Intelligent industrial control system for HDI plate electroplating hole filling process

Abstract

The invention discloses an intelligent industrial control system for an HDI plate electroplating hole filling process, belongs to the field of circuit board intelligent manufacturing industry, and is applied to electroplating equipment comprising a PCB (Printed Circuit Board) as a main cathode, a conductive hanger and a contact thereof, and an auxiliary shielding electrode arranged at the periphery of the contact. The system comprises an initial electric field non-uniformity modeling module, a dynamic electric field compensation control module, a post-plating quality evaluation module and a model parameter adaptive correction module. According to the invention, a post-plating quality evaluation and model parameter adaptive correction module is introduced, and a complete prediction-control-evaluation-correction closed-loop system is constructed. This enables the system to learn, automatically correct and optimize its internal model from each production result to accommodate slow drift of process conditions.

Description

Technical Field

[0001] The present invention relates to the field of intelligent manufacturing industrial control of circuit boards, and in particular to an intelligent industrial control system for an HDI board electroplating hole filling process. Background Art

[0002] During the manufacturing process of high-density interconnect (HDI) printed circuit boards, electroplating and filling vias is a critical step in ensuring reliable electrical interconnection between conductive layers. In traditional electroplating processes, the PCB is typically secured by a metal conductive hanger and immersed in the plating solution as the entire cathode. However, the hanger's contacts, which hold the PCB and conduct electricity, themselves become part of the electric field, significantly shielding and distorting the surrounding electric field lines.

[0003] This physical effect causes the plating current density at the edge of the PCB near the hanger contacts to be significantly lower than that at the center of the board. This ultimately results in an uneven distribution of the copper layer thickness after electroplating, with thinner edges and thicker centers. This unevenness seriously affects the final yield and high-frequency electrical performance of the HDI board.

[0004] To address this issue, existing technologies typically employ passive, empirically driven compensation methods, such as adding physical baffles and adjusting anode position or shape. These methods are limited in effectiveness, unable to accurately address electric field distortion and unable to adapt to variations in PCB board designs from different batches. They also lack intelligence, precision, and adaptability. Therefore, a technical solution that can proactively, accurately, and intelligently address the uneven plating caused by the hanger contacts is urgently needed. Summary of the Invention

[0005] The purpose of the present invention is to overcome the above-mentioned shortcomings of the prior art and provide an intelligent industrial control system for the HDI board electroplating and hole filling process. The control system aims to solve the problems of shielding and disturbance of the electric field at the edge of the PCB board caused by the contacts of the electroplating rack, so as to achieve uniform plating thickness across the entire PCB board surface.

[0006] The technical solution of the present invention is: an intelligent industrial control system for the HDI board electroplating hole filling process, which is applied to an electroplating equipment including a PCB board as a main cathode, a conductive hanger and its contacts, and an auxiliary shielding electrode arranged around the contacts. The system includes: Initial electric field non-uniformity modeling module: Based on the preset electric field model, and according to the geometric layout parameters of the conductive rack contacts and the physical property parameters of the electroplating process, an initial electric field non-uniformity map is generated; Dynamic electric field compensation control module: used to calculate and output compensation control signals for each auxiliary shielding electrode in real time during the electroplating process based on the degree of deviation of each region indicated by the initial electric field non-uniformity map, so as to drive the auxiliary shielding electrode to generate a local compensation electric field to offset the electric field disturbance caused by the conductive hanger contact; A post-plating quality assessment module is configured to collect the actual coating thickness at multiple preset sampling points on the PCB board after the electroplating process is completed to form an actual coating thickness dataset; and compare the actual coating thickness dataset with the target coating thickness and the initial electric field non-uniformity map to generate a model error map, which is used to quantify the prediction deviation of the initial electric field non-uniformity map; The model parameter adaptive correction module is used to calculate one or more correction values ​​of the disturbance intensity coefficients related to the conductive hanger contacts based on the model error map; and use the correction values ​​to update the disturbance intensity coefficients used by the initial electric field non-uniformity modeling module.

[0007] In this embodiment, the initial electric field non-uniformity modeling module is used to generate the initial electric field non-uniformity map, specifically including: dividing the surface of the PCB board into a plurality of independent influence areas corresponding to each conductive hanger contact according to the geometric layout parameters of the conductive hanger contacts; and within each of the influence areas, based on a preset distance attenuation model, calculating the electric field disturbance intensity generated by the corresponding conductive hanger contact, and superimposing the disturbance intensities generated by all contacts, thereby obtaining the expected deviation degree at any position on the surface of the PCB board to construct the initial electric field non-uniformity map.

