed circuit board resin hole filling equipment is difficult to stably obtain the whole board qualified state by adopting unified global process parameters for filling operation. The method comprises the following steps: acquiring hole characteristic information of a printed circuit board to be processed; detecting conflicts between filling pressure parameter boundaries of different local hole groups along a predetermined scraper running path, generating differentiated control instructions for decoupling the conflicts; controlling a pressure execution unit and an ink supply unit of a hole filling device to perform resin hole filling operation according to the differentiated control instructions; mapping defect positions indicated by quality detection data and actual filling pressure changes recorded with spatial positions recorded in the execution process; and adjusting calculation coefficients corresponding to the filling pressure parameter boundaries based on the mapping results.

Description

Technical Field

[0001] This application relates to the field of printed circuit board manufacturing technology, and in particular to an intelligent control method and system for the resin plugging process of printed circuit boards. Background Technology

[0002] In the production of high-density interconnect boards, especially in scenarios involving the co-filling of multi-level blind vias and through-holes, a single printed circuit board often contains both micro-blind vias and large mechanical through-holes. Due to significant differences in their physical properties, these two types of vias respond very differently to resin filling. Specifically, micro-blind vias, due to their tiny size and significant capillary resistance, require higher filling pressures to ensure complete resin filling to the bottom of the via, thus avoiding underfill defects. However, for larger mechanical through-holes, under the same high pressure conditions, resin is prone to excessive overflow, and may even splash, leading to quality problems such as overfill protrusions. Summary of the Invention

[0003] This application provides an intelligent control method and system for the resin via plugging manufacturing process of printed circuit boards. It aims to solve the technical problem that existing resin via plugging equipment for printed circuit boards uses uniform global process parameters for filling operations, which cannot automatically identify and adapt to the different filling parameter requirements of different physical feature hole groups on the board surface. This results in the inevitable simultaneous occurrence of underfilling and overfilling quality defects on the same printed circuit board, making it difficult for the experience-based compromise control method to stably obtain the overall board qualified state, seriously affecting the first-pass yield, and frequently causing downtime for debugging, thus reducing production efficiency.

[0004] In a first aspect, to solve the above-mentioned technical problems, the present invention provides an intelligent control method for the resin via plugging manufacturing process of printed circuit boards. The method includes: acquiring via feature information of the printed circuit board to be processed, and dividing the board surface into multiple local via groups based on the via feature information. The via feature information includes the type, geometric parameters, and location of each via. Each local via group is formed by the aggregation of vias whose filling requirement similarity meets preset conditions. For each local via group, based on the via type and geometric parameters of the vias it contains, and preset resin rheological property data, determining the filling pressure parameter boundary of the local via group. The filling pressure parameter boundary includes a minimum filling pressure lower limit required for filling micro-blind vias and a maximum filling pressure upper limit allowed to prevent overfilling of mechanical through-holes. Along a predetermined scraper travel path, detecting conflicts between the filling pressure parameter boundaries of different local via groups, and determining the conflict when the minimum filling pressure lower limit of a first local via group is greater than the maximum filling pressure upper limit of a second local via group. When the upper limit of the force is reached, differentiated control instructions for decoupling conflicts are generated. The first local orifice group is a local orifice group whose minimum filling pressure lower limit is higher than that of the adjacent orifice group, and the second local orifice group is a local orifice group whose maximum filling pressure upper limit is lower than that of the adjacent orifice group. The differentiated control instructions include at least applying different filling pressures in different local orifice group regions, or splitting a complete filling stroke into multiple strokes with different pressure and speed combinations. The pressure execution unit and ink supply unit of the orifice plugging device are controlled to perform resin orifice plugging operation according to the differentiated control instructions. During the operation, when the actual filling pressure is detected to be lower than the minimum filling pressure lower limit corresponding to the current position, compensation adjustment of ink supply rate and downward stroke is triggered. The quality inspection data after orifice plugging is acquired, and the defect location indicated by the quality inspection data is mapped with the record of the change of actual filling pressure with spatial position during the execution process. The calculation coefficient corresponding to the boundary of filling pressure parameter is adjusted based on the mapping result.

[0005] Secondly, this application provides an intelligent control system for the resin via plugging manufacturing process of printed circuit boards. The system includes: an acquisition unit, used to acquire hole feature information of the printed circuit board to be processed, and divide the board surface into multiple local hole groups based on the hole feature information. The hole feature information includes the type, geometric parameters, and location of each hole. Each local hole group is formed by aggregating holes whose filling requirement similarity meets preset conditions; a determination unit, used to determine the filling pressure parameter boundary of each local hole group based on the hole type and geometric parameters of the holes it contains, as well as preset resin rheological property data. The filling pressure parameter boundary includes a minimum filling pressure lower limit required for filling micro-blind vias and a maximum filling pressure upper limit allowed to prevent overfilling of mechanical through-holes; and a generation unit, used to detect conflicts between the filling pressure parameter boundaries of different local hole groups along a predetermined scraper travel path, and to determine the conflict when the minimum filling pressure lower limit of a first local hole group is greater than the maximum filling pressure of a second local hole group. Within a time limit, differentiated control instructions for decoupling conflicts are generated. The first local orifice group is a local orifice group whose minimum filling pressure is higher than that of adjacent orifice groups, and the second local orifice group is a local orifice group whose maximum filling pressure is lower than that of adjacent orifice groups. The differentiated control instructions include at least applying different filling pressures to different local orifice group regions, or splitting a complete filling stroke into multiple strokes with different pressure and speed combinations. A control unit is used to control the pressure execution unit and ink supply unit of the orifice plugging device to perform resin plugging operations according to the differentiated control instructions. During operation, when the actual filling pressure is detected to be lower than the minimum filling pressure corresponding to the current position, compensation adjustments to the ink supply rate and downward stroke are triggered. An adjustment unit is used to acquire quality inspection data after orifice plugging, map the defect location indicated by the quality inspection data to the record of actual filling pressure changes with spatial position during execution, and adjust the calculation coefficients corresponding to the filling pressure parameter boundaries based on the mapping results.

[0006] This application has at least the following beneficial effects: The intelligent control method for resin via plugging manufacturing process of printed circuit boards disclosed in this application acquires the hole feature information of the printed circuit board to be processed, and divides the board surface into multiple local hole groups based on the hole feature information, thereby achieving refined identification and classification of the hole features of the board. For each local hole group, based on its hole type, geometric parameters, and preset resin rheological property data, the minimum filling pressure lower limit and maximum filling pressure upper limit of the local hole group are determined, providing a scientific pressure boundary basis for subsequent differentiated control. On the doctor blade travel path, the system can detect conflicts between the filling pressure parameter boundaries of different local hole groups, and generate differentiated control instructions for decoupling conflicts when conflicts exist. These instructions may include applying different filling pressures in different local hole group areas, or splitting a complete filling stroke into multiple strokes with different pressure and speed combinations, thereby effectively solving the problems of underfilling and overfilling caused by hole shape differences in traditional via plugging processes. During the via plugging operation, the system can detect the actual filling pressure in real time, and trigger compensation adjustments to the ink supply rate and downward stroke when the pressure is lower than the minimum filling pressure lower limit, ensuring the stability and quality of the filling process. Furthermore, by acquiring quality inspection data after plugging the holes and mapping the defect locations to the actual filling pressure records, the calculation coefficients corresponding to the filling pressure parameter boundaries were adjusted based on the mapping results, thus achieving adaptive optimization of the plugging process. Attached Figure Description

[0007] Figure 1 This is a flowchart illustrating an intelligent control method for resin via plugging manufacturing of printed circuit boards provided in this application. Figure 2 This is a flowchart illustrating another intelligent control method for resin via plugging manufacturing process of printed circuit boards provided in this application. Figure 3 This is a schematic diagram of the structure of an intelligent control system for the resin plugging process of a printed circuit board, as provided in this application. Detailed Implementation

[0008] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0009] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0010] In the production of high-density interconnect boards, especially in scenarios involving the co-filling of multi-level blind vias and through-holes, a single printed circuit board often contains both micro-blind vias and large mechanical through-holes. Due to significant differences in their physical properties, these two types of vias respond very differently to resin filling. Specifically, micro-blind vias, due to their tiny size and significant capillary resistance, require higher filling pressures to ensure complete resin filling to the bottom of the via, thus avoiding underfill defects. However, for larger mechanical through-holes, under the same high pressure conditions, resin is prone to excessive overflow, and may even splash, leading to quality problems such as overfill protrusions.

[0011] Existing resin via-filling equipment for printed circuit boards typically uses uniform global process parameters for filling operations. This control method, relying on a single global command, cannot automatically identify and adapt to the differentiated filling parameter requirements of different physical feature hole groups on the board surface. When the equipment controller sets a high filling pressure to meet the filling requirements of micro-blind vias, the mechanical through-hole area will be overfilled due to excessive pressure; conversely, if a lower filling pressure is used to avoid overfilling of mechanical through-holes, the micro-blind via area will be underfilled due to insufficient pressure. This mismatch between the fixed global parameter settings and the differentiated requirements of local hole filling states inevitably leads to the simultaneous occurrence of underfill and overfilling defects on the same printed circuit board. This makes it difficult for the experience-based compromise control method to consistently achieve a qualified state for the entire board, seriously affecting the first-pass yield, frequently causing downtime for debugging, and reducing production efficiency.

