An intelligent industrial control system for a hole-filling electroplating process of an HDI board

Abstract

The application discloses an intelligent industrial control system for a via filling electroplating process of an HDI board, belongs to the field of intelligent manufacturing of circuit boards, and is applied to an electroplating device comprising a PCB board as a main cathode, a conductive hanger and a contact thereof, and an auxiliary shielding electrode arranged around 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 self-adaptive correction module. The application introduces the post-plating quality evaluation and model parameter self-adaptive correction module, and constructs a complete prediction-control-evaluation-correction closed loop system. This enables the system to learn from the production result each time, automatically correct and optimize the internal model, and adapt to the slow drift of the process conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent manufacturing of circuit boards, in particular to an intelligent industrial control system for the plating and via filling process of HDI boards. BACKGROUND

[0002] In the manufacturing process of high-density interconnection (HDI) printed circuit boards, plating and via filling is a key process to ensure reliable electrical interconnection between conductive layers. In traditional plating processes, PCB boards are usually fixed by metal conductive hangers and immersed in the plating solution as the entire cathode. However, the contacts on the hanger used to hold the PCB board and conduct electricity themselves become part of the electric field, causing significant shielding and distortion effects on the surrounding electric field lines.

[0003] This physical effect results in a significant decrease in plating current density at the edge region of the PCB board near the hanger contacts, compared to the center of the board, ultimately causing the copper layer after plating to exhibit uneven distribution with thin edges and thick center. This unevenness severely affects the final yield rate and high-frequency electrical performance of HDI boards.

[0004] To solve this problem, existing technologies usually use passive, experience-dependent compensation methods such as adding physical baffles, adjusting the position or shape of the anode. These methods have limited effect and cannot accurately address electric field distortion, and cannot adapt to differences brought by different batches and different designs of PCB boards, lacking intelligence, accuracy and adaptability. Therefore, there is an urgent need for a technical solution that can actively, accurately and intelligently solve the problem of plating unevenness caused by hanger contacts. SUMMARY

[0005] The present application aims to overcome the above-mentioned defects of the prior art and provide an intelligent industrial control system for the plating and via filling process of HDI boards, which aims to solve the problem of shielding and disturbance of the electric field at the edge of the PCB board caused by the plating hanger contacts, to achieve uniformization of the plating layer thickness on the entire PCB board surface.

[0006] The technical solution of the present application is an intelligent industrial control system for the plating and via filling process of HDI boards, applied to a plating equipment comprising a PCB board as the main cathode, a conductive hanger and its contacts, and an auxiliary shielding electrode arranged around the contacts, the system comprising:

[0007] An initial electric field non-uniformity modeling module: based on a preset electric field model, and according to the geometric layout parameters of the conductive hanger contacts and the physical attribute parameters of the plating process, an initial electric field non-uniformity map is generated;

[0008] The dynamic electric field compensation control module is configured to calculate and output a compensation control signal for each auxiliary shielding electrode in real time during the electroplating process according to the deviation degree 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 for offsetting the electric field disturbance caused by the conductive fixture contact point.

[0009] The post-plating quality evaluation module is configured to collect actual plating layer thicknesses of a plurality of preset sampling points on the PCB after the electroplating process is completed, to form an actual plating layer thickness data set; and compare the actual plating layer thickness data set with the target plating layer 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.

[0010] The model parameter self-adaptive correction module is configured to calculate a correction amount of one or more disturbance intensity coefficients related to the conductive fixture contact point based on the model error map, and update the disturbance intensity coefficients used by the initial electric field non-uniformity modeling module by using the correction amount.

[0011] In the embodiment, the initial electric field non-uniformity modeling module is configured to generate the initial electric field non-uniformity map, specifically including: dividing the surface of the PCB into a plurality of independent influence areas corresponding to each conductive fixture contact point according to the geometric layout parameters of the conductive fixture contact points; and in each of the influence areas, calculating the electric field disturbance intensity generated by the corresponding conductive fixture contact point based on a preset distance decay model, and superimposing the disturbance intensities generated by all the contact points, to obtain the expected deviation degree of any position on the surface of the PCB, so as to construct the initial electric field non-uniformity map.