[0008] In this embodiment, in the distance attenuation model, the disturbance intensity generated by any conductive hanger contact at any point on the PCB board surface depends on the disturbance intensity coefficient related to the contact itself and the geometric distance between the arbitrary point and the contact.

[0009] In this embodiment, the dynamic electric field compensation control module is used to calculate the compensation control signal, and the specific steps are: S1: for each auxiliary shielding electrode, extracting the average deviation degree of its coverage area from the initial electric field non-uniformity map; S2: Multiply the average deviation degree with a compensation gain coefficient preset for the electrode and the real-time electrical parameters of the main circuit of the electroplating equipment to calculate the compensation voltage value required by the auxiliary shielding electrode at that moment, and output the compensation voltage value as the compensation control signal.

[0010] In this embodiment, the post-plating quality assessment module is used to generate the model error map, specifically: for each preset sampling point, the normalized deviation of the actual coating thickness collected relative to the target coating thickness is calculated; the normalized deviation is compared with the deviation degree predicted by the initial electric field non-uniformity map at this point, and the difference is defined as the model prediction error of the point. The model prediction errors of all sampling points together constitute the model error map.

[0011] In this embodiment, the post-plating quality assessment module is also used to: calculate the ratio of the sample standard deviation of all thickness values ​​in the data set to the arithmetic mean based on the actual coating thickness data set, and define the ratio as a whole-plate coating uniformity deviation index, which is used to quantitatively assess the overall quality of this electroplating.

[0012] In this embodiment, the model parameter adaptive correction module is used to calculate the correction amount of the disturbance intensity coefficient, specifically: for each conductive hanger contact, the average model prediction error of the model error map in its corresponding influence area is extracted, and the average error is multiplied by a preset learning rate parameter to obtain the correction amount of the disturbance intensity coefficient corresponding to the contact.

[0013] In this embodiment, the model parameter adaptive correction module further includes a hierarchical correction strategy before performing the update, for comparing the whole plate coating uniformity deviation index with a preset low deviation threshold and a preset high deviation threshold; If the uniformity deviation index is lower than the low deviation threshold, the current model is deemed accurate and no update is performed; if the uniformity deviation index is between the low deviation threshold and the high deviation threshold, the update is performed.

[0014] In this embodiment, the graded correction strategy is also used: if the whole plate coating uniformity deviation index is higher than the high deviation threshold, then while executing the update, the system will also trigger an alarm signal to indicate that there may be abnormalities in the equipment hardware or the chemical state of the plating solution that require manual intervention.

[0015] The present invention provides an intelligent industrial control system for the HDI board electroplating and hole filling process through improvements. Compared with the existing technology, it has the following improvements and advantages: This invention incorporates post-plating quality assessment and model parameter adaptive correction modules, creating a complete prediction-control-assessment-correction closed-loop system. This enables the system to learn from each production result, automatically correcting and optimizing its internal model to adapt to slow drifts in process conditions, such as hanger wear and aging of the coating solution. This is a self-learning and adaptive capability completely lacking in existing technologies.

[0016] The quantitative metrics generated by this method, such as the initial electric field non-uniformity map, model error map, and whole-panel uniformity deviation index, provide process engineers with unprecedented insight into the process. In particular, the hierarchical correction strategy intelligently distinguishes between model fine-tuning and major equipment failures, significantly improving the efficiency and accuracy of fault diagnosis.

[0017] This invention uses auxiliary shielding electrodes for active electric field compensation, significantly improving control accuracy and flexibility compared to traditional passive methods such as physical baffles. The intensity and shape of the compensation electric field can be dynamically and precisely adjusted as needed.

