A sixteen-line four-roller electroplating line production control method

Abstract

The application discloses a kind of sixteen-line four-roller electroplating line production control methods, it is related to electroplating production line control technical field, including, in all production line deployment edge node, and utilize sensor real-time acquisition electroplating liquid state data;According to electroplating liquid state data, utilize machine learning algorithm, to the state data anomaly of electroplating liquid is identified;Utilize visual sensor and infrared perception equipment acquisition electroplating line visual image data and infrared image data;Based on visual image data and infrared image data, through image processing algorithm, quality evaluation is carried out to electroplating line;According to the abnormal state of electroplating liquid and the quality evaluation result of electroplating line, through adaptive algorithm, electroplating liquid state parameter is adjusted;The application is by comprehensively considering electroplating liquid state and electroplating line quality evaluation result, significantly improves the consistency and stability of electroplating quality.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electroplating production line control, and particularly relates to a sixteen-line four-drum electroplating line production control method. BACKGROUND

[0002] A traditional electroplating production line usually adopts a single electroplating tank and a small number of drums, and this configuration is relatively efficient in small-scale production, but in large-scale production, especially in a sixteen-line four-drum electroplating line in which multiple production lines work in parallel, there is a significant technical bottleneck.

[0003] In a sixteen-line four-drum electroplating production line, it is a great challenge in the current electroplating technical field to ensure that the electroplating quality of each line is optimal and the state of the electroplating solution can be adjusted in real time. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a sixteen-line four-drum electroplating line production control method to solve the problem of real-time adjustment of the state of the electroplating solution.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a sixteen-line four-drum electroplating line production control method, which comprises,

[0008] deploying edge nodes on all production lines and collecting electroplating solution state data in real time using sensors;

[0009] According to the electroplating solution state data, the state data anomaly of the electroplating solution is identified using a machine learning algorithm;

[0010] Collecting visual image data and infrared image data of the electroplating line using a visual sensor and an infrared sensing device;

[0011] Based on the visual image data and the infrared image data, the quality of the electroplating line is evaluated through an image processing algorithm;

[0012] According to the abnormal state of the electroplating solution and the quality evaluation result of the electroplating line, the electroplating solution state parameters are adjusted through an adaptive algorithm.

[0013] As a preferred scheme of the sixteen-line four-drum electroplating line production control method, wherein the step of deploying edge nodes on all production lines and collecting electroplating solution state data in real time using sensors comprises the following steps,

[0014] Collecting the state data of the electroplating solution on all production lines in real time using the sensors on the edge nodes;

[0015] Smooth and standardize the real-time collected electroplating solution state data on the edge node.

[0016] As a preferred scheme of the sixteen-line four-roller electroplating line production control method, wherein: according to the electroplating solution state data, the state data anomaly of the electroplating solution is identified by using a machine learning algorithm, and the specific steps are as follows,

[0017] According to the electroplating solution state data, a state data set is constructed;

[0018] According to the range of the state data of the electroplating solution set on the production line, the data outside the range of the state data set is labeled as abnormal, and the data within the range of the state data set is labeled as normal;

[0019] Mix the state data with abnormal labels and normal labels, and divide them into training set and test set;

[0020] Based on the nonlinear support vector machine (SVM), an electroplating solution state judgment model is constructed on the edge node;

[0021] The training set is input into the electroplating solution state judgment model, and the target loss function is calculated;

[0022] The training set is divided into k subsets, and k-fold cross-validation is used, one subset is used as the validation set and the rest is used as the training set, and training and validation are performed;

[0023] According to the range setting of the state data of the electroplating solution on the production line, the accuracy, precision and recall rate thresholds are preset;

[0024] After each training and validation, the accuracy, precision and recall rate are calculated, and when the accuracy, precision and recall rate are all greater than the accuracy, precision and recall rate thresholds, the training is completed;

[0025] The test set is input into the electroplating solution state judgment model, and the accuracy, precision and recall rate are calculated, and when the accuracy, precision and recall rate are all greater than the accuracy, precision and recall rate thresholds, the electroplating solution state judgment model training is completed;

[0026] The incremental learning mechanism is integrated into the electroplating solution state judgment model, and the electroplating solution state judgment model is updated in real time;

[0027] The real-time collected electroplating solution state data is used to judge the state of the electroplating solution on the production line by using the electroplating solution state judgment model on the edge node.