[0012] In the embodiment, in the distance decay model, the disturbance intensity generated by any conductive fixture contact point at any point on the surface of the PCB depends on the disturbance intensity coefficient related to the contact point itself and the geometric distance between the point and the contact point.

[0013] In the embodiment, the dynamic electric field compensation control module is configured to calculate the compensation control signal, specifically including the following steps:

[0014] S1: for each auxiliary shielding electrode, extracting the average deviation degree of the covered area from the initial electric field non-uniformity map;

[0015] S2: multiplying the average deviation degree by a compensation gain coefficient preset for the electrode and real-time electric parameters of the main loop of the electroplating equipment, to calculate the compensation voltage value required by the auxiliary shielding electrode at the moment, and outputting the compensation voltage value as the compensation control signal.

[0016] In the embodiment, the post-plating quality evaluation module is configured to generate the model error map, specifically, for each preset sampling point, a normalized deviation of the actual plating thickness collected by the sampling point from the target plating thickness is calculated; the normalized deviation is compared with a deviation degree predicted by the initial electric field non-uniformity map at the point, and a difference value is defined as a model prediction error of the point, and model prediction errors of all the sampling points jointly constitute the model error map.

[0017] In the embodiment, the post-plating quality evaluation module is further configured to calculate a ratio of a sample standard deviation to an arithmetic mean of all thickness values in the actual plating thickness data set, and define the ratio as a whole-board plating uniformity deviation index, which is used for quantitatively evaluating the overall quality of the electroplating.

[0018] In the embodiment, the model parameter adaptive correction module is configured to calculate a correction amount of the perturbation intensity coefficient, specifically, for each conductive hanger contact, an average model prediction error of the model error map in a corresponding influence area of the contact is extracted, and the average error is multiplied by a preset learning rate parameter to obtain the correction amount of the perturbation intensity coefficient corresponding to the contact.

[0019] In the embodiment, before performing the updating, the model parameter adaptive correction module further comprises a hierarchical correction strategy configured to compare the whole-board plating uniformity deviation index with a preset low deviation threshold and a preset high deviation threshold.

[0020] If the uniformity deviation index is lower than the low deviation threshold, it is determined that the current model is accurate, and the updating is not performed; if the uniformity deviation index is between the low deviation threshold and the high deviation threshold, the updating is performed.

[0021] In the embodiment, the hierarchical correction strategy is further configured to, if the whole-board plating uniformity deviation index is higher than the high deviation threshold, trigger an alarm signal to prompt that there may be an abnormality of device hardware or electroplating liquid chemical state requiring manual intervention while performing the updating.

[0022] The present application provides an intelligent industrial control system for HDI board hole filling electroplating process, which has the following improvements and advantages compared with the prior art.

[0023] The present application introduces a post-plating quality evaluation module and a model parameter adaptive correction module, and constructs a complete prediction-control-evaluation-correction closed loop system. This enables the system to learn from the production results of each time, automatically correct and optimize its internal model to adapt to the slow drift of process conditions, such as hanger wear and chemical aging, which is a self-learning and self-adaptive ability that the prior art completely lacks.

[0024] The generated initial electric field non-uniformity map, model error map, and whole-board uniformity deviation index, etc. quantitative indicators provide process engineers with unprecedented process insight. In particular, the hierarchical correction strategy can intelligently distinguish whether the model needs fine-tuning or the equipment has a major fault, greatly improving the efficiency and accuracy of fault diagnosis.

[0025] The application adopts auxiliary shielding electrodes for active electric field compensation. Compared with the traditional passive method of physical baffle, the control accuracy and flexibility have achieved a qualitative leap. The strength and form of the compensation electric field can be dynamically and accurately adjusted as needed.

