Multi-parameter adaptive based PCB drug solution concentration closed-loop regulation method and system

Abstract

The present application relates to the technical field of printed circuit board (PCB) manufacturing process, in particular to a PCB chemical concentration closed-loop regulation method and system based on multi-parameter self-adaptation. Different from the prior art, the present application introduces fine parameters such as total pore surface area and substrate properties into the model, and realizes accurate addition amount calculation in multi-specification mixed production scenarios from the source through model parameters pre-trained by historical data. A hierarchical self-adaptive mechanism of "feedforward prediction + immediate feedback + global correction coefficient adjustment" is adopted, wherein the feedforward prediction realizes on-demand supply to reduce lag, the immediate feedback quickly eliminates the current concentration deviation, and the global correction coefficient adjustment adjusts the basic addition amount based on the relative error between the actual demand and the prediction value. Without changing the internal parameters of the mature model, the system deviation caused by factors such as chemical aging and equipment drift is effectively compensated, and the system robustness and precision are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of printed circuit board (PCB) manufacturing technology, and in particular to a closed-loop control method and system for PCB chemical concentration based on multi-parameter adaptive methods. Background Technology

[0002] A PCB (Printed Circuit Board) is the support structure for electronic components, providing electrical connections for them. PCB manufacturing typically involves wet processes such as chemical copper plating, wet electroplating, and wet developing, all of which rely on specific concentrations of chemicals. The concentration of these chemicals directly affects PCB quality: excessively high concentrations can lead to over-etching or uneven deposition, while insufficient concentrations may result in incomplete processing and compromised quality. Therefore, precise control of chemical concentrations is a critical aspect of PCB manufacturing.

[0003] Currently, there are two main methods for controlling the concentration of chemicals. The first is automatic addition at a fixed frequency: chemicals are added at preset production parameters on a timed and quantitative basis. Simultaneously, manual sampling and testing are conducted periodically, and chemicals are replenished and the addition amount optimized based on the test results. This method is inefficient, has a long feedback cycle, and suffers from significant lag in analysis and adjustment. Once the concentration deviates, it often leads to batch defects. The second method is automatic addition based on online analysis: the concentration is calculated by monitoring changes in absorbance or current of the chemicals, and addition is automatic when it falls below a threshold. However, the accuracy of the correlation between absorbance and current changes and concentration is limited, and it only applies to certain chemicals, lacking universality. Even with titration detection, the long detection time results in significant concentration fluctuations within the detection interval, making it difficult to guarantee product consistency.

[0004] More importantly, actual production lines often need to produce multiple PCB models, and the chemical consumption varies significantly between these models due to differences in thickness, layer structure, and other factors. Neither of the two methods mentioned above considers this difference: the fixed-frequency automatic addition scheme becomes ineffective after changing the PCB model, requiring re-optimization; while the online analysis scheme can respond in real time, its limitations in detection accuracy and versatility are further amplified when switching between multiple models. Each change of PCB model means re-exploring control parameters, exacerbating the adverse effects of adjustment lag on product quality. Therefore, resolving the concentration fluctuation problem caused by the mismatch between the chemical addition amount and actual consumption in mixed production scenarios is crucial for improving product yield and quality stability. Summary of the Invention

[0005] Based on this, the purpose of this invention is to overcome the defects or deficiencies of the prior art and provide a closed-loop control method for PCB chemical concentration based on multi-parameter adaptive methods, comprising: Step S10: Obtain the product specification parameters of the current model of PCB board to be produced. The product specification parameters include at least the board area and micro-hole parameters of a single PCB board. The micro-hole parameters are the total surface area of ​​the micro-holes, or the micro-hole parameters include the number, diameter and depth of various hole types. Step S20: Obtain the correction coefficient of the PCB board, calculate the first amount of chemical to be added for producing a preset unit of PCB board by combining the prediction model and product specification parameters, and then execute step S30. Step S30: At the start of the current control cycle, obtain the actual concentration of the reaction tank solution at the first moment as the first concentration; for each preset unit of PCB board produced, add the first amount of chemical solution to the reaction tank until the end of the current control cycle, obtain the actual concentration of the solution at the second moment as the second concentration, and then execute step S40. Step S40: Calculate the first consumption amount corresponding to the preset unit of PCB board based on the first concentration and the second concentration, update the correction coefficient and the first addition amount of the PCB board based on the first consumption amount; calculate the chemical adjustment amount of the bath solution based on the target concentration and the second concentration, adjust the concentration of the bath solution in the reaction tank according to the chemical adjustment amount, and then execute step S30.

[0006] Compared to existing technologies, the dynamic control method of this invention incorporates refined parameters such as the total surface area of ​​micropores into the model. By pre-training model parameters using historical data, it achieves precise dosage calculation for multi-specification mixed production scenarios from the source. Whenever a different type of PCB board is used, this invention can quickly adjust the dosage to be close to the actual consumption value, thereby controlling concentration fluctuations within an extremely narrow range. This effectively avoids processing defects caused by improper chemical concentration, significantly reduces the risk of batch quality accidents, and allows the production line to automatically adapt to production switching of different PCB board types and materials, improving the flexibility of the production line. Furthermore, in embodiment S40, real-time concentration prediction is added to compensate for the inherent detection lag problem of current analytical techniques (such as potentiometric titration).

[0007] In one embodiment, in step S20, the prediction model is a machine learning model pre-trained based on historical production data, and the machine learning model is selected from any one of linear model, logistic regression model, linear regression model, multinomial regression model, neural network model, support vector machine model, and fuzzy logic model.

[0008] In one embodiment, step S20, obtaining the correction coefficient of the PCB board includes: comparing the product specification parameters of the current model PCB board with the product specification parameters of existing PCB boards in the database to determine whether the PCB board is a new model PCB board. If so, the product specification parameters of the PCB board are stored, and the initial parameters are used as the correction parameters of the PCB board; if not, the correction coefficient of the PCB board is retrieved from the database.

[0009] In one embodiment, the product specification parameters also include the reactivity coefficient between the PCB board material and the chemical solution, and the calculation method for the first addition amount in step S20 includes: The predictive model calculates the basic amount of chemicals required for a single PCB board based on product specifications; and calculates the first amount of chemicals required for a pre-set production unit of the same PCB board based on the basic amount and correction factor. The prediction model is as follows: ;in, Add to the base amount, This refers to the board area of ​​a single PCB board. The weighting factor is the plate area. This represents the total surface area of ​​the microvias on a single PCB board. This is the weighting coefficient for the total surface area of ​​the micropores. As a reference constant, The reactivity coefficient, This is the weighting factor for the reactivity coefficient; The formula for calculating the first amount added is: ;in, This is the first amount to add. K To correct the parameters, m The number of PCBs corresponding to the preset unit.

[0010] In one embodiment, the total surface area of ​​the micropores The calculation formula is: ;in, d i For the first i Pore ​​diameter of the pore type, h i For the first i Hole depth of the hole type, N i For the first i The number of hole types; the i These represent different hole types, including blind holes and through holes with different diameters and depths.

[0011] In one embodiment, step S40, calculating the chemical adjustment amount of the tank solution based on the target concentration and the second concentration, includes: The real-time concentration of the bath solution at the third moment is calculated based on the second concentration, the current PCB board production unit, and the first consumption. The third time point is the time when the second concentration result is actually obtained, and the time interval Δ between the second and third time points is... t Total time required to obtain the second concentration; The real-time concentration The calculation formula is as follows: ; in, The second concentration, This is the first amount to add. The first consumption amount, The interval duration Δ t Number of PCBs for internal processing m ≥1, The concentration of the medicine. This refers to the volume of the reaction tank; Based on target concentration and real-time concentration Calculate the chemical adjustment amount of the tank solution at the third time step. The amount of medicine to be adjusted The calculation formula is as follows: .