[0028] As a preferred scheme of the sixteen-line four-roller electroplating line production control method, the incremental learning mechanism is integrated into the electroplating solution state judgment model, and the electroplating solution state judgment model is updated in real time, and the specific steps are as follows,

[0029] The influence degree of the newly collected state data on the electroplating solution state judgment model is judged by calculating the distance between the newly collected state data and the hyperplane of the electroplating solution state judgment model.

[0030] The newly collected state data is added to the support vector set of the electroplating solution state judgment model, and the hyperplane and the decision function are updated according to the new support vector set.

[0031] As a preferred scheme of the sixteen-line four-roller electroplating line production control method, the incremental learning mechanism is integrated into the electroplating solution state judgment model, and the electroplating solution state judgment model is updated in real time, and the specific steps are as follows,

[0032] The visual image data of the electroplating line is collected by using a visual sensor;

[0033] The infrared image data of the electroplating line is collected by using an infrared sensing device;

[0034] The visual image data and the infrared image data are processed.

[0035] As a preferred scheme of the sixteen-line four-roller electroplating line production control method, the incremental learning mechanism is integrated into the electroplating solution state judgment model, and the electroplating solution state judgment model is updated in real time, and the specific steps are as follows,

[0036] The visual image data is denoised by using a bilateral filtering method;

[0037] The contrast of the image is improved by adjusting the brightness distribution of the image by using a histogram equalization method;

[0038] The brightness and contrast of the infrared image are adjusted by weighted averaging similar pixels in the image while removing noise and preserving details by using a non-local mean method.

[0039] As a preferred scheme of the sixteen-line four-roller electroplating line production control method, the incremental learning mechanism is integrated into the electroplating solution state judgment model, and the electroplating solution state judgment model is updated in real time, and the specific steps are as follows,

[0040] The thickness of the electroplated layer is calculated according to the relationship between the infrared radiation intensity of the infrared sensing device and the thickness of the electroplated layer;

[0041] The electroplated layer region of the electroplating line is extracted by an image segmentation method, and the standard deviation of the thickness of the electroplated layer in each region is calculated as an uniformity evaluation index;

[0042] Classify and label known defects in visual image data (including normal, crack, bubble, peel);

[0043] Label multiple defect categories for each image and provide bounding boxes for each defect region;

[0044] Based on convolutional neural network (CNN), build a defect recognition model;

[0045] Divide the visual image data classified and labeled by known defects into training set and validation set, and input the training set into the defect recognition model;

[0046] Train the input defect recognition model by calculating the cross-entropy loss function of the visual image classified and labeled by known defects in the training set;

[0047] Minimize the loss function using gradient descent method and iteratively optimize the network weights, and after each training, evaluate the model performance through the validation set and adjust the hyperparameters;

[0048] Evaluate the performance of the model by calculating the accuracy and recall rate of the defect recognition model;

[0049] Use the trained defect recognition model to detect defects in newly collected visual image data;

[0050] Evaluate the overall quality of the electroplated layer by considering the thickness, uniformity and surface defects of the electroplated layer.

[0051] As a preferred scheme of the sixteen-line four-roller electroplating line production control method described in the present application, the specific steps of adjusting the electroplating liquid state parameters through the self-adaptive algorithm according to the abnormal state of the electroplating liquid and the quality evaluation result of the electroplating line are as follows,

[0052] Normalize the abnormal data of the electroplating liquid and the quality evaluation result of the electroplating line;

[0053] Adopt a self-adaptive control algorithm based on reinforcement learning to guide the adjustment of the electroplating liquid state data through a reward function;

[0054] In the reinforcement learning training process, the adjustment process of the electroplating liquid state data is optimized through a policy gradient method;

[0055] Adjust the control parameters of the electroplating liquid step by step through a backpropagation algorithm;

[0056] At each iteration, adjust the electroplating liquid state data and recalculate the reward function according to the current state data and the quality evaluation result;

[0057] When the reward function reaches the maximum value, it indicates that the electroplating solution state data reaches the best, and the electroplating quality of the electroplating line reaches the best.

[0058] In a second aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the computer program, when executed by the processor, implements any step of the sixteen-line four-drum electroplating line production control method according to the first aspect of the present application.

[0059] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any step of the sixteen-line four-drum electroplating line production control method according to the first aspect of the present application.