[0026] The application establishes a mathematical model of electric field non-uniformity, changes the compensation of electric field disturbance from relying on operator experience to accurate calculation based on the model. This makes the compensation process standardized, repeatable, and able to quickly adapt to different PCB types. BRIEF DESCRIPTION OF DRAWINGS

[0027] The application will be further explained in conjunction with the accompanying drawings and examples:

[0028] Figure 1 is a flow chart of an intelligent industrial control system for the electroplating via filling process of an HDI board. DETAILED DESCRIPTION

[0029] To make the purpose, technical solutions and advantages of the application clearer, the application will be further described in detail below with specific examples;

[0030] Example 1:

[0031] Please refer to Figure 1 The application provides an intelligent industrial control system for the electroplating via filling process of an HDI board technical solution: an intelligent industrial control system for the electroplating via filling process of an HDI board, applied to an electroplating equipment containing a PCB board as a main cathode, a conductive hanger and its contacts, and an auxiliary shielding electrode set around the contacts, the system comprises:

[0032] An initial electric field non-uniformity modeling module: based on a preset electric field model, and according to the geometric layout parameters of the conductive hanger contacts and the physical property parameters of the electroplating process, an initial electric field non-uniformity map is generated;

[0033] Based on a preset electric field model, and according to the geometric layout parameters of the conductive hanger contacts and the physical property parameters of the electroplating process, an initial electric field non-uniformity map is generated; the map is used to represent the deviation of the expected electroplating current density of each region on the PCB board surface from the target value under the condition of no compensation;

[0034] A dynamic electric field compensation control module is configured to calculate and output compensation control signals for each auxiliary shielding electrode in real time during the electroplating process according to the deviation degree 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 fixture contact point.

[0035] A post-plating quality evaluation module is configured to collect actual plating layer thicknesses of a plurality of preset sampling points on the PCB after the electroplating process is completed, to form an actual plating layer thickness data set, and to compare the actual plating layer thickness data set with the target plating layer 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.

[0036] A model parameter self-adaptive correction module is configured to calculate a correction amount of one or more disturbance intensity coefficients related to the conductive fixture contact point based on the model error map, and to update the disturbance intensity coefficients used by the initial electric field non-uniformity modeling module by using the correction amount.

[0037] In the embodiment, the initial electric field non-uniformity modeling module is configured to generate the initial electric field non-uniformity map, specifically including: dividing the surface of the PCB into a plurality of independent influence areas corresponding to each conductive fixture contact point according to the geometric layout parameters of the conductive fixture contact point; and in each of the influence areas, calculating the electric field disturbance intensity generated by the corresponding conductive fixture contact point based on a preset distance decay model, and superimposing the disturbance intensities generated by all the contact points, to obtain the expected deviation degree of any position on the surface of the PCB, so as to construct the initial electric field non-uniformity map.

[0038] Geometric layout parameters of the conductive fixture contact point: a set of position coordinates of all conductive contact points in a two-dimensional coordinate system ;

[0039] Physical property parameters of the electroplating process: for example, the conductivity of the electroplating solution , target plating layer thickness , etc.

[0040] Based on the set of position coordinates of the contact points , the module divides the entire surface of the PCB into a plurality of independent and non-overlapping influence areas using the Voronoi diagram algorithm Each area contains all points closer to the kth contact point than to any other contact point. This division method geometrically defines the main influence range of each contact point, providing an accurate spatial basis for subsequent parameter correction.

[0041] In the present embodiment, the disturbance intensity generated by any contact point of the conductive hanger at any point on the surface of the PCB is determined by a disturbance intensity coefficient associated with the contact point itself and the geometric distance between the point and the contact point;

[0042] The non-uniformity map calculates, in each influence area, the electric field disturbance intensity generated by the corresponding contact point based on a preset distance decay model, the core idea of which is that the influence intensity of the contact point on the electric field decays with the increase of the distance; the disturbance intensity generated by any contact point of the conductive hanger at a point on the surface of the PCB is determined by a disturbance intensity coefficient associated with the contact point itself and the geometric distance between the point and the contact point.

[0043] In order to obtain the total expected deviation degree of any point on the surface of the PCB , the module linearly superimposes the disturbance intensities generated by all the contact points at the point, and the calculation formula is as follows:

[0044] ;

[0045] Formula parameter description:

[0046] : represents the initial electric field non-uniformity index at the coordinate point on the PCB; this is a dimensionless quantity, and the value represents the deviation degree of the expected electroplating current density of the point relative to the target value; the greater the value, the more serious the expected deviation, and generally the current density is too low; : total number of conductive hanger contact points;

[0047] : two-dimensional coordinate vector of the kth contact point;

[0048] : disturbance intensity coefficient of the kth contact point, which is the key parameter of the model, and its physical meaning is to represent the disturbance ability of a single contact point to generate an electric field, which is related to the size, material, shape of the contact point and the conductivity of the electroplating solution , etc. The 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, and its dimension can be considered as , so that is dimensionless;

[0049] : represents the Euclidean distance between the point and the kth contact point ;

[0050] The distance decay exponent is a positive real number, usually taken as 1 or 2 based on electric field theory and experimental data; it determines the rate at which the disturbance intensity decays with distance.