[0012] In view of the above problems, this application provides an intelligent control method for the resin via plugging manufacturing process of printed circuit boards. This method acquires the hole feature information of the printed circuit board to be processed and divides the board surface into multiple local hole groups based on the hole feature information. For each local hole group, a filling pressure parameter boundary is determined. This method can detect conflicts between the filling pressure parameter boundaries of different local hole groups along a predetermined squeegee travel path and generate differentiated control commands to decouple the conflicts when they exist. By controlling the pressure execution unit and ink supply unit of the via plugging equipment to perform resin via plugging operations according to the differentiated control commands, and by performing compensation adjustments during the operation, and by acquiring post-plugging quality inspection data and adjusting the calculation coefficients corresponding to the filling pressure parameter boundaries based on the mapping results, this method effectively solves the problem of underfilling and overfilling caused by hole shape differences in traditional via plugging processes, significantly improving via plugging quality and production efficiency.

[0013] The intelligent control method for resin plugging manufacturing process of printed circuit boards provided in this application will be described in detail and explained through the following specific embodiments.

[0014] Reference Figure 1 This application provides an intelligent control method for the resin via plugging manufacturing process of printed circuit boards, which may include the following steps: S101. Obtain the hole feature information of the printed circuit board to be processed, and divide the board surface into multiple local hole groups according to the hole feature information.

[0015] The hole feature information includes the type, geometric parameters and location of each hole. Each local hole group is formed by aggregating holes whose filling requirement similarity meets the preset conditions.

[0016] S102. For each local pore group, determine the filling pressure parameter boundary of the local pore group based on the pore type and geometric parameters of the pores it contains and the preset resin rheological property data.

[0017] The filling pressure parameter boundaries include the minimum filling pressure lower limit required to fill micro blind holes and the maximum filling pressure upper limit allowed to prevent overfilling of mechanical through holes.

[0018] S103. Along the predetermined scraper travel path, detect the conflict between the boundaries of the filling pressure parameters of different local hole groups. When the minimum filling pressure lower limit of the first local hole group is greater than the maximum filling pressure upper limit of the second local hole group, generate a differentiated control command for decoupling the conflict.

[0019] Among them, the first local hole group is a local hole group whose minimum filling pressure lower limit is higher than that of the adjacent hole group, the second local hole group is a local hole group whose maximum filling pressure upper limit is lower than that of the adjacent hole group, and the differentiated control command includes at least applying different filling pressures in different local hole group regions, or splitting a complete filling stroke into multiple strokes with different pressure and speed combinations. S104. The pressure execution unit and ink supply unit of the control plugging device perform resin plugging operation according to differentiated control instructions. During the operation, when the actual filling pressure is detected to be lower than the minimum filling pressure limit corresponding to the current position, the compensation adjustment of ink supply rate and downward stroke is triggered.

[0020] S105. Obtain the quality inspection data after plugging the hole, map the defect location indicated by the quality inspection data with the actual filling pressure change record recorded during the execution process, and adjust the calculation coefficient corresponding to the boundary of the filling pressure parameter based on the mapping result.

[0021] The intelligent control method for resin via filling manufacturing process of printed circuit boards proposed in this application aims to optimize the resin via filling process of printed circuit boards, especially when processing boards with various via types (such as micro blind vias and mechanical through-holes). The core of this method lies in achieving differentiated and refined control over different areas to ensure the quality of resin filling.

[0022] "Hole feature information" refers to data describing the physical properties of holes on a printed circuit board, including the type of each hole (e.g., micro blind vias, mechanical through-holes), geometric parameters (e.g., hole diameter, hole depth), and spatial location (e.g., center coordinates). This information forms the basis for subsequent board layout and parameter settings.

[0023] "Local via clusters" refer to regions on a printed circuit board surface divided according to via feature information. The vias within each local via cluster have similar filling requirements. This division helps to break down complex global problems into multiple manageable local problems, thereby enabling fine-grained control.

[0024] "Resin rheological property data" refers to data describing the physical properties of resins, such as viscosity and surface tension, under different shear rates, temperatures, and other conditions.

[0025] "Fill pressure parameter boundary" refers to the resin filling pressure range set for each local well group, including the minimum filling pressure lower limit and the maximum filling pressure upper limit. The minimum filling pressure lower limit is to ensure that micro-blind wells can be completely filled and to avoid underfilling; the maximum filling pressure upper limit is to prevent overfilling of mechanically perforated wells.

[0026] "Differentiated control instructions" refer to instructions generated based on the boundary conflicts of filling pressure parameters in different local orifice groups, used to guide the orifice plugging equipment to perform refined operations. These instructions may include applying different filling pressures in different areas, or breaking a complete filling stroke into multiple strokes with different combinations of pressure and speed.

[0027] The pressure actuator and ink supply unit are key components of the plugging device. The pressure actuator controls the pressure applied to the resin by the doctor blade, while the ink supply unit controls the amount and rate of resin supply.

[0028] "Quality inspection data" refers to data about the quality of the plugging operation obtained through optical inspection or other inspection methods after the plugging operation is completed, such as underfill depth, overfill height, and defect location. This data is used to evaluate the plugging effect and provide feedback for adjusting control parameters.

[0029] First, the method involves acquiring the hole feature information of the printed circuit board to be processed, and dividing the board surface into multiple local hole groups based on the hole feature information. The hole feature information can be obtained directly from the design file, such as by reading data like hole position, diameter, and depth from a CAD file. Alternatively, image recognition technology can be used to scan the printed circuit board, identify all holes on the board surface, and measure their geometric parameters. After acquiring the hole feature information, a density-based clustering algorithm, such as the DBSCAN algorithm, can be used to aggregate the holes on the board surface into multiple local hole groups based on the spatial location and geometric parameter similarity of the holes. For example, holes with similar diameters, similar depths, and spatial proximity can be grouped into one local hole group.

[0030] Secondly, for each local pore group, the filling pressure parameter boundary is determined based on the pore type and geometric parameters of its constituent pores, as well as preset resin rheological property data. For example, for a local pore group mainly consisting of micro-blind pores, the minimum filling pressure lower limit required to ensure complete resin filling can be calculated using capillary filling theory, based on parameters such as the pore diameter and depth of the micro-blind pores, as well as the surface tension and viscosity of the resin. For a local pore group mainly consisting of mechanical through-holes, the maximum filling pressure upper limit to avoid overfilling can be calculated using a fluid dynamics model, based on parameters such as the pore diameter and depth of the mechanical through-holes, as well as the viscosity of the resin and the scraper speed. These calculations can be based on preset mathematical models and empirical formulas.

[0031] Next, along the predetermined scraper travel path, conflicts between the filling pressure parameter boundaries of different local orifice groups are detected. For example, the scraper travel path can be divided into a series of continuous micro-regions, and within each micro-region, all local orifice groups currently covered by the scraper are identified. Then, the lower limit of the minimum filling pressure and the upper limit of the maximum filling pressure of these local orifice groups are compared. If the lower limit of the minimum filling pressure of a certain local orifice group is found to be higher than the upper limit of the maximum filling pressure of its adjacent local orifice groups, a pressure conflict is determined to exist. For example, such a conflict may occur when the scraper moves from a micro-blind orifice group area requiring high-pressure filling to a pressure-sensitive mechanical through-hole group area.

[0032] When the minimum filling pressure of the first local aperture group exceeds the maximum filling pressure of the second local aperture group, differentiated control commands are generated to decouple the conflict. For example, when a conflict is detected, a command can be generated to apply a higher filling pressure to the first local aperture group region and a lower filling pressure to the second local aperture group region. This can be achieved by dynamically adjusting the downward pressure of the scraper or the gap between the scraper and the board surface during the scraper's travel. Alternatively, a complete filling stroke can be broken down into multiple strokes. For example, the first stroke can use a lower pressure and a faster speed to prioritize filling mechanical through-holes; the second stroke can use a higher pressure and a slower speed to ensure adequate filling of micro-blind vias.

[0033] Next, the pressure actuator and ink supply unit of the orifice-filling device perform resin orifice-filling operations according to differentiated control commands. For example, the pressure actuator can be a servo motor-driven doctor blade mechanism, which dynamically adjusts the downward pressure of the doctor blade by precisely controlling the speed and position of the servo motor. The ink supply unit can be a peristaltic pump or a gear pump, which regulates the ink supply rate of the resin by controlling the pump speed. During operation, the actual filling pressure can be monitored in real time. For example, by installing a pressure sensor below the doctor blade or near the board surface, the resin pressure below the doctor blade can be acquired in real time. When the actual filling pressure is detected to be lower than the minimum filling pressure limit corresponding to the current position, compensation adjustments to the ink supply rate and downward stroke can be triggered. For example, the ink supply can be increased immediately, and the downward stroke of the doctor blade can be slightly increased to quickly increase the local pressure and ensure the filling of micro-blind holes.