[0018] By establishing a mathematical model of electric field inhomogeneity, the present invention shifts compensation for electric field disturbances from relying on operator experience to precise model-based calculations. This makes the compensation process standardized and repeatable, and enables rapid adaptation to different PCB layouts. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The present invention will be further explained below in conjunction with the accompanying drawings and Examples: Figure 1 The present invention is a flowchart of an intelligent industrial control system for an HDI board electroplating hole filling process. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to specific embodiments. Example 1:

[0021] See also Figure 1 The present invention provides a technical solution for an intelligent industrial control system for the HDI board electroplating and hole filling process: an intelligent industrial control system for the HDI board electroplating and hole filling process, applied to an electroplating device including a PCB board as a main cathode, a conductive hanger and its contacts, and auxiliary shielding electrodes arranged around the contacts. The system includes: Initial electric field non-uniformity modeling module: Based on the preset electric field model, and according to the geometric layout parameters of the conductive rack contacts and the physical property parameters of the electroplating process, an initial electric field non-uniformity map is generated; Based on a preset electric field model, and according to the geometric layout parameters of the conductive rack contacts and the physical property parameters of the electroplating process, an initial electric field non-uniformity map is generated; the map is used to characterize the degree of deviation of the expected electroplating current density in various areas of the PCB surface from the target value under uncompensated conditions; Dynamic electric field compensation control module: used to calculate and output compensation control signals for each auxiliary shielding electrode in real time during the electroplating process based on the degree of deviation of each region indicated by the initial electric field non-uniformity map, so as to drive the auxiliary shielding electrode to generate a local compensation electric field to offset the electric field disturbance caused by the conductive hanger contact; The post-plating quality assessment module is used to collect the actual coating thickness at multiple preset sampling points on the PCB board after the electroplating process is completed to form an actual coating thickness dataset; and compare the actual coating thickness dataset with the target coating thickness and the initial electric field non-uniformity map to generate a model error map. The model error map is used to quantify the prediction deviation of the initial electric field non-uniformity map.

[0022] A model parameter adaptive correction module is used to calculate one or more correction values ​​of the disturbance intensity coefficients related to the conductive hanger contacts based on the model error map; and use the correction values ​​to update the disturbance intensity coefficients used by the initial electric field non-uniformity modeling module; In this embodiment, the initial electric field non-uniformity modeling module is used to generate the initial electric field non-uniformity map, specifically including: dividing the surface of the PCB board into a plurality of independent influence areas corresponding to each conductive hanger contact according to the geometric layout parameters of the conductive hanger contacts; and within each of the influence areas, based on a preset distance attenuation model, calculating the electric field disturbance intensity generated by the corresponding conductive hanger contact, and superimposing the disturbance intensities generated by all contacts, thereby obtaining the expected deviation degree at any position on the surface of the PCB board to construct the initial electric field non-uniformity map.

[0023] The geometric layout parameters of the conductive hanger contacts: that is, the position coordinate set of all conductive contacts in the two-dimensional coordinate system ; Physical properties of the electroplating process: for example, the conductivity of the plating solution , target coating thickness wait; Based on the location coordinate set of the touch point The module uses the Voronoi diagram algorithm to divide the entire surface of the PCB into multiple independent and non-overlapping influence areas. Each area Contains all contacts from the kth point A point that is closer to any other contact point than the other one; this division method geometrically clarifies the main influence range of each contact point and provides an accurate spatial basis for subsequent parameter correction.

[0024] In this embodiment, in the distance attenuation model, the disturbance intensity generated by any conductive hanger contact at any point on the PCB surface depends on the disturbance intensity coefficient associated with the contact itself and the geometric distance between the arbitrary point and the contact. In each affected area, the module calculates the electric field disturbance intensity generated by the corresponding contact based on a preset distance attenuation model. The core idea of ​​this model is that the influence of the contact on the electric field decays with increasing distance. The disturbance intensity generated by any conductive hanger contact at a point on the PCB board surface depends on a disturbance intensity coefficient related to the contact itself and the geometric distance between the point and the contact.

[0025] In order to get any point on the board The module linearly superimposes the disturbance intensity generated by all contacts at this point to determine the total expected deviation. The calculation formula is as follows: ; Formula parameter description: : Indicates the coordinate point on the PCB board The initial electric field non-uniformity index at point φ is a dimensionless quantity whose value represents the degree of deviation of the expected electroplating current density at that point from the target value. The larger the value, the more serious the expected deviation, usually indicating that the current density is too low. : The total number of conductive hanger contacts; : The two-dimensional coordinate vector of the k-th contact point; : The disturbance intensity coefficient of the kth contact. This coefficient is the key parameter of this model. Its physical meaning is to characterize the strength of the electric field disturbance generated by a single contact. It is related to the size, material, shape of the contact and the conductivity of the plating solution. This coefficient is usually initially determined by historical data regression analysis or calibration using a standard test board, and is iteratively optimized in the subsequent adaptive correction module. Its dimension can be regarded as , so that dimensionless; : indicates a point With the kth contact The Euclidean distance between : distance decay exponent, a positive real number, usually 1 or 2 based on electric field theory and experimental data; it determines how quickly the disturbance intensity decays with distance; : A very small positive constant, which is used to prevent the calculation point When it coincides with the contact point position, the denominator is zero, which ensures the numerical stability of the calculation.

