Intelligent control method and system for electroplating soft treatment

Abstract

The invention discloses an intelligent control method and system for electroplating soft processing. The method comprises the steps that current working condition information and real-time monitoring data of a process tank are obtained; based on the real-time monitoring data, key characteristic parameters used for representing the bath solution state are extracted; based on the current working condition information and the key characteristic parameters, matching is carried out through a preset historical working condition database, and matched initial process parameters are obtained; starting processing according to the initial process parameters, acquiring monitoring data in the process tank at regular time in the processing process, and updating the key characteristic parameters according to the monitoring data; based on the updated key feature parameters, generating optimal process parameters through a preset adaptive process model by taking a preset quality target as a guide; and generating a process parameter adjusting instruction by taking the optimal process parameter as a current process parameter set value. According to the invention, a closed-loop control architecture integrating real-time sensing and dynamic optimization is constructed, so that the consistency and stability of the electroplating soft treatment process can be improved.

Description

Technical Field

[0001] This application relates to the field of electroplating soft treatment technology, and in particular to an intelligent control method and system for electroplating soft treatment. Background Technology

[0002] The electroplating soft treatment process refers to all chemical or physicochemical treatment steps in the main electroplating process that do not change the surface state of the workpiece through electrodeposition (i.e., without electricity). It mainly includes key processes such as degreasing, pickling, and passivation.

[0003] In existing electroplating soft treatments, core process parameters such as bath temperature, concentration, and treatment time are usually set by operators based on experience. They cannot be dynamically adjusted according to real-time changes in working conditions (such as the number of workpieces, the degree of contamination, and the state of the bath solution), resulting in large fluctuations in product quality and poor consistency. Summary of the Invention

[0004] The purpose of this application is to provide an intelligent control method and system for electroplating soft processing, so as to realize a fully automated intelligent control scheme with real-time perception and intelligent decision-making, thereby replacing human experience and ensuring that the soft processing process is always in the best state.

[0005] In a first aspect, this application provides an intelligent control method for soft electroplating treatment, comprising: Acquire current operating condition information and real-time monitoring data of the process tank, including pH value, conductivity, temperature and redox potential (ORP) of the solution in the process tank; Based on real-time monitoring data, key characteristic parameters for characterizing the state of the bath solution are extracted. Based on the current operating condition information and key characteristic parameters, the initial process parameters are obtained by matching through a preset historical operating condition database. Using the initial process parameters as the current process parameter settings, the process is started according to the preset process flow. During the process, the monitoring data in the process tank is acquired periodically, and the key characteristic parameters are updated based on the monitoring data. Based on the updated key feature parameters and guided by the preset quality target, the optimal process parameters are generated through a preset adaptive process model. The optimal process parameters are used as the current process parameter settings to generate process parameter adjustment instructions.

[0006] By acquiring the bath solution monitoring data in real time and extracting key characteristic parameters, the state of the bath solution can be accurately understood. Then, personalized initial process parameters can be obtained by intelligent matching of historical data, providing the optimal starting point for the workpiece to be processed. Furthermore, the process parameters are continuously monitored and dynamically optimized during the processing, forming a control closed loop of real-time perception and dynamic optimization. This ensures that the process parameters always adapt to the real-time operating conditions, thereby helping to improve the consistency and stability of the electroplating soft treatment process.

[0007] Optionally, the extraction of key characteristic parameters for characterizing the state of the bath solution based on real-time monitoring data includes: The real-time monitoring data is preprocessed, and the preprocessed real-time monitoring data is aligned and fused in the time dimension to form a unified time series dataset of tank liquid status. Based on the state time series dataset, key feature parameters are obtained through a pre-defined calculation method.

[0008] Optionally, the step of obtaining matching initial process parameters based on current operating condition information and key feature parameters by matching through a preset historical operating condition database includes: The current operating condition information and key feature parameters are encoded to generate an operating condition feature vector; Based on the working condition feature vector, similarity matching is performed through a preset historical working condition database to obtain the historical working condition record with the highest similarity and the corresponding similarity. If the similarity is higher than the preset first similarity threshold, the historical working condition with the highest similarity will be recorded as the corresponding historical process parameters and used as the initial process parameters for matching. If the similarity is lower than the preset first similarity threshold but higher than the preset second similarity threshold, then a valid fusion set is generated for all historical work condition records with similarity higher than the preset second similarity threshold. Based on the effective fusion set, the historical process parameters corresponding to each historical working condition record are weighted and fused to generate the initial process parameters, using the similarity of each historical working condition record as the weight. If the similarity is lower than the preset second similarity threshold, then the preset general safety process parameters are used as the initial process parameters.

[0009] Optionally, the step of generating optimal process parameters based on updated key feature parameters and guided by a preset quality target through a preset adaptive process model includes: Based on the current operating condition information and the updated key feature parameters, a working state vector is generated, which is used to characterize the real-time comprehensive status of the current process tank. Based on the working state vector and the current process parameter settings, the quality prediction value of the processing result is generated through a preset adaptive process model. Based on the predicted quality values, the quality deviation is calculated by setting pre-defined quality targets; Determine whether the quality deviation is below a preset threshold; If so, the current process parameter setting value will be taken as the optimal process parameter; If not, then based on the quality deviation, the optimal process parameters are generated through a preset optimization algorithm.