[0034] Finally, the quality inspection data after plugging is obtained. The defect locations indicated by the quality inspection data are mapped to the records of actual filling pressure changes with spatial position during the process. The calculation coefficients corresponding to the filling pressure parameter boundaries are adjusted based on the mapping results. For example, the plate surface image after plugging is obtained through optical inspection equipment, and the location and type of defects such as underfilling and overfilling are identified. Then, these defect locations are spatially aligned with the actual filling pressure curves recorded during the plugging process. If a local group of holes is found to be underfilled even when the actual filling pressure meets its minimum filling pressure limit, the calculation model for the minimum filling pressure limit of that local group of holes is considered to have a deviation, and its resistance correction coefficient needs to be increased. If a local group of holes is found to be overfilled even when the actual filling pressure is lower than its maximum filling pressure limit, the calculation model for the maximum filling pressure limit of that local group of holes is considered to have a deviation, and its critical flow velocity needs to be reduced. Through this feedback mechanism, the calculation model of the filling pressure parameter boundaries can be continuously optimized, improving the accuracy and stability of plugging.

[0035] The intelligent control method for resin via plugging manufacturing process of printed circuit boards disclosed in this application achieves intelligent control of the resin via plugging process by finely dividing the surface of the printed circuit board and dynamically setting and adjusting the filling pressure parameters for different local hole groups.

[0036] Specifically, this method first acquires the hole feature information of the printed circuit board to be processed, which forms the basis for all subsequent decisions. By comprehensively understanding the hole type, geometric parameters, and location, the system can intelligently divide the board surface into multiple local hole groups. This division is not a simple region segmentation, but rather based on the similarity of filling requirements, ensuring that the holes within each local hole group respond consistently to resin filling, thus laying the foundation for subsequent differentiated control.

[0037] Next, for each local pore group, the system accurately calculates the filling pressure parameter boundary based on the type and geometric parameters of its internal pores, as well as preset resin rheological property data. This boundary includes the minimum filling pressure lower limit required to ensure complete filling of micro-blind holes, and the maximum filling pressure upper limit allowed to prevent overfilling of mechanically perforated holes. This dual-boundary setting fully considers the filling characteristics of different pore types and avoids the drawbacks of the "one-size-fits-all" approach in traditional methods.

[0038] During the scraper's movement, the system continuously monitors the boundaries of filling pressure parameters for different local hole groups along a predetermined path to detect any conflicts. When the minimum filling pressure of the first local hole group is detected to be higher than the maximum filling pressure of the second local hole group, it indicates that there are hole groups with conflicting filling pressure requirements within the same scraper's operating area. In this case, the system intelligently generates differentiated control commands to decouple the conflicts. These commands can include applying different filling pressures to different local hole group areas, or breaking down a complete filling stroke into multiple strokes with different pressure and speed combinations. This dynamic conflict decoupling mechanism is one of the core innovations of this application. It enables the system to flexibly handle complex hole distributions on the board surface, avoiding the problem of both underfilling and overfilling in traditional methods.

[0039] Subsequently, the pressure actuation unit and ink supply unit of the orifice plugging device strictly follow the differentiated control instructions to perform resin orifice plugging operations. During operation, the system continuously monitors the actual filling pressure. Once the actual filling pressure is detected to be lower than the minimum filling pressure limit corresponding to the current position, the system immediately triggers compensation adjustments to the ink supply rate and downward stroke. This real-time feedback and compensation mechanism ensures that even in dynamically changing filling environments, micro-blind orifices can obtain sufficient filling pressure, thereby effectively avoiding underfill defects.

[0040] Finally, to achieve continuous optimization, the system acquires quality inspection data after plugging and maps the defect locations indicated by this data to records of actual filling pressure changes with spatial position during execution. Based on the mapping results, the system intelligently adjusts the calculation coefficients corresponding to the boundaries of the filling pressure parameters. This closed-loop feedback adjustment mechanism allows the system to learn and optimize from actual production, continuously improving the accuracy of filling pressure parameter settings, thereby continuously improving plugging quality and production efficiency.

[0041] In some embodiments described above in this application, reference is made to Figure 2 To achieve intelligent control of the resin-filled via manufacturing process for printed circuit boards (PCBs), it is necessary to acquire the via feature information of the PCB to be processed and divide the board surface into multiple local via groups based on this feature information. Specifically, the steps of acquiring the via feature information of the PCB to be processed and dividing the board surface into multiple local via groups based on this feature information can be further refined into the following operations: S201. Analyze the initial type markings, center coordinates for spatial positioning, hole diameter, and hole depth of each hole to be plugged from the manufacturing design documents of the printed circuit board.

[0042] S202. Based on the preset aperture threshold, the depth characteristics represented by the aperture depth, and the initial type label, classify each aperture into a micro blind aperture or a mechanical through aperture, and determine the overfill sensitivity for each aperture.

[0043] The overfill sensitivity is determined by the ratio of the square of the pore diameter to the square root of the pore depth, and is used to characterize the risk of overfilling of the pore under a given filling pressure. S203. Based on the center coordinates of each hole, assign each hole to the grid that divides the plate surface.

[0044] The side length of the grid is determined by the effective working width of the scraper and the lateral diffusion distance of the resin, and the number of micro-blind holes and mechanical through holes, the average hole diameter and the total volume in each grid are counted.

[0045] S204. Using a preset spatial clustering algorithm, adjacent grids that meet the preset similarity conditions are aggregated into local hole groups based on the spatial adjacency between grids and the distance metric constructed by the difference in average aperture of micro blind holes, the difference in average aperture of mechanical through holes and the weighted average of overfill sensitivity within the grid, and the state vector of each local hole group is output.

[0046] The state vector includes the spatial boundary coordinates determined by the range of aggregated grid coordinates, the dominant pore size represented by the average pore diameter and average pore depth of the same type of pores in the pore group, and the total resin requirement determined by the sum of the theoretical volumes of each pore in the pore group.

[0047] Specifically, this step aims to extract the basic geometric and positional information of each via to be plugged from the manufacturing design documents of the printed circuit board, such as Gerber files or Excellon drill files. This information includes the initial type label of the hole (e.g., whether it is a blind via, through-hole, etc.), the center coordinates for precise spatial positioning, the diameter of the hole, and the depth of the hole. By parsing this raw data, a data foundation is laid for subsequent hole classification and board layout.

[0048] After obtaining the basic information about the pores, each pore needs to be finely classified according to a preset pore diameter threshold, the depth characteristics represented by the pore depth, and the initial type label, for example, distinguishing them as micro-blind pores or mechanically through-holes. Simultaneously, the overfill sensitivity of each pore is determined. Overfill sensitivity is a key indicator measuring the risk of resin overfilling at a specific filling pressure, specifically quantified by the ratio of the square of the pore diameter to the square root of the pore depth. This quantification method can more accurately reflect the differences in the response of different pore types to filling pressure, providing a basis for subsequent determination of pressure parameter boundaries.

[0049] In practical applications, for ease of management and analysis, the surface of the printed circuit board is divided into a series of grids based on the center coordinates of each hole, and each hole is assigned to a corresponding grid. The side length of the grid is not arbitrarily set, but is determined by the effective working width of the squeegee and the lateral diffusion distance of the resin on the board surface, ensuring that the grid division matches the physical characteristics of the actual hole-filling process. Within each grid, the number of micro-blind vias and mechanical through-holes, their average diameter, and total volume are counted. These statistics provide basic information for the aggregation of local hole groups.

[0050] Furthermore, this application utilizes a pre-defined spatial clustering algorithm to aggregate adjacent grids that meet pre-defined similarity conditions into local pore groups. During the clustering process, not only is the spatial adjacency between grids considered, but a distance metric constructed from the difference in average diameter of micro-blind holes, the difference in average diameter of mechanical through holes, and the weighted average of overfill sensitivity within the grid is also used as the basis for similarity judgment. This multi-dimensional distance metric ensures that the aggregated local pore groups are spatially adjacent and that the filling requirements of their internal holes are highly similar. After aggregation, the system outputs a state vector for each local pore group. This state vector details the characteristics of the pore group, including the spatial boundary coordinates determined by the aggregated grid coordinate range, the dominant pore size represented by the average diameter and average depth of similar holes within the pore group, and the overall resin requirement determined by the sum of the theoretical volumes of all holes within the pore group. These state vectors are key inputs for subsequently determining the filling pressure parameter boundaries and generating differentiated control commands.

[0051] The above technical solution enables refined acquisition and processing of via feature information on printed circuit boards (PCBs), overcoming the problems of coarse processing and inability to effectively distinguish the filling requirements of different via types in traditional methods. By introducing overfill sensitivity and performing multi-dimensional clustering, the identified local via groups exhibit a high degree of similarity in filling requirements, thus providing more accurate and reliable basic data for subsequent intelligent control. This refined preprocessing significantly improves the intelligence level and control precision of the via plugging process, helps reduce defects such as underfill and overfill, and ultimately improves the via plugging quality and production efficiency of PCBs.