[0051] A very small positive constant whose purpose is to prevent errors when calculating points. When the position coincides with the contact point, the denominator is zero, ensuring the numerical stability of the calculation.

[0052] The practical significance and function of the formula: This formula constructs a predictive model with clear physical meaning; it quantifies a core physical intuition of electric field disturbance: the influence of the disturbance source (contact point) decays with distance; by superimposing the effects of all disturbance sources, a complete initial electric field non-uniformity spectrum covering the entire PCB board can be generated. This map serves as the direct basis for subsequent precise compensation.

[0053] Quantitative prediction transforms empirical and fuzzy edge effect problems into a computable and quantifiable physical model, providing a scientific basis for precise control.

[0054] The graph allows for precise identification of the location and required level of compensation on the PCB board, representing a leap from global compensation to precise local compensation.

[0055] In this embodiment, the dynamic electric field compensation control module is used to calculate the compensation control signal, and the specific steps are as follows:

[0056] S1: For each auxiliary shielding electrode, extract the average deviation of its coverage area from the initial electric field non-uniformity spectrum;

[0057] S2: Multiply the average deviation by 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.

[0058] S1 extracts the average deviation for each auxiliary shielding electrode (assuming there are m electrodes, with indices j=1,2,...,m). This module first extracts the average deviation from the initial electric field non-uniformity map. Extract the area it covers. Average deviation within The calculation method is based on the area... The value is integrated or averaged; this step simplifies the complex two-dimensional graph information into a scalar control objective for each independent control unit.

[0059] Step 2: The compensation voltage module calculates the compensation voltage value required by each auxiliary shielding electrode at time t according to the following formula:

[0060] ;

[0061] Formula parameter explanation:

[0062] : The compensation voltage value required by the jth auxiliary shielding electrode at time t, in volts (V), which is the final output control signal of the module;

[0063] : Compensation gain coefficient of the jth auxiliary shielding electrode; This is a dimensionless preset parameter obtained through experimental calibration, which represents the compensation efficiency of the electrode, that is, how much compensation electric field can be generated by applying a unit voltage. This coefficient is related to the geometric shape, size, position of the electrode, and its relative relationship with the main cathode;

[0064] : Average non-uniformity index of the jth electrode coverage area extracted from the atlas, dimensionless;

[0065] : 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 strength of the electric field disturbance is proportional to the strength 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;

[0066] This formula combines the predicted, static non-uniformity index with the real-time, dynamic process parameters to generate a dynamic and accurate compensation control instruction ; It ensures that the strength of the compensation electric field can follow the fluctuations of the main electroplating process in real time, achieving dynamic matching.

[0067] The calculated compensation voltage signal is sent to the independent programmable power supply connected to each auxiliary shielding electrode . The power supply drives the corresponding electrode according to the signal, making it generate a local, accurately calculated compensation electric field in terms of direction and strength. The purpose of this compensation electric field is to locally correct the electric field distribution and offset the shielding effect caused by the conductive contact, effectively restoring the electric field strength at the edge of the PCB, thereby achieving uniformization of the current density on the entire board.

[0068] In the embodiment, the post-plating quality evaluation module is configured to generate the model error map, specifically, for each preset sampling point, a normalized deviation of the actual plating thickness collected by the sampling point from the target plating thickness is calculated; the normalized deviation is compared with the deviation degree predicted by the initial electric field non-uniformity map at the point, and a difference value is defined as a model prediction error of the point, and model prediction errors of all sampling points jointly constitute the model error map.

[0069] Step 1: Collecting actual plating thickness data .