[0012] In one embodiment, step S40, adjusting the concentration of the solution in the reaction tank according to the dosage of the chemical solution, includes: Step S421: Determine whether the real-time concentration of the solution at the third moment is less than or equal to the target concentration of the solution. If yes, add more chemicals to the reaction tank according to the chemical adjustment amount, and then proceed to step S30; otherwise, proceed to step S422. Step S422: Obtain the number of PCBs entering the reaction tank as the first board number, and determine whether the chemical adjustment amount is less than or equal to the updated first addition amount. If yes, proceed to step S423; otherwise, proceed to step S424. Step S423: When the number of the first board reaches the number of PCB boards corresponding to the preset unit, add chemical according to the difference between the updated first addition amount and the chemical adjustment amount, and execute step S30; Step S424: When the number of the first board reaches the number of PCB boards corresponding to the preset unit, update the chemical adjustment amount to the difference between the chemical adjustment amount and the first addition amount, then clear the number of the first board and execute step S422.

[0013] In one embodiment, the formula for calculating the correction coefficient in step S40 is as follows: ;in, The first consumption amount, This represents the basic addition amount for a single PCB board, calculated by the prediction model based on the PCB board's product specifications. m The number of PCBs corresponding to the preset unit.

[0014] In one embodiment, the closed-loop control method for PCB solution concentration further includes: Step S50: Determine whether the number of cycles of the control cycle has reached the preset number of cycles, or determine whether the first addition amount has converged to the first consumption amount. If yes, store the corresponding correction coefficient in the database and execute step S60; otherwise, execute step S30. Step S60: Based on the correction coefficient in the database, combined with the prediction model and product specification parameters, calculate the second amount of chemical solution required for producing a preset unit of the PCB board. For each preset unit of PCB board produced, add the chemical solution according to the second amount of chemical solution until the production of the current model of PCB board is completed.

[0015] This invention also provides a system for closed-loop control of PCB solution concentration based on multi-parameter adaptive methods, comprising: The data acquisition unit is used to acquire the product specification parameters of the current model of PCB board to be produced. The product specification parameters include at least the board area and micro-hole parameters of a single PCB board. The micro-hole parameters are the total surface area of ​​the micro-holes, or the micro-hole parameters include the number, diameter and depth of various hole types. The first calculation unit is used to obtain the correction coefficient of the PCB board and calculate the first amount of chemical to be added for producing a preset unit of PCB board by combining the prediction model and product specification parameters. The first operating unit is used to obtain the actual concentration of the reaction tank solution at the beginning of the current control cycle as the first concentration; and to add a first amount of chemical solution to the reaction tank for each preset unit of PCB board produced, until the end of the current control cycle, and then obtain the actual concentration of the solution at the second moment as the second concentration. The feedback correction unit is used to calculate the first consumption amount corresponding to the preset unit of PCB board based on the first concentration and the second concentration, update the correction coefficient and the first addition amount of the PCB board based on the first consumption amount, calculate the real-time concentration of the bath solution at the third time based on the second concentration, the first consumption amount, the number of PCB boards produced within the interval from the second time to the third time, calculate the chemical adjustment amount based on the real-time concentration at the third time and the target concentration, and adjust the concentration of the bath solution in the reaction tank according to the chemical adjustment amount.

[0016] In addition, the present invention provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, causes the processor to perform the closed-loop control method for PCB chemical concentration.

[0017] To better understand and implement this invention, the following detailed description is provided in conjunction with the accompanying drawings. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the closed-loop control method for PCB drug concentration based on multi-parameter adaptive method of the present invention.

[0019] Figure 2 This is a schematic flowchart of the closed-loop control method for PCB drug concentration based on multi-parameter adaptive method according to Embodiment 1 of the present invention.

[0020] Figure 3 This is a schematic diagram of the system structure of Embodiment 1 of the present invention.

[0021] Figure 4 A schematic flowchart of the closed-loop control method for PCB drug concentration based on multi-parameter adaptive method in Embodiment 2 of the present invention. Detailed Implementation

[0022] The chemicals used in PCB production are typically supplied by vendors, along with theoretical consumption values ​​per square meter. Fixed-frequency automatic replenishment methods calculate the total consumption based on these theoretical values, combined with the actual area and quantity of PCBs used, and replenish the chemicals at set times and in precise quantities. However, these theoretical values ​​have a fundamental flaw, resulting in a significant and unavoidable deviation from the actual consumption on the production line.

[0023] First, the supplier's theoretical values ​​are based on standard-specification, standard-material PCBs, calculated under ideal conditions such as fixed temperature and fixed immersion time, considering only the chemical reaction consumption between metal and chemicals. However, actual production lines use a variety of PCB specifications, with copper thicknesses ranging from 0.5oz to 3oz, and differences in circuit density, layer structure, and surface treatment methods (such as OSP and ENIG). Greater copper thickness requires more metal for etching and electroplating; higher circuit density results in a larger effective reaction area and higher chemical consumption. The differences in consumption between different PCB specifications cannot be eliminated by a uniform correction factor.

[0024] Secondly, this theoretical value completely fails to account for the carry-in and removal of chemicals by the micropores of the PCB board. In actual production, after the previous wet process, residual chemicals remain on the surface and in the micropores of the PCB board. When it enters the current process, these residual chemicals are carried into the reaction tank, and the chemicals in this process are also adsorbed by the micropores and carried out with the board surface to the next process. The number of micropores varies greatly among different PCB specifications (usually tens of thousands to hundreds of thousands, and up to millions for high-density boards), and the amount of chemicals carried out / removed by the micropores fluctuates significantly accordingly. The theoretical consumption value provided by the supplier does not take this variable into account at all.

[0025] More importantly, the impact of micropores on consumption goes far beyond just carryover—for processes like chemical copper plating, the total surface area of ​​the micropore walls is the dominant factor in chemical consumption. The copper plating reaction occurs on the copper surface of the hole walls; the more micropores, the larger the hole diameter, and the thicker the board layer, the larger the total surface area of ​​the hole walls, and the higher the amount of chemicals involved in the reaction. This variable varies significantly between different board sizes and cannot be corrected by any uniform coefficient; the supplier's theoretical values ​​completely fail to reflect this.

[0026] Furthermore, this theoretical value is calculated only for a single process and does not consider the cumulative effects between multiple processes. On actual production lines, PCB boards undergo multiple wet processes, including degreasing, micro-etching, pre-dip / activation, and copper plating. Each process generates carry-over losses, and the effectiveness of the previous process directly affects the chemical consumption of subsequent processes—for example, insufficient degreasing can prevent the micro-etching process from uniformly contacting the copper surface, thus affecting chemical consumption and product quality. These cumulative effects across processes cause the actual consumption to deviate significantly from the theoretical calculation value for a single process.

[0027] In summary, the theoretical values ​​provided by suppliers based on standard boards, single processes, and only considering chemical reactions differ significantly and irremovably from the actual consumption amounts across multiple specifications and processes on the production line. This is the fundamental reason why the fixed-frequency automatic replenishment scheme needs to be re-optimized after changing the PCB board model.

[0028] Based on the problems identified in the above research, this invention proposes a closed-loop control method and system for PCB chemical concentration based on multi-parameter adaptive control. The closed-loop control method for PCB chemical concentration includes: Step S10: Obtain the product specifications of the current model of PCB board to be produced.

[0029] Step S20: Obtain the correction coefficient of the PCB board, calculate the first amount of chemical to be added for the production of the preset number of PCB boards by combining the prediction model and product specification parameters, and then execute step S30.

[0030] Step S30: At the start of the current control cycle, obtain the first moment. t The actual concentration of the reaction tank solution is taken as the first concentration; for each preset unit of PCB board produced, the first amount of reagent is added to the reaction tank until the current control cycle ends, and the second time point is obtained. t The actual concentration of the solution in tank 2 is used as the second concentration, and then step S40 is executed.