[0060] The present application has the following beneficial effects: The present application deploys edge nodes on the production line, collects electroplating solution state data in real time, and performs smoothing and standardization processing, thereby improving the reliability of the data, providing high-quality basic data for subsequent anomaly identification and quality control, using a machine learning algorithm to identify abnormal states of the electroplating solution in real time and perform adaptive adjustment, thereby ensuring the stability of the electroplating solution. The collection and processing of visual and infrared image data effectively improve the detection accuracy of the electroplating layer thickness, uniformity, and defects, reduce the limitations of manual inspection, intelligently control the production process by comprehensively considering the electroplating solution state and electroplating line quality evaluation results, significantly improve the consistency and stability of the electroplating quality, reduce production fluctuations and defective products, improve production efficiency, and ensure high standards of product quality. BRIEF DESCRIPTION OF DRAWINGS

[0061] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0062] Figure 1 The flowchart of the sixteen-line four-drum electroplating line production control method in Example 1. DETAILED DESCRIPTION

[0063] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings.

[0064] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be appreciated that the present application can be practiced in a variety of ways beyond the specific details set forth herein, which can be practiced in other embodiments and applied to other ways within the scope of the present application and its equivalents. Thus, the present application should not be limited to the specific details set forth in the following description.

[0065] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an embodiment that is independent or alternative to other embodiments.

[0066] Embodiment 1, reference Figure 1 As the first embodiment of the present application, the embodiment provides a production control method for a sixteen-line four-roller electroplating line, comprising the following steps:

[0067] S1, deploying edge nodes on all production lines and collecting electroplating solution state data in real time using sensors.

[0068] Using sensors on the edge nodes (including but not limited to temperature sensors, flow meters, current and voltage sensors, pH sensors, and chemical composition concentration sensors), real-time collection of state data of electroplating solution on all production lines (including but not limited to temperature, flow rate, current, voltage, pH value, and electroplating ion concentration);

[0069] Smooth and standardize the real-time collected electroplating solution state data on the edge nodes;

[0070] Specifically, using linear interpolation method, the missing values are calculated by interpolation of adjacent two point data and filled;

[0071] Using the sliding window method, the influence of short-term fluctuations is reduced by calculating the mean or median within a certain window range around the data points;

[0072] Using the Z-score standardization method, the data is converted to standard normal distribution by subtracting the mean and dividing by the standard deviation.

[0073] S2, according to the electroplating solution state data, using machine learning algorithm to identify the state data anomaly of electroplating solution.

[0074] According to the electroplating solution state data, a state data set is constructed;

[0075] According to the range set for the state data of the electroplating solution on the production line, data outside the range set for the state data is labeled as abnormal, and data within the range set for the state data is labeled as normal;

[0076] Mix the state data with abnormal labels and normal labels, and divide them into training set and test set;

[0077] Based on nonlinear support vector machine (SVM), build a plating solution state judgment model on the edge node;

[0078] Input the training set into the plating solution state judgment model, calculate the target loss function, and the expression is as follows:

[0079]

[0080] Wherein, L is the target loss function value, n is the number of training samples, i is the index of the number of training samples, α i , α i+1 are the i-th and i+1-th Lagrange multipliers, y i , y i+1 are the labels of the i-th and i+1-th training samples, 1 for normal and -1 for abnormal, b is the bias term of the hyperplane, σ is the standard deviation of α i and α i+1 , and K(x i , x i+1 ) is the kernel function;

[0081] Divide the training set into k subsets, use k-fold cross-validation, and use one subset as the validation set and the rest as the training set alternately for training and validation;

[0082] According to the range setting of the state data of the plating solution on the production line, preset the accuracy, precision and recall rate thresholds;

[0083] After each training and verification, calculate the accuracy, precision and recall rate, and when the accuracy, precision and recall rate are all greater than the accuracy, precision and recall rate thresholds, the training is completed;

[0084] Input the test set into the plating solution state judgment model, calculate the accuracy, precision and recall rate, and when the accuracy, precision and recall rate are all greater than the accuracy, precision and recall rate thresholds, the plating solution state judgment model training is completed;

[0085] Integrate the incremental learning mechanism into the plating solution state judgment model to update the plating solution state judgment model in real time;