[0010] Optionally, the step of generating optimal process parameters based on quality deviation using a preset optimization algorithm includes: Based on the current process parameter settings and quality deviations, a set of candidate process parameters is generated; Traverse the set of candidate process parameters, and for each candidate process parameter, output the corresponding quality prediction result through a preset adaptive process model, and obtain the cost item through a preset process cost calculation method; Based on the quality prediction results and cost items, a comprehensive evaluation score is generated using preset weighting coefficients; After the traversal is complete, the candidate process parameter with the lowest comprehensive evaluation score is recorded as the optimal process parameter.

[0011] Optionally, the process tank includes multiple processing tanks connected in series. Adjacent processing tanks are designated as a preceding processing tank and a succeeding processing tank according to a preset procedure. During the processing, before starting processing in any succeeding processing tank, the following steps are also included: Obtain the final process parameters used in the pretreatment tank and the key characteristic parameters of the tank liquid when the treatment is completed, and record them together as feedforward process information; Add the feedforward process information to the current operating condition information to serve as the updated current operating condition information; Based on the updated current operating condition information, the initial process parameters corresponding to the subsequent processing tank are adjusted to generate the adjusted process parameters.

[0012] Optionally, adjusting the initial process parameters corresponding to the subsequent processing tank based on the updated current operating condition information to generate adjusted process parameters includes: The updated current operating condition information is compared with the initial operating condition information before processing begins to obtain the status feature increment; Based on the incremental state features, the corresponding standardized adjustment scheme is obtained through a preset incremental-parameter correction mapping table; Based on the standard adjustment scheme, the initial process parameters corresponding to the subsequent processing tank are adjusted to generate the adjusted process parameters.

[0013] Optionally, after periodically acquiring monitoring data from the process tank and updating key characteristic parameters based on the monitoring data during the processing, the method further includes: Based on key characteristic parameters, process health indicators are calculated, including component stability index and cumulative pollution rate. Based on the current operating condition information, a dynamic safety threshold is generated for the process health index through a preset dynamic threshold model. The process health indicators are compared with the corresponding dynamic safety thresholds. If any indicator exceeds its corresponding dynamic safety threshold, an abnormal prompt message is output.

[0014] Secondly, this application provides an intelligent control system for soft electroplating processing, comprising: The data acquisition module 101 is used to acquire current operating condition information and real-time monitoring data of the process tank. The monitoring data includes the pH value, conductivity, temperature and redox potential (ORP) of the liquid in the process tank. The initial parameter matching module 102 is used to extract key feature parameters that characterize the state of the bath liquid based on real-time monitoring data, and to obtain the matched initial process parameters by matching the current operating condition information and key feature parameters through a preset historical operating condition database. The data dynamic monitoring module 103 is used to start processing according to the preset process flow by using the initial process parameters as the current process parameter setting value. During the processing, it periodically acquires the monitoring data in the process tank and updates the key feature parameters according to the monitoring data. The parameter dynamic correction module 104 is used to generate optimal process parameters based on the updated key feature parameters and guided by the preset quality target through a preset adaptive process model. The parameter adjustment execution module 105 is used to generate process parameter adjustment instructions using the optimal process parameters as the current process parameter setpoints.

[0015] Thirdly, this application provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed as described above for an intelligent control method for electroplating soft processing.

[0016] In summary, firstly, by employing historical matching and real-time optimization of process parameters, it is ensured that each batch of workpieces and each tank receives personalized and precise process parameters, effectively eliminating quality fluctuations caused by human experience and changes in the tank solution state. Secondly, during the dynamic optimization of process parameters, in addition to being guided by quality objectives, process costs are also taken into consideration. This allows for the simultaneous optimization of energy and material consumption while ensuring quality standards are met, minimizing the process cost per batch. Furthermore, through the automatic transmission and correction of feedforward process information between processing tanks, subsequent processes can adapt to the actual processing results of preceding processes, upgrading isolated single-tank control to integrated line optimization, further improving the quality consistency and operational stability of the electroplating soft treatment process. Attached Figure Description

[0017] Figure 1 This is a flowchart of an intelligent control method for soft electroplating provided in an embodiment of this application; Figure 2 This is a flowchart provided in this application embodiment, which obtains the matched initial process parameters by matching based on the current working condition information and key feature parameters through a preset historical working condition database; Figure 3 This is a flowchart provided in this application embodiment that generates optimal process parameters based on updated key feature parameters, guided by a preset quality target, and through a preset adaptive process model. Figure 4 This is a flowchart provided in an embodiment of the present application, which generates optimal process parameters based on quality deviation using a preset optimization algorithm. Figure 5 This is a schematic diagram of an intelligent control system for soft electroplating provided in an embodiment of this application. Detailed Implementation

[0018] The following is in conjunction with the appendix Figure 1 -Appendix Figure 5 This application will be described in further detail below.