[0070] Step 2: Generating a model error map module ; the calculation steps are as follows: a. calculating the normalized deviation of the actual plating thickness of the point: comparing the actual thickness with the target thickness , and normalizing; b. comparing the actual deviation with the predicted deviation degree of the initial model at the point (retrieved from the initial map); the difference between the two is the prediction error of the model at the point;

[0071] The formula is as follows:

[0072] ;

[0073] Formula parameter description:

[0074] : the model prediction error of the i-th sampling point, dimensionless; the positive and negative signs have clear physical meanings: if positive, it means that the actual plating thickness is thicker than the model prediction, indicating that the model prediction of the disturbance intensity of the point is too high; if negative, it is the opposite;

[0075] : the actual measured thickness of the i-th sampling point;

[0076] : the target plating thickness set by the process;

[0077] : the predicted deviation index of the initial electric field non-uniformity model at the i-th sampling point;

[0078] Interpolating all model prediction errors of the sampling points , a visual model error map can be generated , which intuitively shows which areas the current model overestimates and which areas it underestimates;

[0079] In the embodiment, the post-plating quality evaluation module is further configured to calculate a ratio of a sample standard deviation to an arithmetic mean of all thickness values in the actual plating thickness data set, and define the ratio as a whole-plate plating uniformity deviation index for quantitatively evaluating the overall quality of the current electroplating;

[0080] The whole-plate uniformity deviation index is calculated In order to give a macroscopic and single evaluation index of the overall quality of the current electroplating, the module is further configured to calculate a whole-plate plating uniformity deviation index.

[0081] The formula is as follows:

[0082] ;

[0083] Formula parameter description:

[0084] : the whole-plate plating uniformity deviation index, dimensionless; the smaller the value is, the closer the thickness values of all measurement points are, and the better the plating uniformity is;

[0085] : the sample standard deviation of the actual plating thickness data set;

[0086] : the arithmetic mean of the actual plating thickness data set;

[0087] The following beneficial effects are achieved: closed-loop feedback: key feedback information for model optimization is provided; model error map points out the direction and position that the model needs to be improved;

[0088] quality quantification: uniformity deviation index provides an objective and quantifiable quality evaluation standard for the production process, facilitating quality control and process capability analysis;

[0089] In the embodiment, the model parameter adaptive correction module is configured to calculate a correction amount of the perturbation intensity coefficient, specifically: for each conductive hanger contact, the average model prediction error of the model error map in the corresponding influence area thereof is extracted, and the average error is multiplied by a preset learning rate parameter to obtain the correction amount of the perturbation intensity coefficient corresponding to the contact.

[0090] The correction amount of the perturbation intensity coefficient is calculated based on the model error map and the influence area divided in the first part Calculate the disturbance intensity coefficient for each conductive hanger contact. Correction amount ;

[0091] Specifically, for the k-th contact point, the module calculates the relative correction amount of its disturbance intensity coefficient; the module first extracts its affected area. The average error of all models is used to characterize the relative bias of the model's predictions. This average error is then set to a preset learning rate. and the current disturbance intensity coefficient of the contact point. Multiplying the three together yields the final correction amount. The correction logic is as follows: the magnitude of the correction is proportional not only to the model error but also to the value of the parameter itself. This ensures the stability of parameter adjustment and the consistency of physical meaning. When the model prediction error is positive, meaning the actual coating is thicker than predicted, it indicates that the model has overestimated the disturbance, and the correction is negative, thus reducing the disturbance intensity coefficient. The formula is as follows:

[0092] ;

[0093] Formula parameter description:

[0094] Disturbance intensity coefficient at the k-th contact point The correction amount, whose dimensions are the same as Consistent;

[0095] Learning rate, a dimensionless hyperparameter between 0 and 1;

[0096] The area of ​​influence at the k-th contact point The average model error within the time frame is dimensionless.

[0097] The old disturbance intensity coefficient of the k-th contact before this correction;

[0098] The updated disturbance strength coefficient is obtained by adding the calculated correction amount to the old coefficient, and then using the new disturbance strength coefficient for the production of the next batch of PCBs.

[0099] The update equation is:

[0100] ;

[0101] The practical significance and function of the formula: The update process is a typical iterative optimization process of feedback control and machine learning; by introducing old coefficient values... Correction amount and the size is proportional; this makes the adjustment of parameters more smooth and stable; when the model overestimates the disturbance (positive error), it is reduced in proportion ; when the model underestimates the disturbance (negative error), it is increased in proportion ; this enables the model parameters to converge efficiently and stably towards a direction that can more accurately reflect the physical reality;

[0102] The hierarchical correction strategy makes the system more intelligent, capable of distinguishing between two different nature problems of model optimization and system failure, and making appropriate response, both ensuring the stability of the model and providing valuable equipment diagnosis function.