[0026] The practical significance and function of the formula: This formula constructs a prediction model with clear physical meaning; it quantifies a core physical intuition of electric field disturbance: the influence of the disturbance source (contact) decays with distance; by superimposing the effects of all disturbance sources, a complete map of the initial electric field non-uniformity covering the entire PCB board can be generated. ; This map is the direct basis for subsequent precise compensation.

[0027] Quantitative prediction transforms the empirical and fuzzy edge effect problem into a computable and quantifiable physical model, providing a scientific basis for precise control; The atlas can accurately tell the location and degree of compensation required on the PCB board, achieving a leap from global compensation to local precise compensation.

[0028] In this embodiment, the dynamic electric field compensation control module is used to calculate the compensation control signal, and the specific steps are: S1: for each auxiliary shielding electrode, extracting the average deviation degree of its coverage area from the initial electric field non-uniformity map; S2: multiplying the average deviation degree by a compensation gain coefficient preset for the electrode and the real-time electrical parameters of the main circuit of the electroplating equipment, thereby calculating the compensation voltage value required by the auxiliary shielding electrode at that moment, and outputting the compensation voltage value as the compensation control signal; S1 extracts the average deviation degree for each auxiliary shielding electrode (assuming there are m, indexed as j=1,2,...,m). This module first extracts the average deviation degree from the initial electric field non-uniformity map. Extract the area it covers The average deviation within ; The calculation method is to The values ​​are integrated or averaged; this operation simplifies the complex two-dimensional map information into a scalar control target for each independent control unit.

[0029] Step 2: The compensation voltage calculation module calculates the compensation voltage value required by each auxiliary shielding electrode at the time t in real time according to the following formula: ; Formula parameter description: : The compensation voltage value that needs to be applied to the j-th auxiliary shielding electrode at time t, in volts (V). This is the final control signal output by the module; : The compensation gain coefficient of the jth auxiliary shielding electrode; this is a dimensionless preset parameter obtained through experimental calibration. It characterizes the compensation efficiency of the electrode, that is, the degree of compensation electric field generated by the applied unit voltage. This coefficient is related to the electrode's geometry, size, position, and its relative relationship with the main cathode. : the average non-uniformity index of the j-th electrode coverage area extracted from the atlas, dimensionless; : The real-time electrical parameter of the main circuit of the electroplating equipment, usually the real-time voltage of the main circuit, in volts (V). The significance of introducing this parameter is that the absolute intensity of the electric field disturbance is proportional to the intensity of the main electric field. When the main circuit voltage increases and the electroplating current increases, the required compensation voltage should also increase accordingly to maintain the proportionality of the compensation effect; This formula predicts the static non-uniformity index , and real-time, dynamic process parameters Combined, a dynamic and precise compensation control instruction is generated ; It ensures that the intensity of the compensation electric field can follow the fluctuations of the main electroplating process in real time to achieve dynamic matching.

[0030] Drives the calculated compensation voltage signal The signal is sent to an independent programmable power supply connected to each auxiliary shielding electrode. The power supply drives the corresponding electrode according to the signal, causing it to generate a local compensation electric field with precisely calculated direction and strength. The function of this compensation electric field is to locally correct the electric field distribution and effectively restore the electric field strength in the edge area of ​​the PCB board by offsetting the shielding effect generated by the conductive contact, thereby achieving uniform current density across the entire board surface.

[0031] In this embodiment, the post-plating quality assessment module is used to generate the model error map, specifically: for each preset sampling point, the normalized deviation of the actual coating thickness collected relative to the target coating thickness is calculated; the normalized deviation is compared with the deviation degree predicted by the initial electric field non-uniformity map at this point, and the difference is defined as the model prediction error of the point. The model prediction errors of all sampling points together constitute the model error map.