[0019] This application provides an intelligent control method for soft electroplating treatment; see [link to relevant documentation]. Figure 1 This includes the following steps: S100: Obtain current operating condition information and real-time monitoring data of the process tank.

[0020] S200: Based on real-time monitoring data, extract key characteristic parameters to characterize the state of the bath solution.

[0021] S300: Based on current operating condition information and key characteristic parameters, it matches the initial process parameters by using a preset historical operating condition database.

[0022] S400: Using the initial process parameters as the current process parameter settings, start the process according to the preset process flow. During the process, periodically acquire the monitoring data in the process tank and update the key characteristic parameters based on the monitoring data.

[0023] S500, based on updated key feature parameters and guided by preset quality objectives, generates optimal process parameters through a preset adaptive process model.

[0024] S600: Using the optimal process parameters as the current process parameter setpoints, generate process parameter adjustment instructions.

[0025] In this embodiment of the application, the current working condition information and real-time monitoring data of the process tank are first obtained. The current working condition information refers to the external production variables that affect the result of this soft treatment process, other than the state of the tank liquid itself, including the type of workpiece to be treated, the number of workpieces, the initial state of the workpiece surface, and the cumulative generated load.

[0026] Real-time monitoring data refers to data obtained in real time by sensors before the start of the process, reflecting the state of the tank solution when it is idle, and is used for initial decision-making. Specifically, the monitoring data includes the pH value, conductivity, temperature and redox potential (ORP) of the tank solution.

[0027] The process tank includes multiple processing tanks connected in series, namely a degreasing tank, a pickling tank, and a passivation tank. The three tanks are physically isolated and independent. The workpiece moves between the tanks, and a water washing tank is set between the processing tanks for physical cleaning.

[0028] After acquiring real-time monitoring data, key characteristic parameters for characterizing the state of the bath solution will be extracted based on the real-time monitoring data.

[0029] Key characteristic parameters include effective concentration of the bath solution and pollution load index. Effective concentration of the bath solution represents the real-time estimated concentration of the main chemical components (such as alkalinity, acidity, and active ingredients) in the bath solution. Pollution load index is used to reflect the degree of negative impact of pollutants (such as oil, metal ions, and impurities) accumulated in the bath solution on the treatment process.

[0030] It is important to note that the key characteristic parameters here are a general term for the three types of processes. The specific parameters for each treatment tank are different. For example, the key characteristic parameters for the degreasing tank are effective alkalinity and oil load index; the key characteristic parameters for the pickling tank are free acid concentration and metal ion contamination index; and the key characteristic parameters for the passivation tank are effective film-forming agent concentration and impurity interference factor.

[0031] Specifically, based on real-time monitoring data, key characteristic parameters for characterizing the state of the bath solution are extracted, including the following steps: S210. Preprocess the real-time monitoring data, and align and merge the preprocessed real-time monitoring data in the time dimension to form a unified time-series dataset of tank liquid status.

[0032] S220: Based on the state time series dataset, key feature parameters are obtained through a preset calculation method.

[0033] First, the real-time monitoring data will be preprocessed, such as data cleaning and filtering to remove abnormal data. Then, the real-time monitoring data from different sensors will be aligned and synchronized on the timestamp to form a unified time-series dataset of tank liquid status.

[0034] Next, based on the time series dataset of the bath liquid state, key feature parameters are obtained through a preset method. Here, the preset method refers to a pre-defined physicochemical model for soft processing.

[0035] For example, to calculate the effective alkalinity of the degreasing tank, the measured conductivity is compensated to a uniform standard temperature (usually 25°C), and then substituted into a pre-calibrated alkalinity-conductivity standard curve to directly read and estimate the alkalinity. For the pollution load index, to calculate the oil pollution load index of the degreasing tank, the conductivity of the standard tank solution (newly prepared tank solution) is used as a benchmark. The deviation of the current conductivity from the benchmark is calculated, and then a quantitative value is output through nonlinear normalization.

[0036] After determining the key characteristic parameters, the initial process parameters can be obtained by matching the current operating conditions and key characteristic parameters with a preset historical operating condition database. Process parameters refer to the operational variables that can be adjusted by the control system to change the process and results, including temperature, processing time and the amount of reagent added, and the acceleration rate.

[0037] Specifically, see Figure 2 Based on current operating condition information and key characteristic parameters, the system matches the data with a pre-set historical operating condition database to obtain the matching initial process parameters, including the following steps: S310. Encode the current operating condition information and key feature parameters to generate an operating condition feature vector.

[0038] S320. Based on the working condition feature vector, similarity matching is performed through a preset historical working condition database to obtain the historical working condition record with the highest similarity and the corresponding similarity.

[0039] S330. If the similarity is higher than the preset first similarity threshold, the historical process parameters corresponding to the historical working conditions with the highest similarity will be recorded as the initial process parameters for matching.

[0040] S340. If the similarity is lower than the preset first similarity threshold but higher than the preset second similarity threshold, then generate a valid fusion set for all historical work condition records with similarity higher than the preset second similarity threshold.