[0052] In some embodiments of this application, for each local pore group, it is necessary to determine the filling pressure parameter boundary of the local pore group based on the pore type and geometric parameters of the pores it contains, as well as preset resin rheological property data. Specifically, determining the filling pressure parameter boundary of the local pore group includes: For a local group of holes containing micro-blind holes, the shear rate experienced by the resin when entering the micro-blind holes is estimated based on the scraper travel speed. The apparent dynamic viscosity of the resin at this shear rate is obtained from the preset resin rheological property data. The minimum filling pressure is determined by combining the resin surface tension coefficient, micro-blind hole diameter, micro-blind hole depth, reverse air pressure generated by the adiabatic compression of air at the bottom of the hole during the filling process, and the preset resistance correction coefficient. For a localized pore group containing mechanical through-holes, the upper limit of the maximum filling pressure is determined based on the apparent dynamic viscosity of the resin under the corresponding shear conditions, the preset critical flow rate, and the pore diameter and depth of the mechanical through-holes. The minimum filling pressure lower limit and the maximum filling pressure upper limit are used together as the filling pressure parameter boundary for this local pore group.

[0053] Specifically, when determining the minimum filling pressure limit, the scraper travel speed is a key factor affecting the resin shear rate, which directly determines the resin flow behavior within the micro-blind holes. Using pre-set resin rheological property data, the apparent dynamic viscosity of the resin at a specific shear rate can be accurately obtained, reflecting the internal frictional force of the resin during flow. Simultaneously, the resin surface tension coefficient, the pore size and depth of the micro-blind holes directly affect the capillary action and flow resistance within the micro-blind holes. Air at the bottom of the holes undergoes adiabatic compression during filling, generating a reverse pressure that hinders resin filling and therefore needs to be included in the calculation. A pre-set resistance correction coefficient is used to correct the deviation between the model and actual conditions, ensuring the accuracy of the calculation results. By comprehensively considering these factors, the minimum filling pressure limit required to ensure complete filling of the micro-blind holes can be accurately calculated.

[0054] Among the parameters considered when determining the maximum filling pressure limit, the apparent dynamic viscosity of the resin under corresponding shear conditions is also crucial, as it affects the flow resistance of the resin within the mechanically perforated orifice. The preset critical flow rate refers to the maximum allowable flow rate of the resin within the mechanically perforated orifice without overfilling. The orifice diameter and depth determine the orifice geometry, directly influencing the resin's filling volume and flow path. By comprehensively considering these parameters, the maximum filling pressure limit to avoid overfilling in the mechanically perforated orifice can be calculated.

[0055] The above technical solution enables precise control of the filling pressure for different types of holes on printed circuit boards, significantly improving the quality and efficiency of via plugging. Specifically, by comprehensively considering factors such as squeegee travel speed, resin rheological properties, hole geometry parameters, and bottom air pressure, the minimum filling pressure limit for micro-blind holes is accurately calculated, effectively solving the problem of underfilling in micro-blind holes in traditional methods. Simultaneously, by introducing a critical flow rate and combining it with resin viscosity, hole diameter, and hole depth, the maximum filling pressure limit for mechanical through-holes is precisely determined, effectively avoiding overfilling. This refined method of determining pressure parameter boundaries allows the via plugging process to better adapt to complex and varied printed circuit board designs, thereby reducing defect rates, improving product yield, and lowering rework costs.

[0056] In response, this application further proposes a method for generating differentiated control instructions for decoupling conflicts, including: The plate surface is divided into continuous strip regions along the scraper's travel path, and the minimum filling pressure lower limit and maximum filling pressure upper limit of all local hole groups in each strip region are extracted. For each strip region, determine whether there is an intersection between the boundaries of the filling pressure parameters of each local pore group; when the lower limit of the minimum filling pressure of any local pore group is greater than the upper limit of the maximum filling pressure of another local pore group, the intersection of their pressure intervals is empty, indicating a parameter conflict, and the local pore group with the higher lower limit of the minimum filling pressure is identified as the first local pore group, and the local pore group with the lower upper limit of the maximum filling pressure is identified as the second local pore group; when the pressure intervals of all local pore groups have a common intersection, the median of the intersection is taken as the target filling pressure for the strip region; For strip regions with parameter conflicts, a decoupling strategy is selected based on the pore density distribution of the first local pore group and the second local pore group in the strip region: if the mechanical through-hole density is lower than the first preset threshold, a high-pressure priority strategy is adopted, and the local fluid pressure distribution is changed by adjusting the scraper angle of attack before the scraper reaches the strip region; if the micro-blind hole density and the mechanical through-hole density are not lower than the corresponding preset threshold, control commands for at least two differentiated strokes are generated, wherein the pressure of the first stroke is set to be lower than the maximum filling pressure limit and the scraper speed is higher than the preset normal speed, so that the resin flows into the mechanical through-hole first; the pressure of the second stroke is set to be higher than the minimum filling pressure limit and the scraper speed is lower than the preset normal speed, and the filling of the micro-blind holes is completed by utilizing the liquid resistance formed by the resin already filled in the mechanical through-hole. The filling pressure setpoints for each strip area or the pressure and speed combinations for each stroke are arranged along the scraper's travel position to form a continuous pressure control curve, which serves as a differentiated control command.

[0057] Specifically, dividing the board surface into continuous strip regions along the squeegee's travel path refers to logically dividing the entire printed circuit board surface into a series of adjacent regions of a certain width along the direction of squeegee movement. The width of each strip region can be determined based on the effective working width of the squeegee, the lateral diffusion distance of the resin, and the distribution characteristics of the pore groups, ensuring that the filling pressure can be effectively assessed and managed within each strip region. Extracting the minimum filling pressure lower limit and maximum filling pressure upper limit for all local pore groups within each strip region is to obtain the basic pressure requirements for all pore groups that need to be filled within that region, providing a data foundation for subsequent collision detection and strategy selection.

[0058] The determination of whether the boundaries of the filling pressure parameters of each local pore group overlap aims to identify whether the filling pressure requirements of different local pore groups within the same strip region can be simultaneously met. When the lower limit of the minimum filling pressure of any local pore group is greater than the upper limit of the maximum filling pressure of another local pore group, it means that a single filling pressure value cannot be found that can simultaneously meet the requirements of all pore groups, i.e., a parameter conflict exists. In this case, the local pore group with the higher lower limit of the minimum filling pressure is identified as the first local pore group, whose demand for high pressure is the dominant factor; the local pore group with the lower upper limit of the maximum filling pressure is identified as the second local pore group, whose sensitivity to low pressure is the limiting factor. When the pressure ranges of all local pore groups have a common intersection, it indicates that a single filling pressure value can be found that meets the requirements of all pore groups. In this case, the median of this intersection is used as the target filling pressure for the strip region to take into account the filling effect of each pore group.

[0059] In practical applications, for strip regions with conflicting parameters, a decoupling strategy is selected based on the pore density distribution of the first and second local pore groups within that strip region. Pore density distribution, such as mechanical through-hole density and micro-blind through-hole density, is a key indicator for assessing the dominance of pore type in that region. If the mechanical through-hole density is below a first preset threshold, it indicates that the region is mainly composed of micro-blind through-holes or has a small number of mechanical through-holes; in this case, a high-pressure priority strategy is more appropriate. The high-pressure priority strategy aims to ensure sufficient filling of micro-blind through-holes. By adjusting the doctor blade's angle of attack before it reaches the strip region, the local fluid pressure distribution can be altered, allowing for a higher effective filling pressure to be applied in the micro-blind through-hole region. The doctor blade's angle of attack refers to the angle between the doctor blade and the plate surface; adjusting this angle changes the pressure exerted by the doctor blade on the resin, thus affecting the local filling pressure.

[0060] Furthermore, if both the density of micro-blind holes and the density of mechanically perforated holes are not lower than the corresponding preset thresholds, it indicates that a large number of micro-blind holes and mechanically perforated holes exist in this strip region, and there is a significant conflict in the requirements for filling pressure. In this case, generating control commands for at least two differentiated strokes is an effective decoupling strategy. The pressure of the first stroke is set below the maximum filling pressure limit and the scraper speed is higher than the preset normal speed. The purpose is to allow the resin to flow preferentially into the mechanically perforated holes, avoiding overfilling in the mechanically perforated holes in the early stage. The higher scraper speed helps the resin to pass through the mechanically perforated holes quickly, reducing the residence time at the mechanically perforated hole opening, thereby reducing the risk of overfilling. The pressure of the second stroke is set above the minimum filling pressure limit and the scraper speed is lower than the preset normal speed. The purpose is to use the liquid resistance formed by the resin already filled in the mechanically perforated holes to provide additional filling pressure for the micro-blind holes, thereby completing the full filling of the micro-blind holes. The lower scraper speed allows the resin sufficient time to penetrate and fill in the micro-blind holes.