[0032] Step 1: Collect actual coating thickness data. Use precision measuring equipment such as an X-ray fluorescence thickness gauge to measure the actual coating thickness at multiple sampling points preset on the PCB board. These points should cover areas with high, medium, and low deviations, and form an actual coating thickness data set. .

[0033] Step 2: Generate model error map module. For each sampling point i, calculate its model prediction error. ; The calculation steps are as follows: a. Calculate the normalized deviation of the actual coating thickness at this point: the actual thickness With target thickness Compare and normalize; b. The degree of deviation between this actual deviation and the prediction of the initial model at this point (obtained from the initial map) and compared; the difference between the two is the prediction error of the model at that point; The formula is as follows: ; Formula parameter description: : The model prediction error of the i-th sampling point, dimensionless; its positive and negative signs have clear physical meanings: if it is positive, it means that the actual plating thickness is thicker than the model prediction, indicating that the model's prediction of the disturbance intensity at this point is too high; if it is negative, the opposite is true; : The actual measured thickness of the i-th sampling point; : Target coating thickness set by the process; : The prediction deviation index of the initial electric field inhomogeneity model at the i-th sampling point; The model prediction error of all sampling points By interpolating, a visual model error map can be generated. , which intuitively shows which areas the current model predicts too strongly and which areas the predictions are too weak; In this embodiment, the post-plating quality assessment module is further configured to: calculate, based on the actual coating thickness data set, a ratio of the sample standard deviation to the arithmetic mean of all thickness values ​​in the data set, and define the ratio as a whole-plate coating uniformity deviation index for quantitatively assessing the overall quality of this electroplating; Calculate the uniformity deviation index of the whole plate In order to provide a macro and single evaluation index for the overall quality of this electroplating, the module is also used to calculate the uniformity deviation index of the whole plate coating ; This index is usually defined as the coefficient of variation of the actual coating thickness data set.

[0034] The formula is as follows: ; Formula parameter description: : Deviation index of coating uniformity of the whole plate, dimensionless; the smaller the value, the closer the thickness values ​​of all measurement points are, and the better the coating uniformity; : sample standard deviation of the actual coating thickness data set; : the arithmetic mean of the actual coating thickness data set; It has the following beneficial effects: Closed-loop feedback: provides key feedback information for model optimization; Model error map Point out the direction and location where the model needs to be improved; Quality Quantification: Uniformity Deviation Index It provides an objective and quantifiable quality assessment standard for the production process, facilitating quality control and process capability analysis; In this embodiment, the model parameter adaptive correction module is used to calculate the correction amount of the disturbance intensity coefficient. Specifically, for each conductive hanger contact, the average model prediction error of the model error map within the corresponding influence area is extracted, and the average error is multiplied by a preset learning rate parameter to obtain the correction amount of the disturbance intensity coefficient corresponding to the contact. The correction module for calculating the disturbance intensity coefficient is based on the model error map and the impact area divided in the first part , calculate the disturbance intensity coefficient for each conductive hanger contact Correction amount ; Specifically, for the kth contact, the module calculates the relative correction of its disturbance intensity coefficient; the module first extracts its affected area The average of all model errors in the model, which represents the relative deviation of the model prediction, is used to calculate the average error and a preset learning rate. , and the current disturbance intensity coefficient of the contact Multiply the three together to get the final correction amount The correction logic is: the size of the correction is not only proportional to the model error, but also proportional to the value of the parameter itself. This ensures the stability of the parameter adjustment and the consistency of the physical meaning. When the model prediction error is positive, that is, the actual coating is thicker than predicted, it means that the model overestimates the disturbance, and the correction is negative, thereby reducing the disturbance intensity coefficient. The formula is as follows: ; Formula parameter description: : disturbance intensity coefficient of the kth contact The correction amount, its dimension is the same as consistent; : learning rate, a dimensionless hyperparameter between 0 and 1; : The influence area of ​​the kth contact The average value of the model error within , dimensionless; : The old disturbance intensity coefficient of the kth contact before this correction; Update the disturbance intensity coefficient by adding the calculated correction amount to the old coefficient to obtain a new disturbance intensity coefficient for the production of the next batch of PCBs; The update equation is: ; The actual meaning and function of the formula: The update process is a typical iterative optimization process of feedback control and machine learning; by introducing the old coefficient value , correction amount and This makes parameter adjustment smoother and more stable; when the model overestimates the disturbance (the error is positive), it is reduced proportionally ; When the model underestimates the disturbance (the error is negative), increase it proportionally This allows the model parameters to converge efficiently and stably towards a direction that more accurately reflects physical reality. The hierarchical correction strategy makes the system more intelligent, able to distinguish between two different types of problems: model optimization and system failure, and make appropriate responses, which not only ensures the stability of the model but also provides valuable equipment diagnostic functions.