[0086] By calculating the distance between the newly collected state data and the hyperplane of the plating solution state judgment model, the influence degree of the newly collected state data on the plating solution state judgment model is judged, and the expression is as follows:

[0087] |w×x new +b|<∈;

[0088] Wherein, w is the weight vector of the hyperplane, xnew for the newly collected state data, b is the bias term of the hyperplane, and ∈ is the distance threshold of the state data from the hyperplane;

[0089] When the above formula is satisfied, it indicates that the newly collected state data has a greater impact on the current electroplating solution state judgment model, and the electroplating solution state judgment model needs to be updated;

[0090] The newly collected state data is added to the support vector set of the electroplating solution state judgment model, and the hyperplane and the decision function are updated according to the new support vector set;

[0091] Specifically, the hyperplane is updated by recalculating the weighted average of the support vectors, and the updated support vector set will affect the weight vector and the bias term of the SVM, and further affect the classification decision function;

[0092] Periodically perform incremental verification through a new data set to ensure that the model can make good predictions for new patterns.

[0093] The real-time collected state data of the electroplating solution is used to judge the state of the electroplating solution on the production line by using the electroplating solution state judgment model on the edge node.

[0094] S3, collect visual image data and infrared image data of the electroplating line using visual sensors and infrared sensing devices.

[0095] Collect visual image data of the electroplating line using visual sensors to capture defects on the surface of the electroplating line during the electroplating process;

[0096] Collect infrared image data of the electroplating line using infrared sensing devices, emit infrared radiation and receive reflected signals to accurately measure the thickness and uniformity of the electroplating layer;

[0097] Process the visual image data and the infrared image data;

[0098] Use bilateral filtering method to denoise the visual image data;

[0099] Use histogram equalization method to adjust the brightness distribution of the image and improve the contrast of the image, so that the details of the electroplating layer are clearer;

[0100] Use non-local mean method to perform weighted average on similar pixels in the image, maintain details while removing noise, adjust the brightness and contrast of the infrared image, and make the temperature difference of the electroplating layer more obvious, which is convenient for subsequent thickness measurement.

[0101] S4, based on the visual image data and the infrared image data, perform quality evaluation on the electroplating line through image processing algorithms.

[0102] According to the relationship between the infrared radiation intensity of the infrared sensing device and the thickness of the electroplated layer, the thickness of the electroplated layer is calculated, and the expression is as follows:

[0103]

[0104] Wherein, d is the thickness of the electroplated layer, T1 is the reference infrared reflection intensity before electroplating, T2 is the infrared reflection intensity after electroplating, and β is the infrared conduction coefficient;

[0105] The electroplated layer area of the electroplating line is extracted by the image segmentation method, and the standard deviation of the thickness of the electroplated layer in each area is calculated as the uniformity evaluation index. The smaller the uniformity index, the more uniform the electroplated layer, and the expression is as follows:

[0106]

[0107] Wherein, γ is the uniformity evaluation index, δ is the thickness standard deviation of each area, and μ is the average thickness of each area;

[0108] Classify and label the known defects in the visual image data (including normal, crack, bubble, and peeling) ;

[0109] Label multiple defect categories for each image, and provide a bounding box for each defect area;

[0110] Based on the convolutional neural network (CNN), a defect recognition model is constructed;

[0111] Divide the visual image data classified and labeled by the known defects into a training set and a validation set, and input the training set into the defect recognition model;

[0112] Train the input defect recognition model by calculating the cross-entropy loss function of the visual image classified and labeled by the known defects in the training set, and the expression is as follows:

[0113]

[0114] Wherein, Loss is the cross-entropy loss value, N is the number of training data, m is the index of the number of training data, z m is the true label of the defect, is the predicted label of the defect;

[0115] Use gradient descent method (such as Adam optimizer) to minimize the loss function, and iteratively optimize the network weight. After each training, evaluate the performance of the model through the validation set, and adjust the hyperparameters (such as learning rate, batch size, etc.);

[0116] Evaluate the performance of the model by calculating the accuracy and recall rate of the defect recognition model;

[0117] The trained defect recognition model is used to detect defects in newly collected visual image data.

[0118] The overall quality of the electroplated layer is evaluated by comprehensively considering the thickness, uniformity and surface defects of the electroplated layer, and the expression is as follows:

[0119] Q = 0.3d + 0.3y + 0.4D;

[0120] Wherein, Q is the quality evaluation value of the electroplated layer, and D is the number of defects.