[0031] Step S40: Calculate the first consumption amount corresponding to the preset unit of PCB board based on the first concentration and the second concentration, update the correction coefficient and the first addition amount of the PCB board based on the first consumption amount; calculate the chemical adjustment amount of the bath solution based on the target concentration and the second concentration, adjust the concentration of the bath solution in the reaction tank according to the chemical adjustment amount, and then execute step S30.

[0032] Furthermore, in step S20, obtaining the correction coefficient of the PCB board includes: comparing the product specification parameters of the current model PCB board with the product specification parameters of existing PCB boards in the database to determine whether the PCB board is a new model PCB board. If so, the product specification parameters of the PCB board are stored, and the initial parameters are used as the correction parameters of the PCB board; if not, the correction coefficient of the PCB board is retrieved from the database.

[0033] Furthermore, step S40 also includes: determining whether the current model of PCB board has been produced. If not, step S30 is executed; if yes, the process is stopped, or step S10 is executed for the next model of PCB board.

[0034] Furthermore, to conserve system computing power and avoid endless adjustments to the correction coefficient, the closed-loop control method for PCB solution concentration also includes: Step S50: Determine whether the number of cycles of the control cycle has reached the preset number of cycles, or determine whether the first addition amount has converged to the first consumption amount. If yes, store the corresponding correction coefficient in the database and execute step S60; otherwise, execute step S30.

[0035] Step S60: Calculate the second addition amount for the PCB board based on the correction coefficient, product specification parameters and preset number of boards in the database. Add the chemical solution according to the second addition amount for each preset number of PCB boards produced, until the production of the current model of PCB board is completed.

[0036] The method and system for closed-loop control of PCB drug concentration based on multi-parameter adaptive method of the present invention are specifically described through the following embodiments.

[0037] Example 1 This embodiment provides a closed-loop control method and system for PCB chemical concentration based on multi-parameter adaptive methods. (See also...) Figure 2 and Figure 3 , Figure 2 This is a schematic flowchart of the closed-loop control method for PCB chemical concentration in this embodiment. Figure 3 This is a schematic diagram of the system structure in this embodiment. The system described in this embodiment includes a data acquisition unit 10, a first calculation unit 20, a first running unit 30, and a feedback correction unit 40.

[0038] The data acquisition unit 10 is used to perform step S10: acquire the product specification parameters of the current model of PCB board to be produced.

[0039] Specifically, the product specifications of a PCB include the substrate type and board area. And micropore parameters. The substrate types include FR-4, aluminum-based, high-frequency boards, etc., and the reactivity coefficients of different types of substrates with the chemical solution are also considered. Unlike other methods, their specific values ​​can be obtained by pre-training machine learning algorithms based on historical batch data.

[0040] The micropore parameters can be the total surface area of ​​the micropores. Alternatively, the micropore parameters are used to calculate the total surface area of ​​the micropores. pore size of micropores Hole depth h Number of holes Key parameters, etc. The micropores include through holes and blind holes of different sizes.

[0041] Total surface area of ​​micropores on each PCB board The calculation formula is as follows: ; in, d i For the first i The pore size of the micropores of various specifications h i For the first i The pore depth of various sizes of micropores, N i For the first i The number of pores in a certain specification of micropores.

[0042] Product specifications can be obtained manually or by scanning a barcode. Specifically, the PCB board has a barcode sprayed on its side. Different models of PCB boards have different barcode information. The acquisition unit 10 can obtain the product specifications of the PCB board by scanning the barcode.

[0043] The first calculation unit 20 is used to perform step S20: obtain the correction coefficient of the PCB board, calculate the first amount of chemical to be added for the PCB board required for the production of the preset unit by combining the prediction model and product specification parameters, and then perform step S30.

[0044] Specifically, step S20 includes the following steps: Step S21: Compare the product specifications of the current PCB model with the product specifications of existing PCBs in the database to determine if the PCB is a new model. If so, store the product specifications of the PCB in the database, and use the initial parameters.K 0 is used as the correction parameter for this PCB board. K If not, retrieve the correction factor for the PCB board from the database. K .

[0045] Step S22: Based on the correction coefficient of the PCB board, combined with the prediction model and product specification parameters, calculate the first amount of chemical to be added for the production of the preset unit of PCB board, and then execute step S30.

[0046] Specifically, predictive models are used to calculate the basic addition amount for each PCB board. .

[0047] ;in, This is the parameter vector of the prediction model, which is obtained by pre-training a machine learning algorithm based on historical batch data.

[0048] In this embodiment, the prediction model is a linear model, a vector model. The parameter vector Including plate area weighting factor Total surface area of ​​micropores weighting coefficient Weighting coefficient of reactivity coefficient and reference constant .

[0049] For example, in a linear implementation, the mapping function is: .

[0050] In other embodiments, the prediction model may employ logistic regression, linear regression, multinomial regression, neural network, support vector machine, fuzzy logic, etc.

[0051] Based on the correction factor of the PCB board K Basic addition amount Calculate the correction amount for each PCB board. .

[0052] The first addition amount of the chemical required for a preset unit of PCB board is calculated based on the corrected addition amount. .

[0053] First addition amount The calculation formula is as follows: .

[0054] .

[0055] In this embodiment, the initial correction coefficient It is 1.0. m This represents the number of PCBs produced corresponding to a preset unit. When the preset unit is a preset area, m represents the number of PCBs corresponding to the preset area; when the preset unit is a preset number of boards, m represents the number of PCBs produced. m The preset number of boards. In this embodiment, the preset unit is the preset number of boards. m ≥1.

[0056] The first operating unit 30 is used to execute step S30: at the beginning of the current control cycle, the actual concentration of the reaction tank solution at the end of the previous control cycle is obtained as the first concentration; for each preset unit of PCB board produced, a first amount of chemical solution is added to the reaction tank until the end of the current control cycle, the actual concentration of the solution at the end of the current control cycle is obtained as the second concentration, and then step S40 is executed.

[0057] Specifically, the first operating unit 30 includes a first analysis unit 31, a first judgment unit 32, a first addition unit 33, a second judgment unit 34, and a second analysis unit 35. The preset unit uses a preset number of boards. m .

[0058] The first analysis unit 31 is used to execute step S31: when the current control cycle begins, obtain the first moment. t 1. Actual concentration of the solution in the reaction tank at the start of the current control cycle. As the first concentration.

[0059] The first determination unit 32 is used to execute step S32: obtain the number of PCBs of the current model entering the reaction tank as the first board number. n Determine the number of the first board. n Has the preset number of boards been reached? m If not, do not add any more medicine and continue to determine the number of the first plate. n Has the preset number of boards been reached? m If so, then the first board number will be... n Clear the cache and then proceed to step S33.

[0060] The first adding unit 33 is used to perform step S33: adding the first amount The medicine solution is added to the reaction tank and mixed with the tank liquid.

[0061] The preset number of boards can be preset manually or automatically calculated through a prediction model.

[0062] Preferably, the number of times the data is reset is counted (i.e., based on the first amount added). (Number of times the solution was added) or the cumulative number of PCBs produced within the current control cycle. U Calculate the cumulative amount of medicine added during the current regulation period. .

[0063] The second judgment unit 34 is used to execute step S34: determine the duration of the current control cycle. T Whether the preset time period has been reached, or, to determine the cumulative number of PCBs produced within the current control cycle. U Has the preset quantity been reached? If yes, proceed to step S35; otherwise, proceed to step S32.

[0064] The control cycle can be determined based on continuous production time, for example, the preset duration of each control cycle is 1 hour. The control cycle can also be determined based on the number of PCBs produced. For example, the preset quantity of PCB boards to be produced in each control cycle is 200 pieces.

[0065] The second analysis unit 35 is used to execute step S35: when the current control cycle ends, obtain the second moment. t 2. Actual concentration of the tank solution at the end of the current control cycle. As the second concentration, then step S40 is performed.