[0061] Therefore, the filling pressure setpoints for each strip area or the pressure and speed combinations for each stroke are arranged along the scraper's travel position to form a continuous pressure control curve, which serves as a differentiated control command. This continuous pressure control curve is a dynamically changing sequence of pressure commands. Based on the real-time position of the scraper on the plate surface, it indicates the filling pressure and scraper speed that the plugging equipment should apply in different areas or different strokes, thereby achieving refined and intelligent control of the entire plugging process.

[0062] Through the above technical solutions, this application can significantly improve the quality and efficiency of resin via plugging on printed circuit boards. Compared with methods that only generate general differentiated control instructions, this application achieves a higher level of intelligent control over the via plugging process by introducing strip region division, refined conflict detection, and an adaptive decoupling strategy based on via density distribution. Specifically, the high-pressure priority strategy effectively solves the underfilling problem of micro-blind vias in areas with low density of mechanical through-holes, while the differentiated stroke strategy cleverly balances the filling requirements of the two types of vias in areas with high mixed density of micro-blind vias and mechanical through-holes, avoiding the contradiction of overfilling mechanical through-holes and underfilling micro-blind vias. This refined control capability enables the via plugging equipment to dynamically adjust the filling parameters according to the complexity and diversity of via distribution on the printed circuit board, thereby significantly reducing the via plugging defect rate, improving product yield, and reducing rework costs. In addition, the resulting continuous pressure control curve provides precise guidance for the automated and intelligent operation of the via plugging equipment, further improving production efficiency and stability.

[0063] The pressure actuation unit and ink supply unit of the aforementioned control plugging device perform resin plugging operations according to the aforementioned differentiated control commands, specifically including: Based on the inverse model of the device transfer function, which includes aerodynamic delay, mechanical stiffness, and fluid damping, the continuous pressure control curve is converted into a voltage time series that drives the pressure actuator. The system receives real-time position feedback from the scraper position sensor and adjusts the issuing clock frequency of the voltage time sequence through a phase-locked loop control, so that the change in pressure command is spatially synchronized with the boundary of the local hole group. During the movement of the doctor blade, the actual filling pressure is continuously compared with the continuous pressure control curve. When the actual filling pressure is lower than the minimum filling pressure limit corresponding to the current position, an accelerated ink supply command is sent to the ink supply unit and the doctor blade pressure is increased until the actual filling pressure returns to the preset target range.

[0064] Specifically, the inverse transfer function model of the device refers to a mathematical model that describes the dynamic response characteristics of the pressure actuator in the plugging device from receiving an electrical signal to actually generating mechanical pressure output. This model considers factors such as the delay of the pneumatic system, the stiffness of mechanical components, and the damping effect of the resin fluid in front of the scraper. By constructing its inverse model, the desired continuous pressure control curve (i.e., the change of target pressure with time or position) can be calculated in reverse as the voltage-time series required to drive the pressure actuator. Its purpose is to compensate for the nonlinear response and dynamic characteristics of the device itself, ensuring that the actually applied pressure accurately tracks the preset control curve.

[0065] The squeegee position sensor can be understood as a device used to monitor the real-time position of the squeegee on the printed circuit board surface; it could be an optical encoder, magnetic scale, or laser rangefinder. Its real-time position feedback is crucial for precise control. A phase-locked loop (PLL) control is a feedback control system used to synchronize the phase of one signal with the phase of another reference signal. In this application, the PLL is used to adjust the firing clock frequency of the voltage time series to ensure that the points of change in pressure command (e.g., the change in pressure command when transitioning from one local aperture group to another) are spatially precisely aligned with the actual boundaries of the local aperture groups on the printed circuit board. Its purpose is to eliminate spatial asynchrony problems caused by squeegee speed fluctuations or mechanical transmission errors, ensuring that the correct pressure is applied at the correct physical location.

[0066] In practical applications, during the doctor blade movement, the system continuously acquires actual filling pressure data fed back by pressure sensors (e.g., mounted on the doctor blade or its support structure) and compares it in real time with a pre-generated continuous pressure control curve. When the actual filling pressure is detected to be lower than the minimum filling pressure limit corresponding to the current doctor blade position, it indicates that there may be insufficient resin supply or excessive filling resistance in the current area, potentially leading to underfilling. At this time, the system immediately triggers compensation adjustments, sending an accelerated ink supply command to the ink supply unit to increase the amount of resin in front of the doctor blade, while simultaneously increasing the downward pressure of the doctor blade to improve the local filling pressure. These compensation actions continue until the actual filling pressure returns to the preset target range, thereby effectively preventing underfilling defects.

[0067] Through the above technical solutions, this application significantly improves the control accuracy and robustness of the resin via-plugging process for printed circuit boards. First, by using the inverse model of the equipment transfer function, the continuous pressure control curve is transformed into a precise voltage-time series, effectively overcoming the dynamic response limitations of the equipment itself. This allows the actual filling pressure to closely match the theoretical curve, thereby achieving refined pressure control for different local via groups. Second, the combination of the scraper position sensor and phase-locked loop control ensures precise spatial synchronization of pressure commands, avoiding uneven filling caused by positional deviations. This is especially effective when handling pressure switching at via group boundaries, enabling seamless transitions. Finally, real-time monitoring of the actual filling pressure and dynamic compensation adjustments provide strong adaptive capabilities for the via-plugging process. This allows for timely correction of pressure deviations caused by various uncertainties (such as resin characteristic fluctuations and environmental changes), effectively preventing underfill defects and significantly improving the via-plugging yield and production efficiency.

[0068] After generating differentiated control instructions, the above method also includes: The total resin demand, doctor blade travel speed, initial resin volume in front of the doctor blade, and step response time constant of the ink supply unit and the pressing mechanism are obtained for each of the multiple local hole groups. Along the direction of the doctor blade's movement, the resin requirement of each local hole group is subtracted in sequence and the basic ink supply flow rate is added to generate a spatial prediction curve of the resin inventory in front of the doctor blade. Set a critical resin inventory threshold required to maintain minimum filling pressure, scan the spatial prediction curve, and determine the critical spatial point where the predicted inventory will fall below the critical resin inventory threshold. Multiply the step response time constant by the current speed of the scraper to obtain the compensation advance distance, and superimpose the ink supply acceleration signal and the pressure enhancement signal into the original control command at a position upstream of the critical space point and at a distance from the compensation advance distance; The superimposed control commands are sent to the pressure execution unit and the ink supply unit, so that the compensation action has actually taken effect when the scraper moves to the critical space point.

[0069] Specifically, a series of key parameters need to be obtained before the orifice plugging operation begins. Among these, the overall resin requirement refers to the total amount of resin required to fill each local orifice group, which can be calculated by summing the theoretical volumes of all orifices within the orifice group. The doctor blade set travel speed is the preset doctor blade movement speed of the orifice plugging device during operation. The initial resin volume before the doctor blade refers to the amount of resin accumulated at the leading edge of the doctor blade when the doctor blade begins to travel. The step response time constant of the ink supply unit and the pressing mechanism refers to the time required for these actuators to reach a steady state from receiving the command, reflecting the dynamic response characteristics of the equipment.

[0070] The spatial prediction curve for the resin inventory at the blade tip is generated by simulating the blade's movement across the board surface. Specifically, as the blade moves along a predetermined path, it consumes resin based on the resin requirements of the local aperture groups it passes through, while the ink supply unit continuously replenishes the basic ink supply flow. By accumulating these consumption and replenishment amounts along the blade's direction of travel, the resin inventory at the blade tip at different spatial locations can be predicted.

[0071] In practical applications, the critical resin level threshold is a safety value set based on experience or experimental data. When the resin level before the cutter is lower than this threshold, the plugging pressure will be difficult to maintain above the minimum filling pressure limit, which may lead to underfilling. By scanning the spatial prediction curve, the spatial location where the resin level before the cutter is about to fall below this critical threshold can be identified, i.e., the critical spatial point.

[0072] Furthermore, the calculation of the compensation advance distance is to ensure that the compensation action can take effect in a timely manner. Due to the response delay of the ink supply unit and the pressure mechanism, the compensation command needs to be issued in advance. This advance distance is determined by the product of the device's step response time constant and the current speed of the doctor blade. At a position upstream of the critical spatial point and at a distance from this compensation advance distance, the system superimposes an ink supply acceleration signal and a pressure enhancement signal onto the original differentiated control command. The ink supply acceleration signal aims to increase the resin supply rate, while the pressure enhancement signal is used to increase the squeezing effect of the doctor blade on the resin to increase the local filling pressure.

[0073] Therefore, the superimposed control commands are sent to the pressure execution unit and the ink supply unit. Through this advance control, when the doctor blade actually travels to the predicted critical space point, the compensation actions of ink supply acceleration and pressure enhancement have already taken full effect, thereby effectively avoiding the risk of pressure drop and underfilling caused by insufficient resin.