[0121] S5, according to the abnormal state of the electroplating solution and the quality evaluation result of the electroplating line, the state parameters of the electroplating solution are adjusted through the self-adaptive algorithm.

[0122] The abnormal data of the electroplating solution and the quality evaluation result of the electroplating line are normalized to ensure that all data are in the same dimension;

[0123] An adaptive control algorithm based on reinforcement learning is adopted to guide the adjustment of the electroplating solution state data through a reward function;

[0124] Wherein, the expression of the reward function is:

[0125] R = 0.5AQ + 0.5AX;

[0126] Wherein, R represents the reward function value, AQ is the deviation between the quality evaluation result of the electroplating line and the target quality, and AX is the deviation between the state parameters of the electroplating solution and its target value;

[0127] In the reinforcement learning training process, the adjustment process of the electroplating solution state data is optimized through the policy gradient method;

[0128] The control parameters of the electroplating solution are gradually adjusted through the back propagation algorithm to ensure that the state data of the electroplating solution changes as much as possible to the target value;

[0129] At each iteration, the electroplating solution state data is adjusted and the reward function is recalculated according to the current state data and the quality evaluation result;

[0130] When the reward function reaches the maximum value, it means that the electroplating solution state data reaches the best, and the electroplating quality of the electroplating line reaches the best.

[0131] The embodiment also provides a computer device suitable for the sixteen-line four-roller electroplating line production control method, which comprises a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to realize the sixteen-line four-roller electroplating line production control method proposed in the above embodiment.

[0132] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved by WIFI, a carrier network, NFC (Near Field Communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, a trackball or a touchpad arranged on the shell of the computer device, or an external keyboard, a touchpad or a mouse, etc.

[0133] The embodiment also provides a storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the production control method of the sixteen-line four-roller barrel plating line. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk or an optical disk.

[0134] In summary, the present application improves the reliability of data by deploying edge nodes on the production line, collecting electroplating liquid state data in real time and performing smoothing and standardization processing, provides high-quality basic data for subsequent anomaly identification and quality control, uses machine learning algorithms to identify abnormal states of electroplating liquid in real time and perform adaptive adjustment, and ensures the stability of electroplating liquid. The collection and processing of visual and infrared image data effectively improve the detection accuracy of electroplating layer thickness, uniformity and defects, reduce the limitations of manual inspection, and through comprehensive consideration of electroplating liquid state and electroplating line quality evaluation results, through intelligent production process control, significantly improve the consistency and stability of electroplating quality, reduce production fluctuations and defective products, improve production efficiency, and ensure high standards of product quality.