[0066] In subsequent control cycles, the end of the previous control cycle is the start of the new current control cycle. Therefore, the second concentration at the second moment of the previous control cycle is used as the first concentration at the first moment of the new current control cycle to avoid repeated measurements.

[0067] The first analysis unit 31 and the second analysis unit 35 can be independent, executing steps S31 and S35 respectively, or the system can only have the first analysis unit 31, which is used to execute steps S31 and S35.

[0068] The feedback correction unit 40 is used to execute step S40: enter the next control cycle, and at the same time calculate the first consumption of the PCB board corresponding to the preset unit according to the first concentration and the second concentration, update the correction coefficient and the first addition amount of the PCB board according to the first consumption; calculate the chemical adjustment amount of the tank solution according to the target concentration and the second concentration, adjust the concentration of the tank solution in the reaction tank according to the chemical adjustment amount, and then execute step S30.

[0069] Specifically, the feedback correction unit 40 includes a first data processing unit 41 and a medicine adjustment unit 42.

[0070] The first data processing unit 41 is used to execute step S41: enter the next control cycle, and simultaneously calculate the first consumption amount corresponding to the preset unit of PCB board based on the first concentration and the second concentration. According to the first consumption Update the correction factor for the current PCB model.K According to the correction factor K Update the first addition amount ; Calculate the chemical adjustment volume of the reaction tank solution based on the target concentration and the second concentration. .

[0071] Calculate the actual volume of medicine required to maintain the target concentration during the current control period based on the first and second concentrations, i.e., the total consumption. .

[0072] As an example, total consumption .

[0073] in, This represents the actual volume of medicine added during the current adjustment period. This represents the actual concentration of the tank solution at the start of the current control cycle. This represents the actual concentration of the tank solution at the end of the current control cycle. This refers to the volume of the reaction tank.

[0074] Based on the cumulative number of PCBs produced during the current regulatory cycle This allows for further calculation of the unit consumption per PCB board. .

[0075] Then calculate the preset unit (i.e., the preset quantity). m） The first consumption corresponding to the PCB board .

[0076] Based on unit consumption Or the first consumption Update the correction coefficient for the next adjustment cycle. : .

[0077] The first addition amount for the next adjustment cycle will be updated based on the updated correction coefficient. .

[0078] Calculate the chemical adjustment amount for the reaction tank solution based on the target concentration and the secondary concentration. The calculation formula is as follows: ; in, This represents the actual concentration of the tank solution at the end of the current control cycle. The target concentration of the bath solution. This refers to the concentration of the mother liquor, which is used to adjust the actual concentration of the reaction tank solution. This represents the volume of the reaction tank.

[0079] The medicine adjustment unit 42 is used to perform step S42: adjusting the amount of medicine. Adjust the concentration of the solution in the reaction tank, and then proceed to step S31.

[0080] Specifically, step S42 includes: Step S421: Determine whether the second concentration of the recently concluded regulation cycle is less than or equal to the target concentration. If so, adjust the dosage according to the medicine. Add the chemical solution to the reaction tank and mix it with the tank solution, then proceed to step S31; otherwise, proceed to step S422.

[0081] In step S421, if the actual concentration of the tank solution... equal to target concentration If the potion adjustment amount is 0, no additional potion needs to be added.

[0082] If the second concentration is less than the target concentration And the amount of medicine to be adjusted When the concentration is less than the preset threshold, the deviation between the actual concentration of the solution and the target concentration is small and can be ignored. Therefore, no chemical replenishment is required, and step S31 is executed directly. The preset threshold can be set according to the type of chemical.

[0083] If the medicine was added and the dosage adjusted according to step S421 The actual volume of the medicine added in the new current regulation cycle is... The formula for the sum of the total amount of potion added according to the first addition amount and the total amount of potion added according to the potion adjustment amount is as follows: .

[0084] Step S422: Compare the adjustment amount of the medicine With the first addition of the update If the dosage of the medicine is adjusted Less than or equal to the first addition amount of the update Then proceed to step S423, if the amount of medicine is adjusted. Greater than the first addition amount of the update Then proceed to step S424.

[0085] Step S423: Obtain the number of PCBs of the current model entering the reaction tank as the first board count. n， When the first board number n Reaching the preset number of boards m At that time, add according to the first amount added in the update. Adjusting the dosage of medicine The difference V 3. Add the potion, then add the first plate number. n Clear to zero and proceed to step S31.

[0086] .

[0087] Step S424: Obtain the number of PCBs of the current model entering the reaction tank as the first board count. n， When the first board number n Reaching the preset number of boards m At that time, without adding any medicine, count the first batch of plates. n Reset to zero and adjust the potion dosage. Updated to adjust potion dosage With the first addition amount The difference V 4. Then proceed to step S422.

[0088] .

[0089] If the deduction was made according to the drug adjustment amount in the current control cycle according to steps S422~S424, then the actual volume of drug added in the new current control cycle is... : .

[0090] Preferably, step S42 further includes: before performing step S421, determining whether the PCB board of the current model has been manufactured; if not, then performing steps S421~S424 to adjust the amount of chemical solution. Adjust the concentration of the solution in the reaction tank, and then execute step S31. If so, stop the operation or execute step S10 for the next model of PCB board. More preferably, before executing step S10, adjust the concentration of the solution in the reaction tank according to the chemical adjustment amount.

[0091] The method for determining whether the production of the current model of PCB board has been completed includes: obtaining the production target (i.e., total production volume) for the current model of PCB board; calculating the total production volume of the current model of PCB board based on the total number of control cycles and the cumulative number of PCB boards produced in each control cycle; and determining that the production of the current model of PCB board has been completed when the total production volume reaches the production target.

[0092] In step S40, by adding and subtracting chemicals, the concentration of the solution in the reaction tank can be quickly restored to the target concentration. This reduces the fluctuation range of the chemical concentration in the reaction tank and improves the quality stability of the PCB board.

[0093] Example 2 To reduce the computational load and save computing power, this embodiment provides a closed-loop control method and system for PCB chemical concentration based on multi-parameter adaptive control. (See attached document.) Figure 4The similarities with Embodiment 1 will not be repeated here. The difference is that the system in this embodiment also includes a third judgment unit 50 and a second running unit 60.

[0094] Step S42: Adjust the dosage according to the medicine solution Adjust the concentration of the solution in the reaction tank, and then proceed to step S50.

[0095] The third judgment unit 50 is used to execute step S50: judging whether the number of cycles of the control cycle has reached the preset number of cycles, or judging the first addition amount. Does it converge to the first consumption? If yes, then store the corresponding correction coefficient in the database and execute step S60; otherwise, execute step S30.

[0096] The second operating unit 60 is used to execute step S60: based on the correction coefficient in the database, combined with the prediction model and product specification parameters, calculate the second amount of chemical solution required for producing a preset unit of the PCB board, and add chemical solution according to the second amount for each preset unit of PCB board produced, until the production of the current model of PCB board is completed.

[0097] Specifically, the second operating unit 60 includes a second calculation unit 61, a third judgment unit 62, a second addition unit 63, and a fourth judgment unit 64.

[0098] The second calculation unit 61 is used to execute step S61: based on the correction coefficient in the database, combined with the prediction model and product specification parameters, calculate the second amount of chemical to be added for the production of the PCB board of the preset unit. : .

[0099] .

[0100] in, K The correction coefficients are stored in the database in step S50.

[0101] The third judgment unit 62 is used to execute step S62: obtain the number of PCBs of the current model entering the reaction tank as the first board number. n Determine the number of the first board. n Has the preset number of boards been reached? m If not, do not add any more medicine and continue to determine the number of the first plate. n Has the preset number of boards been reached? m If so, then the first board number will be... n Clear the cache and then proceed to step S63.

[0102] The second adding unit 63 is used to perform step S63: adding the second amount The medicine solution is added to the reaction tank and mixed with the tank liquid.