[0041] S350. Based on the effective fusion set, the historical process parameters corresponding to each historical working condition record are weighted and fused to generate the initial process parameters, using the similarity of each historical working condition record as the weight.

[0042] S360. If the similarity is lower than the preset second similarity threshold, then the preset general safety process parameters are used as the initial process parameters.

[0043] First, the current operating condition information and key feature parameters are encoded into a multi-dimensional feature vector, denoted as the operating condition feature vector.

[0044] Then, based on the operating condition feature vector, similarity matching is performed through a preset historical operating condition database to obtain the historical operating condition record with the highest similarity and the corresponding similarity. The preset historical operating condition database stores multiple historical operating condition records. Each record is associated with at least a historical operating condition feature vector composed of historical operating condition information and historical key feature parameters, as well as the historical process parameters used to achieve the quality target under that historical operating condition.

[0045] Next, the similarity is compared with a preset first similarity threshold. If the similarity is higher than the preset first similarity threshold, the historical process parameters corresponding to the historical operating condition with the highest similarity are recorded as the initial process parameters for matching. The preset first similarity threshold is the lower confidence limit for determining that the current operating condition and historical experience are highly similar and can be directly reused. Its value is determined based on the distribution of historical data and process sensitivity analysis, and is usually set in the high confidence interval between 0.80 and 0.90.

[0046] If the similarity is lower than the preset first similarity threshold but higher than the preset second similarity threshold, then a valid fusion set is generated for all historical work condition records with similarity higher than the preset second similarity threshold. The preset second similarity threshold is the lower confidence limit for determining that the current work condition has a certain similarity with historical experience and is therefore eligible to participate in the weighted fusion to generate suggested parameters. Its value is usually set in an acceptable confidence range between 0.50 and 0.75.

[0047] Once the effective fusion set is determined, the historical process parameters corresponding to each historical working condition record can be weighted and fused based on the effective fusion set, using the similarity of each historical working condition record as the weight, to generate the initial process parameters.

[0048] Once the initial process parameters are determined, they can be used as the current process parameter settings, and the process can be started according to the preset process flow.

[0049] Although setting the initial process parameters is equivalent to establishing a basic plan for the entire electroplating soft treatment process, ensuring that each treatment tank has a scientific and safe starting point, in the actual treatment process, it is still necessary to dynamically adjust the process parameters according to the real-time changes in the state of the tank solution in order to cope with disturbances, compensate for deviations, and ensure that the final quality target is accurately achieved.

[0050] Therefore, during the processing, monitoring data from the process tank will be acquired periodically, and key characteristic parameters will be updated based on the monitoring data. The monitoring data here is consistent with the real-time monitoring data mentioned earlier, only the timing of acquisition is different; the former is acquired during the processing, while the latter is acquired before the processing.

[0051] Then, based on the updated key feature parameters and guided by the preset quality target, the optimal process parameters are generated through a preset adaptive process model.

[0052] Specifically, see Figure 3 Based on the updated key feature parameters and guided by the preset quality objectives, the optimal process parameters are generated through a preset adaptive process model, including the following steps: S510. Based on the current operating condition information and the updated key feature parameters, generate the working state vector.

[0053] S520. Based on the working state vector and the current process parameter settings, the quality prediction value of the current processing result is generated through a preset adaptive process model.

[0054] S530. Based on the predicted quality value, calculate the quality deviation by setting a preset quality target.

[0055] S540. Determine whether the quality deviation is lower than the preset threshold.

[0056] S550 If so, the current process parameter setting value shall be taken as the optimal process parameter.

[0057] S560. If not, then based on the quality deviation, the optimal process parameters are generated through a preset optimization algorithm.

[0058] First, based on the key feature parameters after the minimum update and combined with the current operating condition information, a working state vector is generated to represent the real-time comprehensive status of the current processing slot.

[0059] Then, the working state vector and the current process parameter settings are input into the preset adaptive process model to generate the quality prediction value of the processing result.

[0060] The preset adaptive process model here is generated based on deep learning (such as neural networks and reinforcement learning), which can predict the final quality result of the workpiece if it continues to be processed according to the current process parameters in the current state.

[0061] By comparing the predicted quality values ​​with the preset quality targets, the quality deviation is calculated. Here, the preset quality targets are used to characterize the quantifiable expected standards for the output results of the soft treatment process, such as cleanliness, surface activation, film thickness, etc.

[0062] Finally, it is determined whether the quality deviation is lower than the preset threshold. If the quality deviation is lower than the preset threshold, it means that the deviation is within the acceptable range. Then, the current process parameter setting value is taken as the optimal process parameter, that is, the current process parameter is maintained.

[0063] If the quality deviation is not lower than the preset threshold, it means that the deviation exceeds the acceptable range. Therefore, based on the quality deviation, the optimal process parameters are generated through the preset optimization algorithm.

[0064] Specifically, see Figure 4 Based on the quality deviation, the optimal process parameters are generated using a preset optimization algorithm, including the following steps: S561. Based on the current process parameter settings and quality deviations, generate a set of candidate process parameters.

[0065] S562. Traverse the set of candidate process parameters. For each candidate process parameter, output the corresponding quality prediction result through a preset adaptive process model, and obtain the cost item through a preset process cost calculation method.