[0103] In the embodiment, the model parameter adaptive correction module further comprises a hierarchical correction strategy before performing the updating, for comparing the whole-plate plating layer uniformity deviation index with a preset low deviation threshold and a preset high deviation threshold;

[0104] If the uniformity deviation index is lower than the low deviation threshold, it is determined that the current model is accurate, and the updating is not performed; if the uniformity deviation index is between the low deviation threshold and the high deviation threshold, the updating is performed;

[0105] In the embodiment, the hierarchical correction strategy is further used for: if the whole-plate plating layer uniformity deviation index is higher than the high deviation threshold, the system will trigger an alarm signal to prompt that there may be abnormality of the device hardware or electroplating liquid chemical state requiring manual intervention while performing the updating;

[0106] In order to increase the robustness and practicability of the system, the module will first perform a hierarchical judgment before performing the parameter updating; it compares the whole-plate uniformity deviation index calculated by the post-plating quality evaluation module with two preset thresholds, a low deviation threshold and a high deviation threshold .

[0107] Case one: (mild deviation) At this time, the system determines that the current electroplating quality is already very high, and the model is accurate enough; in order to avoid unnecessary fluctuations of the model parameters (overfitting) caused by random factors such as measurement noise, the system does not perform parameter updating;

[0108] Case two: (moderate deviation) At this time, the system determines that the electroplating quality has room for improvement, and the deviation is within the correctable range of the model; the system performs the above parameter updating process to adaptively optimize the model; this is a normal learning and iteration process;

[0109] Case three: (Serious deviation) At this time, the system determines that the uniformity deviation is extremely large, which is likely not caused by inaccurate model prediction, but by a systematic hardware failure (such as severe wear or poor contact of a certain hanger contact) or a chemical state anomaly (such as a serious deviation of the electroplating solution composition from the process window); in this case, the system triggers an alarm signal to prompt the operator to perform manual inspection and intervention to eliminate potential equipment or chemical problems while performing parameter updates;

[0110] The system has learning ability and can automatically adapt to slow changes caused by factors such as hanger wear and electroplating solution aging, thereby maintaining high-precision control for a long time.

[0111] It should be noted that all calculation formulas in the present application file use regression analysis including but not limited to machine learning algorithms to deeply analyze the collected relevant parameters, identify their natural trends and mutual relationships; professional software such as Python's Scikit-learn library or R language is used to automatically generate mathematical models matching the data; then, the model performance is objectively evaluated through cross-validation and other methods, and combined with continuous feedback and optimization to ensure that the created formula truly reflects the inherent law of the data, thereby ensuring its effectiveness and accuracy, and ensuring that the calculation process conforms to the constraints of natural laws, rather than being based on artificially set rules.

[0112] The technical solutions of the present application can be embodied in the form of a software product, which 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., including a number of instructions to make a computer device (which can be a personal computer, server, or network device, etc.) execute the method of various embodiments of the present application.

[0113] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered a list of executable instructions for implementing logic functions, which can be specifically embodied in any computer-readable medium for use by an instruction execution system, device or apparatus, such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions, or in conjunction with these instruction execution systems, devices or apparatus; for the purpose of this specification, the computer-readable medium can be any device that can contain, store, communicate, propagate or transport programs for use by an instruction execution system, device or apparatus or in conjunction with these instruction execution systems, devices or apparatus.