[0103] Preferably, the number of times the data is reset is counted (i.e., based on the second addition amount). (Number of times the solution was added) or the cumulative number of PCBs produced within the current control cycle. U Calculate the cumulative amount of medicine added during the current regulation period. .

[0104] The fourth judgment unit 64 is used to execute step S64: determine whether the PCB board of the current model has been produced. If not, execute step S62. If yes, stop running or execute step S10 for the next model of PCB board.

[0105] Example 3 In PCB solution concentration control systems, there is an inherent delay in concentration measurement: the system experiences a delay from sampling (second time point). t 2. From collecting the bath solution sample to outputting the analytical data (the actual result of obtaining the second concentration), it typically takes several minutes. During this time, the system continues to run—processing the PCB board and replenishing the solution according to the original initial addition amount. This means that the second concentration measured by the system actually reflects the state of the bath solution several minutes ago, rather than the true concentration at the current moment. In other words, when the second concentration measurement is completed, the real-time concentration in the reaction tank may deviate from the actual concentration. t The actual concentration at time 2 (whether it deviates, and the degree of deviation, mainly depends on the deviation between the calculated first addition amount and the actual consumption amount). If the adjustment is made directly based on the second concentration, there will still be an error between the actual concentration and the target concentration.

[0106] To reduce the deviation between the real-time concentration and the target concentration, this embodiment adopts a predictive compensation control strategy: ① Real-time concentration prediction: Based on the measured second concentration, the third time step is calculated. t 3. Real-time concentration of the tank solution at the moment when the system actually measures the second concentration. C t ② Precise adjustment calculation: based on predicted concentration C t Calculate the required amount of medicine to adjust; ③ Implement control: Add or subtract medicine according to the calculated amount of medicine to make the actual concentration approach the target concentration.

[0107] in, t 2 to t 3 interval duration Δ t The time required for the system to complete the second concentration measurement, t 1 to t The duration of 3 is the duration of one control cycle. During this period, the volume of potion continuously added by the system is recorded as follows:V t The number of PCBs processed is denoted as P t —Both as predictive models C t The key input parameters.

[0108] The similarities between this embodiment and Embodiment 1 will not be repeated here. The difference lies in the following steps S30 to S40 of this embodiment: Step S30: At the start of the current control cycle, obtain the first moment. t The actual concentration of the reaction tank solution is taken as the first concentration; within the current control cycle, for every preset unit of PCB board produced, the first amount of reagent is added to the reaction tank; when the current control cycle is about to end, the second time point is collected. t The concentration of the solution in tank 2 was measured to obtain the second time step. t The actual concentration of the solution in tank 2 is used as the second concentration, and then step S40a is executed.

[0109] Step S40a: Proceed to the next control cycle and calculate the first consumption of the PCB board per preset unit based on the first and second concentrations. According to the first consumption Update the correction factor for the current PCB model. K According to the correction factor K Update the first addition amount .

[0110] Specifically, based on the first and second concentrations, the duration from the first moment to the second moment of the current control cycle is calculated as the actual volume of medicine to be added to maintain the target concentration, i.e., the total consumption. .

[0111] As an example, total consumption .

[0112] in, This represents the actual volume of medicine added during the time interval from the first moment to the second moment of the current control cycle. This represents the actual concentration of the solution at the first moment. This represents the actual concentration of the solution in the tank at the second moment. This refers to the volume of the reaction tank.

[0113] Based on the cumulative number of PCBs produced during the first and second moments of the current control cycle. This allows for further calculation of the unit consumption per PCB board. .

[0114] Then calculate the preset unit (i.e., the preset quantity). mThe first consumption corresponding to the PCB board. .

[0115] Based on unit consumption Or the first consumption Update the correction coefficient for the next adjustment cycle. : .

[0116] The first addition amount for the next adjustment cycle will be updated based on the updated correction coefficient. .

[0117] Step S40b: Based on the second concentration and the first consumption or unit consumption Calculate the third time point t 3. The real-time concentration of the tank solution at the end of the current control cycle is calculated using the following formula: .

[0118] Right now .

[0119] in, For the third moment t Real-time concentration of the solution in tank 3. The target concentration of the bath solution. This refers to the concentration of the mother liquor of the chemical solution, which is used to adjust the actual concentration of the solution in the reaction tank. The volume of the reaction tank. The interval duration Δ t The volume of the added medicine, The interval duration Δ t Number of PCBs for internal processing This refers to the volume of chemicals required for a single PCB board. The first consumption amount, m The number of PCBs produced corresponding to the preset unit.

[0120] Step S40c: Based on the target concentration and real-time concentration Calculate the third time point t 3. Adjustment amount of chemical solution for the corresponding tank solution The calculation formula is as follows: ; in, For the third moment t Real-time concentration of the solution in tank 3. The target concentration of the bath solution. This refers to the concentration of the mother liquor, which is used to adjust the actual concentration of the reaction tank solution. This represents the volume of the reaction tank.

[0121] Step S40d: Determine the real-time concentration corresponding to time t3. Is it less than or equal to the target concentration? If so, adjust the dosage according to the medicine. Add the chemical solution to the reaction tank and mix it with the tank solution, then proceed to step S31; otherwise, proceed to step S40e.

[0122] Step S40e: Compare the adjustment amount of the medicine With the first addition of the update If the dosage of the medicine is adjusted Less than or equal to the first addition amount of the update Then proceed to step S40f, if the amount of medicine is adjusted. Greater than the first addition amount of the update Then proceed to step S40g.

[0123] Step S40f: Obtain the number of PCBs of the current model entering the reaction tank as the first board count. n， When the first board number n Reaching the preset number of boards m At that time, add according to the first amount added in the update. Adjusting the dosage of medicine The difference V 3. Add the potion, then add the first plate number. n Clear to zero and proceed to step S31.

[0124] .

[0125] Step S40g: Obtain the number of PCBs of the current model entering the reaction tank as the first board count. n， When the first board number n Reaching the preset number of boards m At that time, without adding any medicine, count the first batch of plates. n Reset to zero and adjust the potion dosage. Updated to adjust potion dosage With the first addition amount The difference V 4. Then proceed to step S40e.

[0126] .

[0127] Referring to the descriptions in Examples 1-3, the following section uses an actual production process as an example to specifically illustrate the method for dynamic control of the chemical plating solution in a PCB factory: Chemical plating reactors (CCP reactors) are used to deposit conductive copper layers on the walls of vias in PCBs. Copper ions (Cu) within the reactor... 2+ Concentration is a key control indicator, and the target concentration is... Effective volume of the tank The mother liquor of the medicine is a high-concentration copper sulfate solution, and its concentration... The production line operates on a multi-specification mixed production model, with products exhibiting significant differences in single-board area, number of holes, hole diameter, hole depth, and substrate type.

[0128] The system for the chemical copper plating wire includes an addition control unit and a metering pump. The addition control unit drives the metering pump to add a pre-concentrated copper sulfate solution (high-concentration copper sulfate solution) to the reaction tank at a preset dosage.

[0129] Step S10: Scan the barcode information on the PCB board of the current model to be produced to obtain the product specification parameters of the PCB board (as shown in Table 1).

[0130] Table 1 Product specifications of the PCB boards to be processed

[0131] Calculate the total surface area of ​​micropores : m 2 / pcs.

[0132] Step S20: Obtain the correction coefficient of the PCB board, calculate the first amount of chemical to be added for the PCB board required for the production of the preset unit by combining the prediction model and product specification parameters, and then execute step S30.

[0133] In this embodiment, the product specifications of the PCB board differ from those of existing PCB boards in the database, indicating it is a new model PCB board. Therefore, the correction factor for this PCB board is... K Use initial coefficients, i.e. K =1.0.

[0134] During the first control cycle, the system operates according to steps S20-1 to S50-1.

[0135] Step S20-1: Obtain the correction coefficient of the PCB board, and calculate the first amount of chemical to be added for the PCB board required for the production of the preset unit by combining the prediction model and product specification parameters.