[0066] S563. Based on the quality prediction results and cost items, a comprehensive evaluation score is generated using preset weighting coefficients.

[0067] S564. After the traversal is completed, the candidate process parameter with the smallest comprehensive evaluation score is recorded as the optimal process parameter.

[0068] First, a set of candidate process parameters is generated based on the current process parameter settings and quality deviation. That is, based on the quality deviation, the direction and range of parameter adjustment are determined according to the direction and magnitude of the quality deviation. If the predicted quality value is less than the preset quality target, it is a negative deviation and the adjustment direction is to enhance it. If the predicted quality value is greater than the preset quality target, it is a positive deviation and the adjustment direction is to weaken it.

[0069] Adjustable range is denoted as , , It can be represented as: in, Set the current process parameters to their current values. This represents the absolute value of the quality deviation. The process sensitivity coefficient corresponding to the parameter represents the parameter adjustment required per unit mass deviation, and can be calibrated through process experiments or historical data. The maximum adjustment amplitude coefficient is a preset constant (0 < 0). ≤1), used to limit the maximum amplitude of a single adjustment.

[0070] After adjusting the direction, the adjustment range is: ( >0), or ( <0).

[0071] After determining the adjustment range, the set of candidate process parameters can be determined according to the preset step size. For example, if >0, then the candidate process parameter set Next, the set of candidate process parameters is traversed. For each candidate process parameter, the corresponding quality prediction result is output through a preset adaptive process model, and the cost item is obtained through a preset process cost calculation method.

[0072] The preset process cost calculation method is a method for quantifying the resource costs required to execute the candidate parameters. It includes energy cost, time cost, and variable cost. Energy cost can be estimated by combining the temperature setting and processing time with the heater power model. Time cost is the absolute value of the processing time, reflecting the impact on production capacity. Variable cost is the difference between the candidate process parameters and the current process parameters, used to penalize excessively large adjustment actions and ensure control stability.

[0073] Then, based on the quality prediction results and cost items, a comprehensive evaluation score is generated using preset weighting coefficients.

[0074] The quality prediction result is denoted as Preset quality targets, denoted as The cost item is , The comprehensive cost value is a weighted average of all cost items, denoted by S. S can be expressed as: in, and These are the weighting coefficients. + =1, used to balance the two core objectives of quality compliance and cost saving.

[0075] Finally, after the traversal is complete, the candidate process parameter with the lowest comprehensive evaluation score is recorded as the optimal process parameter.

[0076] Once the optimal process parameters are determined, they can be used as the current process parameter settings to generate process parameter adjustment commands, which are then sent to the actuators associated with the current processing tank to perform the adjustment actions, so that the actual process parameters approach the optimal process parameter settings.

[0077] As mentioned earlier, the process tank consists of multiple processing tanks connected in series. Although each processing tank is physically independent, there are corresponding physical and chemical relationships between them. Therefore, the process treatment of each processing tank cannot be completely controlled independently and requires coordinated adjustment.

[0078] The core of coordinated adjustment lies in assessing whether the processing results of preceding processing slots will affect the processing of subsequent processing slots, thereby enabling targeted coordinated optimization. For example, under independent control, the pickling tank is unaware of the actual degreasing effect of the degreasing tank. If a batch of workpieces is not thoroughly degreased (heavy oil residue remains), and the pickling tank is still set with parameters according to the standard clean workpiece model, it may result in insufficient pickling.

[0079] Therefore, in this embodiment, two adjacent processing slots are respectively designated as the preceding processing slot and the following processing slot according to a preset procedure. During the processing, before starting processing for any subsequent processing slot, the following steps are also included: S710. Obtain the final process parameters used in the preceding treatment tank and the key characteristic parameters of the tank liquid when the treatment is completed, and record them together as feedforward process information.

[0080] S720. Add the feedforward process information to the current operating condition information as the updated current operating condition information.

[0081] S730. Based on the updated current operating condition information, adjust the initial process parameters corresponding to the subsequent processing tank to generate the adjusted process parameters.

[0082] First, obtain the final process parameters used in the pretreatment tank, which are the optimal process parameters generated during the dynamic optimization process, as well as the associated characteristic parameters of the tank liquid when the treatment is completed. That is, after the treatment is completed, monitoring data characterizing the state of the tank liquid will also be obtained. Then, based on the monitoring data, key characteristic parameters are extracted, and the final process parameters used are combined with the key characteristic parameters and recorded together as feedforward process information.

[0083] Then, the feedforward process information is added to the current operating condition information to serve as the updated current operating condition information. In this way, the current operating condition information not only includes its inherent static attribute information, but also includes the feedforward process information as dynamic processing data accumulated with the processing flow.

[0084] Finally, based on the updated current operating information, the initial process parameters corresponding to the subsequent processing tank can be adjusted to generate the adjusted process parameters.

[0085] Specifically, based on the updated current operating condition information, the initial process parameters corresponding to the subsequent processing tank are adjusted to generate the adjusted process parameters, including the following steps: S731. Compare the updated current operating condition information with the initial operating condition information before processing begins to obtain the status feature increment.