[0074] Through the above technical solution, this application enables more refined and proactive control over the resin via-filling process of printed circuit boards. Compared to solutions that rely solely on real-time detection for feedback adjustment, this application significantly reduces the risk of localized underfilling due to equipment response lag by predicting the resin inventory before the die-cutting process and compensating in advance, especially in areas with drastic changes in via density or high resin demand. This proactive, preventative compensation mechanism not only improves the uniformity and integrity of via filling and reduces the defect rate, but also enhances the stability and reliability of the entire manufacturing process, thereby effectively improving the overall manufacturing quality and production efficiency of printed circuit boards.

[0075] In response, this application further proposes a more refined and intelligent method for adjusting the calculation coefficients corresponding to the boundary of the aforementioned filling pressure parameters, specifically including: Receive quality inspection data containing defect coordinates and defect types, wherein the defect types include underfill depth and overfill height; The defect coordinates are spatially aligned with the curve of actual filling pressure changing with position recorded in real time during this plugging process. For the defect coordinates located within a local hole group, if the actual filling pressure at that location is not lower than the minimum filling pressure lower limit but there is underfilling, the resistance correction coefficient corresponding to the local hole group is increased; if the actual filling pressure at that location is not higher than the maximum filling pressure upper limit but there is overfilling, the preset critical flow velocity corresponding to the local hole group is decreased. An empirical coefficient matrix is ​​constructed using the pore group feature vector, which is composed of the dominant pore size within the local pore group and the overall resin demand, as an index. Using the sum of the squares of the underfill depth and overfill height as the error function, the Kalman gain of the current error observation is determined based on the area ratio of the defect region and the detection confidence. The Kalman gain is then used to weight and fuse the current error observation with the model prediction error to obtain a smoothed effective error. The resistance correction coefficients and the preset critical flow velocity in the empirical coefficient matrix are updated by gradient descent using the smoothed effective error, and boundary constraints based on physical limits are applied to the updated coefficients.

[0076] Specifically, receiving quality inspection data refers to acquiring defect information on the printed circuit board after the via plugging operation through methods such as automated optical inspection (AOI), X-ray inspection, or manual visual inspection. This information not only includes the precise coordinates of the defects on the board surface but also details the type of defect, such as underfilling due to insufficient resin filling or overfilling due to resin overflowing from the via opening. For underfilling, its depth is quantified and recorded; for overfilling, its height is also quantified and recorded. This quantified defect information forms the basis for subsequent precise adjustments.

[0077] Spatial alignment of the defect coordinates with the actual filling pressure curve refers to precisely matching the defect location indicated in the quality inspection data with the actual filling pressure data recorded in real time by the pressure sensor during the plugging process, which changes with the scraper's position. This spatial alignment accurately identifies the actual filling pressure value applied by the plugging equipment at the defect location, thus establishing a direct correlation between the defect and the pressure conditions.

[0078] In practical applications, the handling logic for defects located within specific local pore groups differs. If underfill is detected within a local pore group, and the actual filling pressure at that location is not lower than the minimum filling pressure required to fill micro-blind holes, this indicates that even if the minimum pressure requirement is met, the resin has not been fully filled, possibly due to an underestimation of the pore resistance. In this case, the resistance correction factor corresponding to that local pore group will be increased to improve the estimation of the required pressure for this type of pore group in subsequent calculations. Conversely, if overfill is detected, and the actual filling pressure at that location is not higher than the maximum allowable filling pressure limit to prevent overfilling of mechanically perforated holes, this may mean that the critical flow rate of the resin has been overestimated, causing the resin to flow too quickly within the pores. In this case, the preset critical flow rate corresponding to that local pore group will be decreased to more strictly limit the filling speed in subsequent calculations and avoid overfilling.

[0079] Furthermore, to achieve more intelligent and generalized adjustments, this application constructs an empirical coefficient matrix. This matrix uses a pore group feature vector, composed of the dominant pore size (e.g., average pore diameter, average pore depth) and the overall resin requirement of the local pore group, as an index. This means that for local pore groups with similar pore characteristics and resin requirements, their empirical coefficients can be shared or referenced, thereby accelerating the learning process and improving the accuracy of adjustments.

[0080] Based on this, the sum of the squares of the underfill depth and overfill height is used as an error function to quantify the overall quality deviation of the plugging operation. To improve the robustness and stability of the adjustment, this application introduces a Kalman gain. This Kalman gain is dynamically determined based on the area ratio of the defect region and the detection confidence level, and is used to weight and fuse the observed error with the model prediction error to obtain a smooth and effective error estimate. This fusion mechanism can effectively suppress measurement noise and uncertainty, making the error estimate more accurate.

[0081] Finally, the resistance correction coefficients and the preset critical flow velocity in the empirical coefficient matrix are updated using gradient descent based on the smoothed effective error. Gradient descent is an iterative optimization algorithm that gradually reduces the error by adjusting the coefficients along the direction of the steepest descent of the error function. Simultaneously, to ensure the physical rationality of the adjustment results, boundary constraints based on physical limits are applied to the updated coefficients. For example, the resistance correction coefficient cannot be negative, and the critical flow velocity cannot exceed the maximum flow velocity of the resin, preventing unrealistic parameter values.

[0082] In response, this application proposes an optimization scheme for intelligent processing when the hole feature information is incomplete, so as to improve the robustness and adaptability of the hole plugging process.

[0083] The above method also includes: when acquiring the hole feature information of the printed circuit board to be processed, if the depth information of some holes is detected to be missing, spatial interpolation estimation is performed using the depth information of known holes of the same type on the same board surface, or the standard depth parameters of the same product model in the historical database are called to fill the gap, and after filling, the pressure parameter boundary of the local hole group where the hole is located is multiplied by a preset safety factor to expand the pressure tolerance range; during the execution of the differentiated control command, the real-time filling status of the missing information area is recorded, and the area is marked as a risk area to be reviewed.

[0084] Specifically, when the system acquires hole feature information of a printed circuit board, if it finds that the depth data of some holes could not be successfully read or recorded, i.e., there is a lack of depth information, a data filling mechanism will be activated. This mechanism can use two methods: one is spatial interpolation estimation, which analyzes the depth information of other known holes with similar types and location characteristics on the same board surface and uses interpolation algorithms (such as linear interpolation, nearest neighbor interpolation, or Kriging interpolation) to predict and fill in the missing depth values. The other method is to call a historical database, which retrieves and uses standard or average depth parameters from historical production data storing the same product model or similar process parameters to fill in the missing information.

[0085] After the depth information is filled, to address potential uncertainties arising from estimation or filling, the system adjusts the filling pressure parameter boundaries of the local orifice group containing that orifice. Specifically, it multiplies the pressure parameter boundaries of the local orifice group (including the minimum filling pressure lower limit and the maximum filling pressure upper limit) by a preset safety factor. This safety factor is typically greater than 1, its purpose being to moderately expand the pressure tolerance range, providing greater margin of error for subsequent orifice plugging operations, thereby reducing the risk of orifice plugging defects caused by estimation errors.

[0086] Furthermore, during the resin plugging operation performed by the plugging equipment according to differentiated control instructions, the system specifically records the real-time filling status of areas that previously had missing information and have been filled. This includes changes in key parameters such as actual filling pressure, doctor blade speed, and ink supply volume as a function of spatial location. These areas are then marked as high-risk areas requiring review. This is intended to alert operators or subsequent quality control systems to focus on and review these areas, enabling timely detection and resolution of potential quality issues and providing data support for subsequent process optimization.

[0087] Furthermore, even after filling in the depth information, the pressure boundary calculated directly using the filled data may still pose risks due to potential biases in estimations or historical data. To address this, this solution introduces a preset safety factor, expanding the permissible pressure range for this local orifice group to provide additional tolerance for the plugging operation. This means that even if the filled depth information deviates somewhat from the actual situation, the plugging equipment can still operate within a relatively relaxed pressure range, effectively reducing the risk of plugging defects caused by data uncertainty.

[0088] Meanwhile, during the execution of differentiated control instructions, the real-time filling status of areas with missing information is recorded and marked as risk areas requiring review. This enables the system to continuously monitor potential risk points. Once a significant deviation between the actual filling status and the expected status is detected, timely intervention or focused review during subsequent quality inspections can be implemented, thus forming a closed-loop risk management process and further improving the reliability and quality control level of the entire plugging process.

[0089] Through the above technical solution, this application can effectively address the problem of incomplete hole feature information in the printed circuit board manufacturing process, significantly improving the robustness and adaptability of the via plugging process. Compared with the basic solution, which may face the risk of control failure or increased defect rate when data is missing, this solution, through an intelligent data filling mechanism, ensures that even when some depth information is missing, a relatively reasonable filling pressure parameter boundary can still be determined for local hole groups, avoiding production interruptions or quality control failures caused by incomplete data.