[0136] In this embodiment, the initial correction coefficient The weighting factor for the plate area is 1.0. Micropore surface area weighting coefficient Substrate weighting coefficient reference constant The production cycle is 200 PCB boards per cycle, with a preset unit of 5 PCB boards. That is, the mother liquor is added to the reaction tank once every 5 PCB boards are produced.

[0137] Initial vector .

[0138] Basic copper ion addition per PCB board for: .

[0139] Correction addition amount per PCB board .

[0140] The first addition amount corresponding to the preset unit PCB board = 68.8 mL.

[0141] Step S31-1: When opening the line (first moment) t 1-1 The actual concentration of the solution in the reaction tank As the first concentration.

[0142] Step S32-1: Obtain the number of PCBs of the current model entering the reaction tank as the first board count. n Determine the number of the first board. n Has the preset number of boards been reached? m If not, do not add any more medicine and continue to determine the number of the first plate. n Has the preset number of boards been reached? m If so, then the first board number will be... n Clear the cache and then proceed to step S33-1.

[0143] The system triggers an addition once per preset unit (every 5 pieces produced).

[0144] Step S33-1: The first addition unit of the system drives the metering pump to automatically inject 68.80 mL of high-concentration copper sulfate solution into the reaction tank.

[0145] Step S34-1: Determine the duration of the current adjustment cycle. T Whether the preset time period has been reached, or, to determine the cumulative number of PCBs produced within the current control cycle. U Has the preset quantity been reached? If yes, proceed to step S35-1; otherwise, proceed to step S32-1.

[0146] Step S35-1: The first control cycle is about to end (second moment) t 2-1 ), collect the concentration of the tank solution at this time for concentration detection, and at the third timet 3-1 Get the second moment t 2-1 The actual concentration of the tank solution is measured, and this result is taken as the second concentration, i.e., the second concentration is... .in, t 1-1 to t 3-1 The interval length is the total duration of the first control cycle.

[0147] Step S40a-1: Enter the second control cycle, and simultaneously calculate the first consumption of the PCB board per preset unit based on the first concentration and the second concentration of the first control cycle. According to the first consumption Update the correction factor for the current PCB model. K According to the correction factor K Update the first addition amount .

[0148] t 1-1 to t 2-1 The actual volume of potion added within the specified time period .

[0149] t 1-1 Concentration of the tank solution , t 2-1 Actual concentration of the tank solution , t 1-1 to t 2-1 A total of 200 PCB boards were produced during the production period. t 1-1 to t 2-1 The total amount of high-concentration copper sulfate solution added during the specified time period is: .

[0150] t 1-1 to t 2-1 The actual volume of medicine added required to maintain the target concentration within the specified time period, i.e., the total medicine consumption. for: .

[0151] Unit consumption per PCB board =3.252 / 200=16.26mL / pcs.

[0152] First consumption =81.3mL.

[0153] Based on unit consumption Or the first consumption Update the correction factor for the current PCB model. K .

[0154] =16.26 / 13.76=1.1817.

[0155] Based on the updated correction coefficients Update the first addition amount : = 81.30mL.

[0156] Step S40b-1: Based on the second concentration and the first consumption... or unit consumption And the current PCB board manufacturing unit calculates the third moment of the bath solution. t 3-1 Real-time concentration of the bath solution .

[0157] From the time the second concentration test sampling was completed to the time the results were obtained, 15 PCB boards were produced. P t It is 15. =0.0688×(15 / 5)=0.2064.

[0158] =(1.98×1000+0.2064×40-15×0.01626×40) / 1000=1.978g / L.

[0159] or, =1.98+(0.0688-0.0813)×(15 / 5)×40) / 1000=1.978g / L.

[0160] Step S40c-1: Based on the target concentration and real-time concentration Calculate the third time point t 3-1 Corresponding tank solution chemical adjustment volume .

[0161] .

[0162] in, For the third moment t 3-1Real-time concentration of the bath solution (copper ions), The target concentration of the bath solution (copper ions). This refers to the concentration of copper ions in the mother liquor (a high-concentration copper sulfate solution). This represents the volume of the reaction tank.

[0163] Step S40d-1: Third Time t 3-1 Real-time concentration of the tank solution (1.978 g / L) is less than the target concentration. (2.00g / L), medication adjustment volume The amount (0.55L) is greater than the preset threshold (0.03L), therefore 0.55L of high-concentration copper sulfate solution is added, and then steps S31-2 to S42-2 are executed. In the second control cycle, the correction coefficient... K use K new That is, 1.1817.

[0164] Step S31-2: Obtain the timeframe when the first control cycle is about to end (i.e., the second moment of the first control cycle). t 2-1 The first moment of the second regulatory cycle t 1-2 The actual concentration of the solution in the reaction tank at that time is taken as the first concentration, i.e., the first concentration. It is 1.98 g / L.

[0165] The second moment at the end of the first regulation cycle is the first moment at the beginning of the second regulation cycle. Therefore, the second concentration at the second moment at the end of the first regulation cycle is directly taken as the first concentration at the first moment at the beginning of the second regulation cycle.

[0166] Step S32-2: Obtain the number of PCBs of the current model entering the reaction tank as the first board count. n Determine the number of the first board. n Has the preset number of boards been reached? m If not, do not add any more medicine and continue to determine the number of the first plate. n Has the preset number of boards been reached? m If so, then the first board number will be... n Reset to zero, then proceed to step S33-2. The system triggers an addition once per preset unit (every 5 pieces produced).

[0167] Step S33-2: The first addition unit of the system drives the metering pump to automatically inject 81.30 mL of high-concentration copper sulfate solution into the reaction tank.

[0168] Step S34-2: Determine the duration of the current regulatory cycle. TWhether the preset time period has been reached, or, to determine the cumulative number of PCBs produced within the current control cycle. U Has the preset quantity been reached? If yes, proceed to step S35-2; otherwise, proceed to step S32-2.

[0169] Step S35-2: The second control cycle is about to end (i.e., the second moment of the second control cycle). t 2-2 ), collect the concentration of the tank solution at this time for concentration detection, and at the third time t 3-2 Get the second moment t 2-2 The actual concentration of the tank solution was measured, and this result was taken as the second concentration. Specifically, the second concentration... .in, t 1-2 to t 3-2 The interval length is the total duration of the second control cycle.

[0170] Step S40a-2: Enter the third control cycle, and simultaneously calculate the first consumption of the PCB board per preset unit based on the first and second concentrations of the second control cycle. According to the first consumption Update the correction factor for the current PCB model. K According to the correction factor K Update the first addition amount .

[0171] Second regulatory cycle t 1-2 to t 2-2 The actual volume of potion added within the specified time: .

[0172] The first concentration in the second regulation cycle Second concentration Then the second regulatory cycle. t 1-2 to t 2-2 The total amount of potion consumed within the specified time is: Total medicine consumption .

[0173] Consumption per PCB board: =3.327 / 200×1000=16.63mL / pcs.

[0174] First consumption =16.635×5=83.17mL.

[0175] Based on unit consumption Or the first consumption Update the correction factor for the current PCB model. K .

[0176] =16.63 / 13.76=1.2086.

[0177] Based on the updated correction coefficients Update the first addition: =16.63mL.

[0178] =83.15mL.

[0179] Step S40b-2: Based on the second concentration and the first consumption... or unit consumption And the current PCB board manufacturing unit calculates the third moment of the bath solution. t 3-2 Real-time concentration of the bath solution From the time the second concentration test sampling was completed to the time the results were obtained, 15 PCB boards were produced.

[0180] =(1.999×1000+0.0813×3×40-15×0.01663×40) / 1000=1.999g / L.

[0181] Step S40c-2: Based on the target concentration and real-time concentration Calculate the third time point t 3-2 Corresponding tank solution chemical adjustment volume .

[0182] Adjustment of medicine dosage .