[0086] S732. Based on the incremental state features, obtain the corresponding standardized adjustment scheme through a preset incremental-parameter correction mapping table.

[0087] S733. Based on the standard adjustment scheme, adjust the initial process parameters corresponding to the subsequent processing tank to generate the adjusted process parameters.

[0088] As mentioned earlier, the initial process parameters are set based on the current operating conditions and the key characteristic parameters of each processing tank. However, now, considering the processing status of the preceding processing tanks, the current operating conditions also include feedforward process information. Therefore, the previously set initial process parameters can be fine-tuned based on the feedforward process information.

[0089] First, the updated current operating condition information is compared with the initial operating condition information before processing begins to obtain the status feature increment.

[0090] The reason for making a comparison here, instead of directly obtaining the state feature increment based on the feedforward process information, is that the preceding processing slot and the following processing slot are not unique, and all the feedforward process information obtained previously may affect the following processing slot. It should also be noted that the initial operating condition information before the start of processing here is operating condition information that only includes its static attribute information.

[0091] State feature increments are used to quantify the cumulative changes in a specific process dimension of the batch of workpieces after being processed by the pre-processing tank, including cumulative increments in cleanliness, cumulative increments in metal etching, and cumulative increments in foreign matter introduction.

[0092] The cumulative increase in cleanliness can be estimated based on the actual parameters and effects of the preceding degreasing tank; the cumulative increase in metal etching can be estimated based on the actual parameters and effects of the preceding pickling tank; and the cumulative increase in foreign matter introduction can be estimated based on the changes in the contamination index of the preceding tank solution.

[0093] Then, based on the state feature increment, the corresponding standardized adjustment scheme is obtained through a preset increment-parameter correction mapping table.

[0094] Among them, the preset addition-parameter correction mapping table is a two-dimensional structured knowledge base that uses state feature increments as indexes and standardized parameter adjustment schemes as outputs.

[0095] The standardized parameter adjustment scheme is a structured output object of the mapping table query results, which includes a target parameter set and an adjustment scheme set. The target parameter set is the set of process parameters that need to be adjusted, such as {processing time, film-forming agent concentration}. The adjustment scheme set is the adjustment method and adjustment value corresponding to the parameters to be adjusted, such as {proportional coefficient: 0.92, proportional coefficient: 1.05}.

[0096] Finally, based on the standard adjustment scheme, the initial process parameters corresponding to the subsequent processing tank can be adjusted to generate the adjusted process parameters.

[0097] In this way, by introducing coordinated adjustment between treatment tanks, the processing results of the preceding treatment tank can assist in guiding the parameter setting of the subsequent treatment tank, thereby achieving linkage optimization of the entire electroplating soft treatment process.

[0098] In addition, considering that a water washing tank is also provided between the pre-treatment tank and the post-treatment tank, in addition to adjusting the process parameters of the post-treatment tank in accordance with the above-mentioned coordinated adjustment, key process parameters of the water washing tank can also be added. The key process parameters of the water washing tank include water washing temperature, conductivity and / or pH value.

[0099] Therefore, the initial process parameters of the subsequent treatment tank can be modified based on the key process parameters of the water washing tank. The logic of this modification is that the worse the water washing quality (the more residue), the greater the modification force of the subsequent treatment tank. For example, taking the pickling tank as an example, the conductivity of the water washing tank reflects the amount of alkaline residue on the workpiece surface, so the initial concentration of pickling can be modified.

[0100] In addition to ensuring that process parameters match real-time operating conditions, it is also necessary to pay attention to risk perception during the electroplating soft treatment process, that is, to detect abnormalities in advance and intervene in a timely manner in order to extend the service life of the plating solution.

[0101] Therefore, in this embodiment of the application, after periodically acquiring monitoring data from the process tank and updating key feature parameters based on the monitoring data, the following steps are also included: S810. Calculate process health indicators based on key characteristic parameters.

[0102] S820: Based on the current operating condition information, a dynamic safety threshold is generated for the process health index through a preset dynamic threshold model.

[0103] S830. Compare the process health indicators with the corresponding dynamic safety thresholds. If any indicator exceeds its corresponding dynamic safety threshold, output an abnormality prompt message.

[0104] First, based on key characteristic parameters, process health indicators are calculated, including component stability index and contamination accumulation rate.

[0105] The component stability index is calculated based on the degree of fluctuation (standard deviation) of selected key characteristic parameters within a sliding time window; the contamination accumulation rate is calculated based on the rate of change per unit time of key characteristic parameters (such as the contamination load index) characterizing the degree of contamination during the current batch processing.

[0106] Then, based on the current operating condition information, a corresponding dynamic safety threshold is generated for the process health index through a preset dynamic threshold model. The preset dynamic threshold model is an algorithm that receives the operating condition information as input and outputs the dynamic safety threshold. This algorithm can be an empirical formula that regresses historical technical data.