[0035] In this embodiment, the model parameter adaptive correction module further includes a hierarchical correction strategy before performing the update, for comparing the whole plate coating uniformity deviation index with a preset low deviation threshold and a preset high deviation threshold; If the uniformity deviation index is lower than the low deviation threshold, the current model is deemed accurate and no update is performed; if the uniformity deviation index is between the low deviation threshold and the high deviation threshold, the update is performed; In this embodiment, the hierarchical correction strategy is further configured to: if the whole-plate coating uniformity deviation index is higher than the high deviation threshold, then, while executing the update, the system will also trigger an alarm signal to indicate that there may be an abnormality in the equipment hardware or the chemical state of the electroplating solution that requires manual intervention; In order to increase the robustness and practicality of the system, the module will first make a classification judgment before executing the parameter update; it will calculate the uniformity deviation index of the whole board calculated by the post-plating quality assessment module. With two preset thresholds - low deviation threshold and high deviation threshold ——Make comparison.

[0036] Case 1: (Slight deviation) At this point, the system determines that the current electroplating quality is already very high and the model is sufficiently accurate. To avoid unnecessary fluctuations in model parameters (overfitting) due to random factors such as measurement noise, the system does not perform parameter updates. Case 2: (Moderate Deviation) At this point, the system determines that there is room for improvement in electroplating quality and that the deviation is within the model's correctable range. The system executes the aforementioned parameter update process to adaptively optimize the model. This is a normal learning and iterative process. Case 3: (Severe Deviation) In this case, the system determines that the uniformity deviation is extremely large. This is most likely not caused by inaccurate model predictions, but rather by a systemic hardware failure (such as severe wear or poor contact on a hanger contact) or an abnormal chemical state (such as the plating solution composition significantly deviating from the process window). In this case, the system will trigger an alarm signal while performing the parameter update, prompting the operator to perform manual inspection and intervention to eliminate potential equipment or chemical problems. The system is equipped with learning capabilities, enabling it to automatically adapt to slow changes caused by factors such as hanger wear and plating solution aging, and maintain high-precision control for a long time.

[0037] It should be noted that all calculation formulas in this application document use, including but not limited to, regression analysis in machine learning algorithms to deeply analyze the collected relevant parameters and identify their natural trends and interrelationships; use professional software, such as Python's Scikit-learn library or R language, to automatically generate mathematical models that match the data; then, objectively evaluate the model performance through methods such as cross-validation, and combine continuous feedback and optimization to ensure that the created formulas truly reflect the inherent laws of the data, thereby ensuring their effectiveness and accuracy, and ensuring that the calculation process complies with the constraints of natural laws rather than being based on artificially set rules.

[0038] The technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory, random access memory, flash memory, hard disk or optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0039] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other system that can fetch instructions from and execute instructions on an instruction execution system, apparatus, or device), or in conjunction with such instruction execution systems, apparatuses, or devices; for the purposes of this specification, a computer-readable medium can be any device that can contain, store, communicate, propagate, or transport a program for use by an instruction execution system, apparatus, or device, or in conjunction with such instruction execution systems, apparatuses, or devices.

[0040] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An intelligent industrial control system for the HDI board electroplating and hole filling process, applied to an electroplating equipment comprising a PCB board as a main cathode, a conductive hanger and its contacts, and auxiliary shielding electrodes arranged around the contacts, characterized in that: The system comprises: Initial electric field non-uniformity modeling module: Based on the preset electric field model, and according to the geometric layout parameters of the conductive rack contacts and the physical property parameters of the electroplating process, an initial electric field non-uniformity map is generated; Dynamic electric field compensation control module: used to calculate and output compensation control signals for each auxiliary shielding electrode in real time during the electroplating process based on the degree of deviation of each region indicated by the initial electric field non-uniformity map, so as to drive the auxiliary shielding electrode to generate a local compensation electric field to offset the electric field disturbance caused by the conductive hanger contact; A post-plating quality assessment module is configured to collect the actual coating thickness at multiple preset sampling points on the PCB board after the electroplating process is completed to form an actual coating thickness dataset; and compare the actual coating thickness dataset with the target coating thickness and the initial electric field non-uniformity map to generate a model error map, which is used to quantify the prediction deviation of the initial electric field non-uniformity map; The model parameter adaptive correction module is used to calculate one or more correction values ​​of the disturbance intensity coefficients related to the conductive hanger contacts based on the model error map; and use the correction values ​​to update the disturbance intensity coefficients used by the initial electric field non-uniformity modeling module.