[0090] Furthermore, the introduction of a safety factor to expand the pressure tolerance range provides necessary error margin for via plugging operations, effectively reducing the risk of via plugging defects caused by data estimation errors, thereby improving the first-pass yield. Simultaneously, real-time status recording and risk marking of information-deficient areas enable continuous monitoring and focused attention on potential risks during production, facilitating timely identification and resolution of problems and providing valuable data support for subsequent process optimization and quality traceability. These additional technical features work together to ensure the entire printed circuit board resin via plugging manufacturing process maintains efficient and stable operation even in the face of uncertainty, significantly improving the level of production intelligence and product quality.

[0091] In response, this application proposes an intelligent control method for the resin via plugging manufacturing process of printed circuit boards. The method further includes: when the pressure execution unit and ink supply unit of the via plugging equipment perform resin via plugging operation according to the differentiated control instructions, if the sensor signal used to provide feedback on the actual filling pressure is lost, the method immediately switches from the feedback adjustment mode based on the actual filling pressure to the open-loop predictive execution mode, and drives the pressure execution unit and ink supply unit to complete the via plugging process of the current board entirely according to the differentiated control instructions. After the current board is completed, the method automatically suspends subsequent production tasks and triggers an equipment maintenance alarm.

[0092] Specifically, during the resin plugging operation of the plugging equipment, the pressure actuator and ink supply unit operate according to pre-generated differentiated control commands. At this time, the system continuously monitors the sensor signal used to provide feedback on the actual filling pressure. When the sensor signal is detected as lost, for example due to sensor failure, broken connecting wires, or signal interference, resulting in the inability to obtain valid actual filling pressure data, the system immediately takes countermeasures. The phrase "switching from a feedback adjustment mode based on the actual filling pressure to an open-loop predictive execution mode" means that under normal operation, the system is in a closed-loop control state, that is, it compares the real-time detected actual filling pressure with the target pressure and makes dynamic adjustments. Once the sensor signal is lost, the system will no longer rely on real-time feedback but will switch to a preset control mode that does not rely on real-time feedback. In this open-loop predictive execution mode, the pressure actuator and ink supply unit of the plugging equipment will be driven entirely according to the pre-generated differentiated control commands, without real-time pressure feedback adjustment. This means that even without feedback on the actual filling pressure, the system will complete the plugging process of the current printed circuit board according to the predetermined pressure curve and speed combination. "Automatically pausing subsequent production tasks and triggering equipment maintenance alarms after the current board is completed" means that, to avoid potential impacts on the quality of subsequent products in the absence of critical feedback, the system will automatically stop subsequent production tasks after the via plugging operation is completed on the currently being processed printed circuit board. Simultaneously, the system will immediately trigger an equipment maintenance alarm to notify operators or the maintenance team, enabling timely troubleshooting and repair of sensor malfunctions and ensuring the equipment returns to normal operation.

[0093] Through the above technical solution, this application significantly improves the robustness and reliability of the resin-filled via manufacturing process for printed circuit boards. When a critical sensor malfunctions and causes signal loss, this solution can prevent sudden production line shutdowns and scrapping of products being processed, thereby effectively reducing production losses and downtime. By allowing the current board to complete processing in open-loop mode, production continuity is ensured. Simultaneously, by promptly pausing subsequent tasks and triggering maintenance alarms, the overall production quality is kept controllable, and timely equipment maintenance is promoted, reducing potential risks and operating costs caused by sensor failures.

[0094] In some embodiments, reference is made to Figure 3This application proposes an intelligent control system for the resin via plugging manufacturing process of printed circuit boards, comprising: an acquisition unit 301, used to acquire the hole feature information of the printed circuit board to be processed, and divide the board surface into multiple local hole groups according to the hole feature information, the hole feature information including the type, geometric parameters and position of each hole, each local hole group being formed by the aggregation of holes whose filling requirement similarity meets preset conditions; a determination unit 302, used to determine the filling pressure parameter boundary of each local hole group according to the hole type and geometric parameters of the holes it contains and preset resin rheological property data, the filling pressure parameter boundary including the minimum filling pressure lower limit required for filling micro-blind vias and the maximum filling pressure upper limit allowed to prevent overfilling of mechanical through-holes; and a generation unit 303, used to detect conflicts between the filling pressure parameter boundaries of different local hole groups along a predetermined scraper travel path, and when the minimum filling pressure lower limit of the first local hole group is greater than the maximum filling pressure upper limit of the second local hole group, a conflict is detected. The system generates differentiated control instructions for decoupling conflicts. The first local orifice group is a local orifice group with a minimum filling pressure lower limit higher than adjacent orifice groups, and the second local orifice group is a local orifice group with a maximum filling pressure upper limit lower than adjacent orifice groups. The differentiated control instructions include at least applying different filling pressures to different local orifice group regions, or splitting a complete filling stroke into multiple strokes with different pressure and speed combinations. The control unit 304 controls the pressure execution unit and ink supply unit of the orifice plugging device to perform resin orifice plugging operations according to the differentiated control instructions. During operation, when the actual filling pressure is detected to be lower than the minimum filling pressure lower limit corresponding to the current position, it triggers compensation adjustments to the ink supply rate and downward stroke. The adjustment unit 305 acquires quality inspection data after orifice plugging, maps the defect location indicated by the quality inspection data to the record of actual filling pressure changes with spatial position during execution, and adjusts the calculation coefficients corresponding to the filling pressure parameter boundaries based on the mapping results.

[0095] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A smart control method for the resin via plugging manufacturing process of printed circuit boards, characterized in that, include: Obtain hole feature information of the printed circuit board to be processed, and divide the board surface into multiple local hole groups according to the hole feature information. The hole feature information includes the type, geometric parameters and position of each hole. Each local hole group is formed by aggregating holes whose filling requirement similarity meets the preset conditions. For each of the aforementioned local pore groups, the filling pressure parameter boundary of the local pore group is determined based on the pore type and geometric parameters of the pores it contains and the preset resin rheological property data. The filling pressure parameter boundary includes the minimum filling pressure lower limit required to fill micro blind holes and the maximum filling pressure upper limit allowed to prevent overfilling of mechanical through holes. Along a predetermined scraper travel path, conflicts between the filling pressure parameter boundaries of different local hole groups are detected. When the minimum filling pressure lower limit of the first local hole group is greater than the maximum filling pressure upper limit of the second local hole group, a differentiated control command for decoupling the conflict is generated. The first local hole group is a local hole group whose minimum filling pressure lower limit is higher than that of the adjacent hole group, and the second local hole group is a local hole group whose maximum filling pressure upper limit is lower than that of the adjacent hole group. The differentiated control command includes at least applying different filling pressures in different local hole group regions, or splitting a complete filling stroke into multiple strokes with different pressure and speed combinations. The pressure execution unit and ink supply unit of the control plugging device perform resin plugging operation according to the differentiated control instructions. During the operation, when the actual filling pressure is detected to be lower than the minimum filling pressure limit corresponding to the current position, the compensation adjustment of the ink supply rate and the downward stroke is triggered. After obtaining the quality inspection data after plugging the hole, the defect location indicated by the quality inspection data is mapped to the actual filling pressure change with spatial position recorded during the execution process, and the calculation coefficient corresponding to the boundary of the filling pressure parameter is adjusted based on the mapping result.

2. The method according to claim 1, characterized in that, The step of acquiring the hole feature information of the printed circuit board to be processed, and dividing the board surface into multiple local hole groups based on the hole feature information, includes: The initial type markings, center coordinates for spatial positioning, hole diameter, and hole depth of each hole to be plugged are analyzed from the manufacturing design documents of the printed circuit board. Based on the preset aperture threshold, the depth characteristics represented by the aperture depth, and the initial type label, each aperture is classified into a micro blind aperture or a mechanical through aperture, and an overfill sensitivity is determined for each aperture. The overfill sensitivity is determined based on the ratio of the square of the aperture to the square root of the aperture depth, and is used to characterize the risk tendency of the aperture to overfill under a given filling pressure. Based on the center coordinates of each hole, each hole is assigned to a grid that divides the plate surface. The side length of the grid is determined by the effective working width of the scraper and the lateral diffusion distance of the resin. The number of micro blind holes and mechanical through holes, the average hole diameter, and the total volume in each grid are counted. Using a preset spatial clustering algorithm, adjacent grids that meet preset similarity conditions are aggregated into local pore groups based on the spatial adjacency between grids and the distance metric constructed from the difference in average diameter of micro-blind holes, the difference in average diameter of mechanical through holes, and the weighted average of overfill sensitivity within the grid. The state vector of each local pore group is output, and the state vector includes the spatial boundary coordinates determined by the coordinate range of the aggregated grids, the dominant pore size represented by the average diameter and average depth of the same type of pores within the pore group, and the total resin requirement determined by the sum of the theoretical volumes of each pore within the pore group.