[0114] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. An intelligent industrial control system for an HDI board electroplating via filling process, applied to an electroplating equipment comprising a PCB board as a main cathode, a conductive hanger and its contacts, and an auxiliary shielding electrode arranged around the contacts, characterized in that, The system comprises: An initial electric field non-uniformity modeling module: based on a preset electric field model, and according to the geometric layout parameters of the conductive fixture contacts and the physical attribute parameters of the electroplating process, an initial electric field non-uniformity atlas is generated; A dynamic electric field compensation control module: for calculating and outputting compensation control signals for each auxiliary shielding electrode in real time during the electroplating process, according to the deviation degree of each region indicated by the initial electric field non-uniformity atlas, to drive the auxiliary shielding electrode to generate a local compensation electric field to offset the electric field disturbance caused by the conductive fixture contacts; A post-plating quality evaluation module: for collecting the actual plating layer thickness of a plurality of preset sampling points on the PCB after the electroplating process is completed, forming an actual plating layer thickness dataset; and comparing the actual plating layer thickness dataset with the target plating layer thickness and the initial electric field non-uniformity atlas to generate a model error atlas, which is used to quantify the prediction deviation of the initial electric field non-uniformity atlas; A model parameter self-adaptive correction module: for calculating a correction amount of one or more disturbance intensity coefficients related to the conductive fixture contacts based on the model error atlas; and updating the disturbance intensity coefficients used by the initial electric field non-uniformity modeling module using the correction amount.

2. The intelligent industrial control system for HDI board via-filling plating process according to claim 1, wherein, The initial electric field non-uniformity modeling module is used to generate the initial electric field non-uniformity atlas, specifically comprising: dividing the surface of the PCB into a plurality of independent influence areas corresponding to each conductive fixture contact according to the geometric layout parameters of the conductive fixture contacts; and in each of the influence areas, calculating the electric field disturbance intensity generated by the corresponding conductive fixture contact based on a preset distance decay model, and superimposing the disturbance intensities generated by all contacts to obtain the expected deviation degree of any position on the surface of the PCB, thereby constructing the initial electric field non-uniformity atlas.

3. The intelligent industrial control system for HDI board via plating process according to claim 2, characterized in that, In the distance decay model, the disturbance intensity generated by any conductive fixture contact at any point on the surface of the PCB depends on the disturbance intensity coefficient related to the contact itself and the geometric distance between the point and the contact.

4. The intelligent industrial control system for HDI board electroplating via fill process of claim 1, wherein, The dynamic electric field compensation control module is used to calculate the compensation control signal, specifically comprising the following steps: S1: for each auxiliary shielding electrode, extracting the average deviation degree of its covered area from the initial electric field non-uniformity atlas; S2: multiplying the average deviation degree by a preset compensation gain coefficient for the electrode and the real-time electric parameters of the main loop of the electroplating equipment, thereby calculating the real-time required compensation voltage value of the auxiliary shielding electrode, and outputting the compensation voltage value as the compensation control signal.

5. The intelligent industrial control system for HDI board electroplating via fill process of claim 1, wherein, The post-plating quality evaluation module is used to generate the model error atlas, specifically comprising: for each preset sampling point, calculating the normalized deviation of the actual plating layer thickness collected relative to the target plating layer thickness; comparing the normalized deviation with the deviation degree predicted by the initial electric field non-uniformity atlas at the point, and the difference is defined as the model prediction error of the point, and the model prediction errors of all sampling points together constitute the model error atlas.

6. The intelligent industrial control system for HDI board via plating process according to claim 5, wherein, The post-plating quality evaluation module is further configured to calculate a ratio of a sample standard deviation to an arithmetic mean of all thickness values in the actual plating thickness dataset, and define the ratio as a whole-board plating uniformity deviation index for quantitatively evaluating the overall quality of the current electroplating.

7. The intelligent industrial control system for HDI board via plating process according to claim 5, wherein, The model parameter adaptive correction module is configured to calculate a correction amount of the perturbation intensity coefficient, specifically, for each conductive hanger contact, extract an average model prediction error of the model error map in the corresponding influence area thereof, and multiply the average model prediction error by a preset learning rate parameter to obtain the correction amount of the perturbation intensity coefficient corresponding to the contact.

8. The intelligent industrial control system for HDI board via plating process according to claim 7, wherein, The model parameter adaptive correction module further comprises a hierarchical correction strategy before performing the updating, which is configured to compare the whole-board plating 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, it is determined that the current model is accurate, and the updating is not performed; if the uniformity deviation index is between the low deviation threshold and the high deviation threshold, the updating is performed.

9. The intelligent industrial control system for HDI board via plating process according to claim 8, wherein, The hierarchical correction strategy is further configured to, if the whole-board plating uniformity deviation index is higher than the high deviation threshold, trigger an alarm signal to prompt that there may be an abnormality in the device hardware or electroplating liquid chemical state requiring manual intervention while performing the updating.

Images