[0183] Step S40d-2: Current real-time concentration of the medicine (1.999 g / L) is less than the target concentration. (2.00g / L), and the dosage of the medicine was adjusted. When the concentration is less than the preset threshold (0.03L), no additional solution is added, and step S31-3 is executed. At this time, the copper ion concentration in the reaction tank solution is 1.999g / L, and the correction factor is... K use K new That is, 1.2086.

[0184] Step S31-3: Obtain the moment when the second control cycle is about to end (i.e., the first moment of the third control cycle). t 1-3 , t 1-3 That is, the second moment of the second regulation cycle. t 2-2 The concentration of the solution in the reaction tank is taken as the first concentration, specifically, the first concentration is 1.999 g / L.

[0185] Step S32-3: Obtain the number of PCBs of the current model entering the reaction tank as the first board count. n Determine the number of the first board. n Has the preset number of boards been reached? m If not, do not add any more medicine and continue to determine the number of the first plate. n Has the preset number of boards been reached? m If so, then the first board number will be... n Reset to zero, then proceed to step S33-3. The system triggers an addition once per preset unit (every 5 pieces produced).

[0186] Step S33-3: The first addition unit of the system drives the metering pump to automatically inject 83.15 mL of high-concentration copper sulfate solution into the reaction tank.

[0187] Step S34-3: Determine the duration of the current adjustment cycle. T Whether the preset time period has been reached, or, to determine the cumulative number of PCBs produced within the current control cycle. U Has the preset quantity been reached? If yes, proceed to step S35-3; otherwise, proceed to step S32-3.

[0188] Step S35-3: The third control cycle is about to end (i.e., the second moment of the third control cycle). t 2-3 ), collect the concentration of the tank solution at this time for concentration detection, and at the third time t 3-3 Get the second moment t 2-3 The actual concentration of the tank solution was measured, and this result was taken as the second concentration. Specifically, the second concentration was... .

[0189] Step S40a-3: Enter the fourth control cycle, and simultaneously calculate the first consumption of the PCB board per preset unit based on the first and second concentrations of the third control cycle. According to the first consumption Update the correction factor for the current PCB model. K According to the correction factor KUpdate the first addition amount .

[0190] Third regulatory cycle t 1-3 to t 2-3 The actual volume of potion added within the specified time period .

[0191] The first concentration in the third regulation cycle Second concentration Then the third regulatory cycle. t 1-3 to t 2-3 The total amount of potion consumed within the specified time is: .

[0192] Consumption per PCB board: =3.251 / 200×1000=16.25mL / pcs.

[0193] First consumption =81.25mL.

[0194] Based on unit consumption Or the first consumption Update the correction factor for the current PCB model. K .

[0195] =16.25 / 13.76=1.1809.

[0196] Based on the updated correction coefficients Update the first addition: =81.25mL.

[0197] Step S40b-3: Based on the second concentration and the first consumption... or unit consumption And the current PCB board manufacturing unit calculates the third moment of the bath solution. t 3-3 Real-time concentration of the bath solution From the time the second concentration test sampling was completed to the time the results were obtained, 15 PCB boards were produced.

[0198] =(2.002×1000+0.08315×3×40-15×0.01625×40) / 1000=2.002g / L.

[0199] Step S40c-3: Based on the target concentration and real-time concentration Calculate the third time point t 3-3 Corresponding tank solution chemical adjustment volume .

[0200] Adjustment of medicine dosage .

[0201] Step S40d-3: Third moment of the bath solution t 3-3 real-time concentration (2.002 g / L) is greater than the target concentration. (2.00g / L).

[0202] Step S40e-3: Adjusting the amount of medicine (0.05L, or 50.00mL) is less than the updated addition amount. (81.25mL).

[0203] Step S40f-3: Calculate the first addition amount after the update. Adjust the dosage with the medicine The difference between : =81.25-50=31.25mL.

[0204] When the first batch of PCBs reaches the preset unit (5pcs), a control unit drives a metering pump to inject 31.25mL of high-concentration copper sulfate solution into the reaction tank. Then, the first batch of PCBs is reset to zero, and step S31-4 is executed, in which 81.25mL of high-concentration copper sulfate solution is automatically injected into the reaction tank for every 5pcs of PCBs produced.

[0205] The dynamic control methods for the fourth control cycle and subsequent control cycles are the same as those for the first to third control cycles, and will not be repeated here.

[0206] If the amount of a single PCB board added during the current adjustment period... Stable convergence to consumption When, or when the number of cycles in the control cycle reaches the preset number of cycles, the correction coefficient of the current control cycle is stored, and step S60 is executed until the PCB board of the current model is processed.

[0207] The unit consumption, correction coefficient, correction addition amount, and concentration fluctuation range for the first, second, third, and subsequent control cycles (from the fourth to the zth control cycle) are shown in Table 2.

[0208] Table 2 Parameters for each control cycle

[0209] As can be seen, when directly using the correction coefficient to perform overall scaling compensation on the base addition calculated by the prediction formula, the system only needs one cycle to adjust the correction coefficient. Adjusted to a value close to the true value (accurately identified at the end of the first control cycle). Minor deviations in subsequent control cycles caused by measurement noise or discrete additive actions will be addressed through... The adaptive update automatically corrects the concentration fluctuation range, keeping it within the acceptable range. Within this range, it is far superior to traditional methods. Even if the actual consumption during production changes slowly due to aging chemicals or equipment conditions, the system can still detect this through back-calculation in each cycle. Continuous updates and tracking enable long-term, high-precision control.

[0210] In summary, this invention constructs a three-layer control system of "multi-dimensional parameter-driven theoretical prediction + real-time concentration feedback correction + global correction coefficient self-updating", specifically reflected in the following aspects: (1) The first multi-dimensional calculation model of drug consumption based on micro-geometric features: abandoning the traditional single macro-parameter estimation method, the model incorporates refined parameters such as total surface area of ​​micropores and substrate properties, and realizes accurate addition calculation in multi-specification mixed production scenarios from the source by using model parameters pre-trained from historical data.

[0211] (2) The hierarchical adaptive mechanism of “feedforward prediction + real-time feedback + global correction coefficient adjustment”: the feedforward realizes on-demand supply to reduce lag, the real-time feedback quickly eliminates the current concentration deviation, and the global correction coefficient adjustment scales the basic addition amount based on the relative error between actual demand and prediction value. Without changing the internal parameters of the mature model, it effectively compensates for the system deviation caused by factors such as drug aging and equipment drift, and the system robustness and accuracy are significantly improved.

[0212] (3) Generalizability of the model structure: The prediction model protected by this invention is not limited to a specific linear formula, but covers all learnable models that map multidimensional feature parameters to additive quantities (such as linear regression, neural networks, support vector machines, fuzzy logic, etc.). The correction mechanism is independent of the internal structure of the model and only performs uniform compensation on the final output, which is applicable to all types of deployed prediction models.

[0213] Compared with existing technologies, the closed-loop control method and system for PCB chemical concentration of the present invention has the following advantages: (1) Significantly improves concentration stability and product yield. By controlling concentration fluctuations within an extremely narrow range, processing defects caused by improper drug concentration are effectively avoided, and the risk of batch quality accidents is greatly reduced.

[0214] (2) Perfectly adaptable to flexible manufacturing with mixed production of multiple specifications. Without the need for frequent manual intervention to adjust parameters, the system can automatically adapt to the production switching of different plate types and materials, thereby improving the flexibility of the production line.

[0215] (3) Reduce operating costs and chemical waste. Precise on-demand addition reduces excessive consumption of chemicals and waste discharge, while also reducing reliance on manual sampling and testing and highly experienced operators.

[0216] (4) Extend the service life of equipment and chemicals. A stable chemical environment reduces the corrosion of the tank by extreme concentrations and avoids premature failure of chemicals due to drastic concentration fluctuations.