3. The method according to claim 1, characterized in that, For each of the aforementioned local pore groups, the filling pressure parameter boundary of the local pore group is determined based on the pore type and geometric parameters of the pores it contains, as well as preset resin rheological property data, including: For a local group of holes containing micro-blind holes, the shear rate experienced by the resin when entering the micro-blind holes is estimated based on the scraper travel speed. The apparent dynamic viscosity of the resin at this shear rate is obtained from the preset resin rheological property data. The minimum filling pressure lower limit is determined by combining the resin surface tension coefficient, micro-blind hole diameter, micro-blind hole depth, reverse air pressure generated by the adiabatic compression of air at the bottom of the hole during the filling process, and the preset resistance correction coefficient. For a localized group of pores containing mechanical through-holes, the upper limit of the maximum filling pressure is determined based on the apparent dynamic viscosity of the resin under the corresponding shear conditions, the preset critical flow rate, and the pore diameter and depth of the mechanical through-holes. The minimum filling pressure lower limit and the maximum filling pressure upper limit are used together as the filling pressure parameter boundary for this local pore group.

4. The method according to claim 1, characterized in that, The generation of differentiated control instructions for decoupling conflicts includes: The plate surface is divided into continuous strip regions along the scraper's travel path, and the minimum filling pressure lower limit and the maximum filling pressure upper limit of all local hole groups in each strip region are extracted. For each strip region, determine whether there is an intersection between the boundaries of the filling pressure parameters of each local pore group; when the lower limit of the minimum filling pressure of any local pore group is greater than the upper limit of the maximum filling pressure of another local pore group, the intersection of their pressure intervals is empty, it is determined that there is a parameter conflict, and the local pore group with the higher lower limit of the minimum filling pressure is identified as the first local pore group, and the local pore group with the lower upper limit of the maximum filling pressure is identified as the second local pore group; when the pressure intervals of all local pore groups have a common intersection, the median of the intersection is taken as the target filling pressure of the strip region; For strip regions with parameter conflicts, a decoupling strategy is selected based on the pore density distribution of the first local pore group and the second local pore group in the strip region: if the mechanical through-hole density is lower than the first preset threshold, a high-pressure priority strategy is adopted, and the local fluid pressure distribution is changed by adjusting the scraper angle of attack before the scraper reaches the strip region; if the micro-blind hole density and the mechanical through-hole density are not lower than the corresponding preset threshold, control commands for at least two differentiated strokes are generated, wherein the pressure of the first stroke is set to be lower than the maximum filling pressure upper limit and the scraper speed is higher than the preset normal speed, so that the resin flows into the mechanical through-hole first; the pressure of the second stroke is set to be higher than the minimum filling pressure lower limit and the scraper speed is lower than the preset normal speed, and the filling of the micro-blind holes is completed by utilizing the liquid resistance formed by the resin already filled in the mechanical through-hole. The filling pressure setting value of each strip area or the pressure and speed combination of each stroke is arranged along the scraper's travel position to form a continuous pressure control curve, which serves as the differentiated control command.

5. The method according to claim 4, characterized in that, The pressure actuation unit and ink supply unit of the control plugging device perform resin plugging operation according to the differentiated control command, including: Based on the inverse model of the device transfer function, which includes aerodynamic delay, mechanical stiffness, and fluid damping, the continuous pressure control curve is converted into a voltage time series that drives the pressure actuator. The system receives real-time position feedback from the scraper position sensor and adjusts the issuing clock frequency of the voltage time sequence through a phase-locked loop control, so that the change in pressure command is spatially synchronized with the boundary of the local hole group. During the movement of the doctor blade, the actual filling pressure is continuously compared with the continuous pressure control curve. When the actual filling pressure is lower than the minimum filling pressure limit corresponding to the current position, an accelerated ink supply command is sent to the ink supply unit and the doctor blade pressure is increased until the actual filling pressure returns to the preset target range.

6. The method according to claim 1, characterized in that, After generating the differentiated control instructions, the method further includes: The total resin demand, doctor blade travel speed, initial resin volume in front of the doctor blade, and step response time constant of the ink supply unit and the pressing mechanism are obtained for each of the multiple local hole groups. Along the direction of the doctor blade's movement, the resin requirement of each local hole group is subtracted in sequence and the basic ink supply flow rate is added to generate a spatial prediction curve of the resin inventory in front of the doctor blade. Set a critical resin inventory threshold required to maintain minimum filling pressure, scan the spatial prediction curve, and determine the critical spatial point where the predicted inventory will fall below the critical resin inventory threshold. Multiply the step response time constant by the current speed of the scraper to obtain the compensation advance distance, and superimpose the ink supply acceleration signal and the pressure enhancement signal into the original control command at a position upstream of the critical space point and at a distance from the compensation advance distance; The superimposed control commands are sent to the pressure execution unit and the ink supply unit, so that the compensation action has actually taken effect when the scraper moves to the critical space point.

7. The method according to claim 1, characterized in that, The process involves acquiring quality inspection data after plugging the holes, mapping the defect locations indicated by the quality inspection data to records of actual filling pressure changes with spatial position during execution, and adjusting the calculation coefficients corresponding to the filling pressure parameter boundaries based on the mapping results. This includes: Receive quality inspection data containing defect coordinates and defect types, wherein the defect types include underfill depth and overfill height; The defect coordinates are spatially aligned with the curve of actual filling pressure versus position recorded in real time during the plugging process. For a defect coordinate located within a local hole group, if the actual filling pressure at the defect coordinate is not lower than the minimum filling pressure lower limit but there is underfilling, the resistance correction coefficient corresponding to the local hole group is increased; if the actual filling pressure at the defect coordinate is not higher than the maximum filling pressure upper limit but there is overfilling, the preset critical flow velocity corresponding to the local hole group is decreased. An empirical coefficient matrix is ​​constructed using a pore group feature vector, composed of the dominant pore size within the local pore group and the overall resin demand, as an index. The Kalman gain of the current error observation is determined based on the area ratio of the defect region and the detection confidence level, using the sum of the squares of the underfill depth and overfill height as the error function. This Kalman gain is then used to weight and fuse the current error observation with the model prediction error to obtain a smoothed effective error. The smoothed effective error is then used to update the resistance correction coefficients and the preset critical flow velocity in the empirical coefficient matrix using gradient descent, and boundary constraints based on physical limits are applied to the updated coefficients.

8. The method according to claim 1, characterized in that, The method further includes: when acquiring the hole feature information of the printed circuit board to be processed, if the depth information of some holes is detected to be missing, spatial interpolation estimation is performed using the depth information of known holes of the same type on the same board surface, or the standard depth parameters of the same product model in the historical database are called to fill the gap, and after filling, the pressure parameter boundary of the local hole group where the hole is located is multiplied by a preset safety factor to expand the pressure tolerance range; during the execution of the differentiated control command, the real-time filling status of the missing information area is recorded, and the area is marked as a risk area to be reviewed.

9. The method according to claim 1, characterized in that, The method further includes: when the pressure execution unit and ink supply unit of the control plugging device perform resin plugging operation according to the differentiated control command, if the sensor signal used to provide feedback on the actual filling pressure is lost, the method immediately switches from the feedback adjustment mode based on the actual filling pressure to the open-loop predictive execution mode, and drives the pressure execution unit and ink supply unit to complete the plugging process of the current board entirely according to the differentiated control command. After the current board is completed, the method automatically pauses subsequent production tasks and triggers an equipment maintenance alarm.

10. An intelligent control system for the resin via plugging manufacturing process of printed circuit boards, characterized in that, include: The acquisition unit is used to acquire hole feature information of the printed circuit board to be processed, and divide the board surface into multiple local hole groups according to the hole feature information. The hole feature information includes the type, geometric parameters and position of each hole. Each local hole group is formed by aggregating holes whose filling requirement similarity meets the preset conditions. The determining unit is used to determine the filling pressure parameter boundary of each local pore group based on the pore type and geometric parameters of the pores it contains and preset resin rheological property data. The filling pressure parameter boundary includes the minimum filling pressure lower limit required to fill micro blind holes and the maximum filling pressure upper limit allowed to prevent overfilling of mechanical through holes. A generation unit is used to detect conflicts between the boundaries of filling pressure parameters of different local hole groups along a predetermined scraper travel path. When the minimum filling pressure lower limit of a first local hole group is greater than the maximum filling pressure upper limit of a second local hole group, a differentiated control instruction for decoupling the conflict is generated. The first local hole group is a local hole group whose minimum filling pressure lower limit is higher than that of the adjacent hole group, and the second local hole group is a local hole group whose maximum filling pressure upper limit is lower than that of the adjacent hole group. The differentiated control instruction includes at least applying different filling pressures in different local hole group regions, or splitting a complete filling stroke into multiple strokes with different pressure and speed combinations. The control unit is used to control the pressure actuation unit and ink supply unit of the plugging device to perform resin plugging operation according to the differentiated control instructions, and during the operation, when the actual filling pressure is detected to be lower than the minimum filling pressure limit corresponding to the current position, it triggers the compensation adjustment of ink supply rate and downward stroke. The adjustment unit is used to acquire the quality inspection data after plugging the hole, map the defect location indicated by the quality inspection data with the actual filling pressure change record recorded during the execution process, and adjust the calculation coefficient corresponding to the boundary of the filling pressure parameter based on the mapping result.

Images