[0135] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit it, although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. A production control method for a 16-line, four-roller electroplating line, characterized in that: include, Edge nodes are deployed on all production lines, and sensors are used to collect real-time data on the state of the electroplating solution. Based on the state data of the electroplating solution, machine learning algorithms are used to identify anomalies in the state data of the electroplating solution. The specific steps are as follows. Construct a state dataset based on the state data of the electroplating solution; Based on the range of state data set for the electroplating solution on the production line, data that is outside the range of state data is labeled as abnormal, and data that is within the range of state data is labeled as normal. Mix state data with abnormal and normal labels and divide them into training and test sets; A nonlinear support vector machine (SVM) is used to construct an electroplating solution state judgment model at the edge nodes. The training set is input into the electroplating solution state judgment model, and the target loss function is calculated; Divide the training set into For each subset, use k-fold cross-validation, taking turns using one subset as the validation set and the rest as the training set for training and validation; Based on the range of state data of the electroplating solution on the production line, preset accuracy, precision and recall thresholds are set. After each training and validation cycle, the accuracy, precision, and recall are calculated. Training ends when the accuracy, precision, and recall are all greater than their respective thresholds. The test set is input into the electroplating solution state judgment model, and the accuracy, precision and recall are calculated. When the accuracy, precision and recall are all greater than the accuracy, precision and recall thresholds, the electroplating solution state judgment model training is complete. An incremental learning mechanism is incorporated into the electroplating solution state judgment model to update the model in real time. The state data of the electroplating solution is collected in real time, and the state judgment model of the electroplating solution on the edge node is used to judge the state of the electroplating solution on the production line. The visual image data and infrared image data of the electroplating line are collected using visual sensors and infrared sensing devices. The specific steps are as follows. Visual sensors are used to collect visual image data of the electroplating line to capture defects on the surface of the electroplating line during the electroplating process; Infrared image data of the electroplating line is collected using infrared sensing equipment. By emitting infrared radiation and receiving reflected signals, the thickness and uniformity of the electroplating layer can be accurately measured. Processing visual image data and infrared image data; Based on visual and infrared image data, image processing algorithms are used to assess the quality of electroplating lines. The specific steps are as follows. The thickness of the electroplating layer is calculated based on the relationship between the infrared radiation intensity of the infrared sensing device and the thickness of the electroplating layer. The electroplating area of ​​the electroplating line is extracted by image segmentation method, and the standard deviation of the electroplating thickness in each area is calculated as a uniformity evaluation index. Classify and label known defects in visual image data, including normal, cracks, bubbles, and peeling; Each image is labeled with multiple defect categories, and a bounding box is provided for each defect region; A defect identification model is constructed based on a convolutional neural network (CNN). Visual image data with known defects classified and labeled is divided into training and validation sets, and the training set is input into the defect recognition model; The input defect recognition model is trained by calculating the cross-entropy loss function of visual images classified and labeled with known defects in the training set. The loss function is minimized using gradient descent, and the network weights are iteratively optimized. After each training iteration, the model's performance is evaluated using a validation set, and the hyperparameters are adjusted accordingly. The performance of the defect identification model is evaluated by calculating its accuracy and recall. The trained defect recognition model is used to detect defects in newly acquired visual image data; The overall quality of the electroplated layer is evaluated by comprehensively considering its thickness, uniformity, and surface defects. Based on the abnormal state of the electroplating solution and the quality assessment results of the electroplating line, the state parameters of the electroplating solution are adjusted using an adaptive algorithm.

2. The production control method for a sixteen-line, four-roller electroplating line as described in claim 1, characterized in that: The specific steps for deploying edge nodes on all production lines and using sensors to collect real-time electroplating solution status data are as follows: By using sensors on edge nodes, the status data of electroplating solutions on all production lines can be collected in real time; The real-time electroplating solution status data collected at the edge nodes is smoothed and standardized.

3. The production control method for a sixteen-line, four-roller electroplating line as described in claim 1, characterized in that: The incremental learning mechanism is incorporated into the electroplating solution state judgment model to update the model in real time. The specific steps are as follows: The influence of the newly acquired state data on the electroplating solution state judgment model is determined by calculating the distance between the newly acquired state data and the hyperplane of the electroplating solution state judgment model. The newly collected state data is added to the support vector set of the electroplating solution state judgment model, and the hyperplane and decision function are updated based on the new support vector set.

4. The production control method for a sixteen-line, four-roller electroplating line as described in claim 1, characterized in that: The specific steps for processing the visual image data and infrared image data are as follows. The visual image data is denoised using a bilateral filtering method. Histogram equalization is used to improve image contrast by adjusting the brightness distribution of the image. By using the nonlocal mean method, a weighted average is applied to similar pixels in the image, which removes noise while preserving details and adjusts the brightness and contrast of the infrared image.

5. The production control method for a sixteen-line, four-roller electroplating line as described in claim 1, characterized in that: Based on the abnormal state of the electroplating solution and the quality assessment results of the electroplating line, the state parameters of the electroplating solution are adjusted using an adaptive algorithm. The specific steps are as follows: Abnormal data of electroplating solution and quality assessment results of electroplating line are normalized. An adaptive control algorithm based on reinforcement learning is adopted, which guides the adjustment of electroplating solution state data through a reward function; During reinforcement learning training, the process of adjusting the state data of the electroplating solution is optimized through the policy gradient method. The control parameters of the electroplating solution are gradually adjusted using a backpropagation algorithm. By adjusting the electroplating solution state data and recalculating the reward function based on the current state data and quality assessment results at each iteration; When the reward function reaches its maximum value, it indicates that the electroplating solution state data has reached its optimal level, and the electroplating quality of the electroplating line has reached its optimal level.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the production control method for a sixteen-line four-roller electroplating line as described in any one of claims 1 to 5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the production control method for a sixteen-line four-roller electroplating line as described in any one of claims 1 to 5.

Images