[0217] (5) Protect existing model investments. Once the model weight parameters are trained, they can remain stable without online modification. They can adapt to changes in operating conditions simply by using external correction coefficients, thus reducing the complexity of model operation and maintenance.

[0218] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the embodiments of this application. The singular forms “a,” “the,” and “the” used in the embodiments and claims of this application are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that, unless otherwise stated, the terms “first,” “second,” “third,” etc., are used only to distinguish and not to describe a particular order or sequence, nor should they be construed as indicating or implying relative importance. The term “and / or” as used herein refers to and includes any or all possible combinations of one or more associated listed items. When the above description relates to drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. In the description of this application, those skilled in the art will be able to understand the specific meaning of the above terms in this application according to the specific circumstances.

[0219] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A closed-loop control method for PCB chemical concentration based on multi-parameter adaptive regulation, characterized in that, include: Step S10: Obtain the product specification parameters of the current model of PCB board to be produced. The product specification parameters include at least the board area and micro-hole parameters of a single PCB board. The micropore parameters are the total surface area of ​​the micropores, or the micropore parameters include the number, diameter and depth of various pore types; Step S20: Obtain the correction coefficient of the PCB board, calculate the first amount of chemical to be added for producing a preset unit of PCB board by combining the prediction model and product specification parameters, and then execute step S30. Step S30: At the start of the current control cycle, obtain the actual concentration of the reaction tank solution at the first moment as the first concentration; for each preset unit of PCB board produced, add the first amount of chemical solution to the reaction tank until the end of the current control cycle, obtain the actual concentration of the solution at the second moment as the second concentration, and then execute step S40. Step S40: Calculate the first consumption amount corresponding to the preset unit of PCB board based on the first concentration and the second concentration, update the correction coefficient and the first addition amount of the PCB board based on the first consumption amount; calculate the chemical adjustment amount of the bath solution based on the target concentration and the second concentration, adjust the concentration of the bath solution in the reaction tank according to the chemical adjustment amount, and then execute step S30.

2. The method according to claim 1, characterized in that, In step S20, the prediction model is a machine learning model pre-trained based on historical production data. The machine learning model is selected from any one of the following: linear model, logistic regression model, linear regression model, multinomial regression model, neural network model, support vector machine model, and fuzzy logic model.

3. The method according to claim 1, characterized in that, In step S20, obtaining the correction coefficient of the PCB board includes: comparing the product specification parameters of the current model PCB board with the product specification parameters of the existing PCB boards in the database to determine whether the PCB board is a new model PCB board. If so, the product specification parameters of the PCB board are stored, and the initial parameters are used as the correction parameters of the PCB board; if not, the correction coefficient of the PCB board is retrieved from the database.

4. The method according to claim 1, characterized in that, The product specifications also include the material of the PCB board and the reactivity coefficient of the chemical solution. In step S20, the calculation method for the first addition amount includes: The predictive model calculates the basic amount of chemicals required for a single PCB board based on product specifications; and calculates the first amount of chemicals required for a pre-set production unit of the same PCB board based on the basic amount and correction factor. The prediction model is as follows: ;in, Add to the base amount, This refers to the board area of ​​a single PCB board. The weighting factor is the plate area. This represents the total surface area of ​​the microvias on a single PCB board. This is the weighting coefficient for the total surface area of ​​the micropores. As a reference constant, The reactivity coefficient, This is the weighting factor for the reactivity coefficient; The formula for calculating the first amount added is: ;in, This is the first amount to add. K To correct the parameters, m The number of PCB boards corresponding to the preset unit. m ≥1.

5. The method according to claim 4, characterized in that, The total surface area of ​​the micropores The calculation formula is: ;in, d i For the first i Pore ​​diameter of the pore type, h i For the first i Hole depth of the hole type, N i For the first i The number of hole types; the i These represent different hole types, including blind holes and through holes with different diameters and depths.

6. The method according to claim 1, characterized in that, In step S40, calculating the chemical adjustment amount of the tank solution based on the target concentration and the second concentration includes: The real-time concentration of the bath solution at the third moment is calculated based on the second concentration, the current PCB board production unit, and the first consumption. Wherein, the third time point is the time when the second concentration result is actually obtained, and the time interval Δ between the second time point and the third time point is... t Total time required to obtain the second concentration; The real-time concentration The calculation formula is as follows: ； in, The second concentration, This is the first amount to add. The first consumption amount, The interval duration Δ t Number of PCBs for internal processing m ≥1, The concentration of the medicine. This refers to the volume of the reaction tank; Based on target concentration and real-time concentration Calculate the chemical adjustment amount of the tank solution at the third time step. The amount of medicine to be adjusted The calculation formula is as follows: 。 7. The method according to claim 6, characterized in that, In step S40, adjusting the concentration of the solution in the reaction tank according to the dosage of the chemical solution includes: Step S421: Determine whether the real-time concentration of the solution at the third moment is less than or equal to the target concentration of the solution. If yes, add more chemicals to the reaction tank according to the chemical adjustment amount, and then proceed to step S30; otherwise, proceed to step S422. Step S422: Obtain the number of PCBs entering the reaction tank as the first board number, and determine whether the chemical adjustment amount is less than or equal to the updated first addition amount. If yes, proceed to step S423; otherwise, proceed to step S424. Step S423: When the number of the first board reaches the number of PCB boards corresponding to the preset unit, add chemical according to the difference between the updated first addition amount and the chemical adjustment amount, and execute step S30; Step S424: When the number of the first board reaches the number of PCB boards corresponding to the preset unit, update the chemical adjustment amount to the difference between the chemical adjustment amount and the first addition amount, then clear the number of the first board and execute step S422.

8. The method according to claim 7, characterized in that, In step S40, the formula for calculating the correction coefficient is as follows: ;in, The first consumption amount, This represents the basic addition quantity for a single PCB board, calculated by the prediction model based on the PCB board's product specifications. m The number of PCBs corresponding to the preset unit.

9. The method according to any one of claims 1 to 8, characterized in that, Also includes: Step S50: Determine whether the number of cycles of the control cycle has reached the preset number of cycles, or determine whether the first addition amount converges to the first consumption amount; If so, store the corresponding correction coefficient in the database and proceed to step S60; If not, proceed to step S30; Step S60: Based on the correction coefficient in the database, combined with the prediction model and product specification parameters, calculate the second amount of chemical solution required to produce a preset unit of the PCB board. For each preset unit of PCB board produced, add the second amount of chemical solution to the reaction tank until the production of the current model of PCB board is completed.

10. A system for closed-loop control of PCB solution concentration based on multi-parameter adaptive regulation, characterized in that, include: The data acquisition unit is used to acquire the product specification parameters of the current model of PCB board to be produced. The product specification parameters include at least the board area and micro-hole parameters of a single PCB board. The micropore parameters are the total surface area of ​​the micropores, or the micropore parameters include the number, diameter and depth of various pore types; The first calculation unit is used to obtain the correction coefficient of the PCB board and calculate the first amount of chemical to be added for producing a preset unit of PCB board by combining the prediction model and product specification parameters. The first operating unit is used to obtain the actual concentration of the reaction tank solution at the beginning of the current control cycle as the first concentration; and to add a first amount of chemical solution to the reaction tank for each preset unit of PCB board produced, until the end of the current control cycle, and then obtain the actual concentration of the solution at the second moment as the second concentration. The feedback correction unit is used to calculate the first consumption amount corresponding to the preset unit of PCB board based on the first concentration and the second concentration, update the correction coefficient and the first addition amount of the PCB board based on the first consumption amount, calculate the real-time concentration of the bath solution at the third time based on the second concentration, the first consumption amount and the number of PCB boards produced within the interval from the second time to the third time, calculate the chemical adjustment amount based on the real-time concentration at the third time and the target concentration, and adjust the concentration of the bath solution in the reaction tank according to the chemical adjustment amount.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program that, when executed by a processor, causes the processor to perform the method according to any one of claims 1 to 9.

Images