[0107] The reason for setting dynamic safety thresholds is to take into account that the actual situation is different under different working conditions. For example, under the current working condition, the workpiece is heavily contaminated with oil and the load is large, so it is necessary to dynamically increase the warning threshold of "contamination accumulation rate" to avoid false alarms.

[0108] In addition, the dynamic safety threshold can be adaptively adjusted according to the cumulative production load of the treatment tank solution. For example, for the contamination accumulation rate, the dynamic safety threshold is positively correlated with the tank solution life. The longer the tank solution life, the higher the rate of contamination increase when treating the same load. To avoid false alarms, the dynamic threshold is adjusted accordingly. For the component stability index, the dynamic safety threshold is negatively correlated with the tank solution life. The longer the tank solution life, the lower its buffering capacity and the lower its tolerance to fluctuations. To detect the risk of stability degradation in a timely manner, the dynamic threshold is adjusted accordingly.

[0109] Finally, the process health indicators are compared with the corresponding dynamic safety thresholds. If any indicator exceeds its corresponding dynamic safety threshold, an abnormality prompt message is output to prompt relevant personnel to intervene and handle the situation in a timely manner.

[0110] This application also provides an intelligent control system for soft electroplating processing. See [link to relevant documentation] Figure 5 The system includes: a data acquisition module 101, an initial parameter matching module 102, a data dynamic monitoring module 103, a parameter dynamic correction module 104, and a parameter adjustment execution module 105.

[0111] The data acquisition module 101 is used to acquire current operating condition information and real-time monitoring data of the process tank.

[0112] The initial parameter matching module 102 is used to extract key feature parameters that characterize the state of the bath liquid based on real-time monitoring data, and to obtain the matched initial process parameters by matching the current operating condition information and key feature parameters through a preset historical operating condition database.

[0113] The data dynamic monitoring module 103 is used to start processing according to the preset process flow by using the initial process parameters as the current process parameter setting value. During the processing, it periodically acquires the monitoring data in the process tank and updates the key characteristic parameters based on the monitoring data.

[0114] The parameter dynamic correction module 104 is used to generate optimal process parameters based on the updated key feature parameters and guided by the preset quality target through a preset adaptive process model.

[0115] The parameter adjustment execution module 105 is used to generate process parameter adjustment instructions using the optimal process parameters as the current process parameter setpoints.

[0116] In this embodiment of the application, the data acquisition module 101 is specifically used to acquire current operating condition information and real-time monitoring data of the process tank, wherein the monitoring data includes the pH value, conductivity, temperature and redox potential (ORP) of the liquid in the process tank.

[0117] The initial parameter matching module 102 is specifically used to extract key feature parameters for characterizing the state of the bath liquid based on the real-time monitoring data obtained by the data acquisition module 101, and to obtain the matched initial process parameters based on the current operating condition information and key feature parameters by matching them through a preset historical operating condition database.

[0118] The data dynamic monitoring module 103 is specifically used to start processing according to the preset process flow by using the initial process parameters obtained by the initial parameter matching module 102 as the current process parameter setting value. During the processing, it periodically acquires the monitoring data in the process tank and updates the key feature parameters according to the monitoring data.

[0119] The parameter dynamic correction module 104 is specifically used to generate optimal process parameters based on the updated key feature parameters obtained by the data dynamic monitoring module 103, guided by the preset quality target, through a preset adaptive process model.

[0120] The parameter adjustment execution module 105 is specifically used to generate a process parameter adjustment command by using the optimal process parameters generated by the parameter dynamic correction module 104 as the current process parameter setting value.

[0121] This application also provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed by any of the above-described intelligent control methods for electroplating soft processing.

[0122] The embodiments described in this application are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the principles of this application should be included within the scope of protection of this application.

Claims

1. An intelligent control method for soft electroplating treatment, characterized in that, include: Acquire current operating condition information and real-time monitoring data of the process tank, including pH value, conductivity, temperature and redox potential (ORP) of the solution in the process tank; Based on real-time monitoring data, key characteristic parameters for characterizing the state of the bath solution are extracted. Based on the current operating condition information and key characteristic parameters, the initial process parameters are obtained by matching through a preset historical operating condition database. Using the initial process parameters as the current process parameter settings, the process is started according to the preset process flow. During the process, the monitoring data in the process tank is acquired periodically, and the key characteristic parameters are updated based on the monitoring data. Based on the updated key feature parameters and guided by the preset quality target, the optimal process parameters are generated through a preset adaptive process model. The optimal process parameters are used as the current process parameter settings to generate process parameter adjustment instructions.

2. The intelligent control method for soft electroplating treatment according to claim 1, characterized in that, The key characteristic parameters for characterizing the state of the bath solution, extracted based on real-time monitoring data, include: The real-time monitoring data is preprocessed, and the preprocessed real-time monitoring data is aligned and fused in the time dimension to form a unified time series dataset of tank liquid status. Based on the state time series dataset, key feature parameters are obtained through a pre-defined calculation method.