2. The intelligent industrial control system for HDI board electroplating and hole filling process according to claim 1 is characterized in that: The initial electric field nonuniformity modeling module is used to generate the initial electric field nonuniformity map, specifically including: dividing the surface of the PCB board into multiple independent influence areas corresponding to each conductive hanger contact based on the geometric layout parameters of the conductive hanger contacts; and within each of the influence areas, calculating the electric field disturbance intensity generated by the corresponding conductive hanger contact based on a preset distance attenuation model, and superimposing the disturbance intensities generated by all contacts to obtain the expected deviation degree at any position on the PCB board surface to construct the initial electric field nonuniformity map.

3. The intelligent industrial control system for HDI board electroplating and hole filling process according to claim 2, characterized in that: In the distance attenuation model, the disturbance intensity generated by any conductive hanger contact at any point on the PCB surface depends on the disturbance intensity coefficient associated with the contact itself and the geometric distance between the arbitrary point and the contact.

4. The intelligent industrial control system for HDI board electroplating and hole filling process according to claim 1, characterized in that: The dynamic electric field compensation control module is used to calculate the compensation control signal, and the specific steps are: S1: for each auxiliary shielding electrode, extracting the average deviation degree of its coverage area from the initial electric field non-uniformity map; S2: Multiply the average deviation degree with a compensation gain coefficient preset for the electrode and the real-time electrical parameters of the main circuit of the electroplating equipment to calculate the compensation voltage value required by the auxiliary shielding electrode at that moment, and output the compensation voltage value as the compensation control signal.

5. The intelligent industrial control system for HDI board electroplating and hole filling process according to claim 1 is characterized in that: The post-plating quality assessment module is used to generate the model error map, specifically: for each preset sampling point, the normalized deviation of the actual coating thickness collected is calculated relative to the target coating thickness; the normalized deviation is compared with the deviation degree predicted by the initial electric field non-uniformity map at that point, and the difference is defined as the model prediction error at that point. The model prediction errors of all sampling points together constitute the model error map.

6. The intelligent industrial control system for HDI board electroplating and hole filling process according to claim 5, characterized in that: The post-plating quality assessment module is also used to: calculate the ratio of the sample standard deviation of all thickness values ​​in the data set to the arithmetic mean based on the actual coating thickness data set, and define the ratio as a whole-plate coating uniformity deviation index for quantitatively evaluating the overall quality of this electroplating.

7. The intelligent industrial control system for HDI board electroplating and hole filling process according to claim 5, characterized in that: The model parameter adaptive correction module is used to calculate the correction amount of the disturbance intensity coefficient. Specifically, for each conductive hanger contact, the average model prediction error of the model error map in its corresponding influence area is extracted, and the average error is multiplied by a preset learning rate parameter to obtain the correction amount of the disturbance intensity coefficient corresponding to the contact.

8. The intelligent industrial control system for HDI board electroplating and hole filling process according to claim 7, characterized in that: The model parameter adaptive correction module further includes a hierarchical correction strategy before executing the update, for comparing the whole plate coating uniformity deviation index with a preset low deviation threshold and a preset high deviation threshold; If the uniformity deviation index is lower than the low deviation threshold, the current model is deemed accurate and no update is performed; if the uniformity deviation index is between the low deviation threshold and the high deviation threshold, the update is performed.

9. The intelligent industrial control system for HDI board electroplating and hole filling process according to claim 8, characterized in that: The graded correction strategy is also used: if the whole plate coating uniformity deviation index is higher than the high deviation threshold, then while executing the update, the system will also trigger an alarm signal to indicate that there may be abnormalities in the equipment hardware or the chemical state of the electroplating solution that require manual intervention.

Images