3. The intelligent control method for soft electroplating treatment according to claim 1, characterized in that, The process of obtaining initial process parameters based on current operating condition information and key characteristic parameters by matching them against a preset historical operating condition database includes: The current operating condition information and key feature parameters are encoded to generate an operating condition feature vector; Based on the working condition feature vector, similarity matching is performed through a preset historical working condition database to obtain the historical working condition record with the highest similarity and the corresponding similarity. If the similarity is higher than the preset first similarity threshold, the historical working condition with the highest similarity will be recorded as the corresponding historical process parameters and used as the initial process parameters for matching. If the similarity is lower than the preset first similarity threshold but higher than the preset second similarity threshold, then a valid fusion set is generated for all historical work condition records with similarity higher than the preset second similarity threshold. Based on the effective fusion set, the historical process parameters corresponding to each historical working condition record are weighted and fused to generate the initial process parameters, using the similarity of each historical working condition record as the weight. If the similarity is lower than the preset second similarity threshold, then the preset general safety process parameters are used as the initial process parameters.

4. The intelligent control method for soft electroplating treatment according to claim 1, characterized in that, The process of generating optimal process parameters based on updated key feature parameters, guided by a preset quality target, and through a preset adaptive process model, includes: Based on the current operating condition information and the updated key feature parameters, a working state vector is generated, which is used to characterize the real-time comprehensive status of the current process tank. Based on the working state vector and the current process parameter settings, the quality prediction value of the processing result is generated through a preset adaptive process model. Based on the predicted quality values, the quality deviation is calculated by setting pre-defined quality targets; Determine whether the quality deviation is below a preset threshold; If so, the current process parameter setting value will be taken as the optimal process parameter; If not, then based on the quality deviation, the optimal process parameters are generated through a preset optimization algorithm.

5. The intelligent control method for soft electroplating treatment according to claim 4, characterized in that, The step of generating optimal process parameters based on quality deviation using a preset optimization algorithm includes: Based on the current process parameter settings and quality deviations, a set of candidate process parameters is generated; Traverse the set of candidate process parameters, and for each candidate process parameter, output the corresponding quality prediction result through a preset adaptive process model, and obtain the cost item through a preset process cost calculation method; Based on the quality prediction results and cost items, a comprehensive evaluation score is generated using preset weighting coefficients; After the traversal is complete, the candidate process parameter with the lowest comprehensive evaluation score is recorded as the optimal process parameter.

6. The intelligent control method for soft electroplating treatment according to claim 1, characterized in that, The process tank includes multiple processing tanks connected in series. Adjacent processing tanks are designated as a preceding processing tank and a succeeding processing tank according to a preset procedure. During the processing, before starting processing in any succeeding processing tank, the following steps are also included: Obtain the final process parameters used in the pretreatment tank and the key characteristic parameters of the tank liquid when the treatment is completed, and record them together as feedforward process information; Add the feedforward process information to the current operating condition information to serve as the updated current operating condition information; Based on the updated current operating condition information, the initial process parameters corresponding to the subsequent processing tank are adjusted to generate the adjusted process parameters.

7. The intelligent control method for soft electroplating treatment according to claim 6, characterized in that, The step of adjusting the initial process parameters corresponding to the subsequent processing tank based on the updated current operating condition information to generate adjusted process parameters includes: The updated current operating condition information is compared with the initial operating condition information before processing begins to obtain the status feature increment; Based on the incremental state features, the corresponding standardized adjustment scheme is obtained through a preset incremental-parameter correction mapping table; Based on the standard adjustment scheme, the initial process parameters corresponding to the subsequent processing tank are adjusted to generate the adjusted process parameters.

8. The intelligent control method for soft electroplating treatment according to claim 1, characterized in that, The process, after periodically acquiring monitoring data from the process tank and updating key characteristic parameters based on the monitoring data, also includes: Based on key characteristic parameters, process health indicators are calculated, including component stability index and cumulative pollution rate. Based on the current operating condition information, a dynamic safety threshold is generated for the process health index through a preset dynamic threshold model. The process health indicators are compared with the corresponding dynamic safety thresholds. If any indicator exceeds its corresponding dynamic safety threshold, an abnormal prompt message is output.

9. An intelligent control system for soft electroplating treatment, characterized in that, include: The data acquisition module (101) is used to acquire current operating condition information and real-time monitoring data of the process tank. The monitoring data includes the pH value, conductivity, temperature and redox potential (ORP) of the liquid in the process tank. The initial parameter matching module (102) is used to extract key feature parameters for characterizing the state of the bath liquid based on real-time monitoring data, and to obtain the matched initial process parameters by matching the current working condition information and key feature parameters through a preset historical working condition database. The data dynamic monitoring module (103) is used to start processing according to the preset process flow with the initial process parameters as the current process parameter setting value. During the processing, it periodically acquires the monitoring data in the process tank and updates the key feature parameters according to the monitoring data. The parameter dynamic correction module (104) is used to generate the optimal process parameters based on the updated key feature parameters and guided by the preset quality target through the preset adaptive process model. The parameter adjustment execution module (105) is used to generate process parameter adjustment instructions using the optimal process parameters as the current process parameter setting values.

10. A computer-readable storage medium storing a computer program capable of being loaded by a processor and executing a smart control method for electroplating soft processing as described in any one of claims 1 to 8.

Images