PCB electroplating parameter prediction model training method and device, equipment and storage medium

Abstract

The invention relates to a PCB electroplating parameter prediction model training method and device, equipment and a storage medium. The method comprises the following steps: acquiring initial data of PCB electroplating parameters; cleaning the initial data to obtain target data; dividing the target data into training data and test data based on a random number generator; an initial PCB electroplating parameter prediction model is trained based on the training data to obtain target PCB electroplating parameter prediction model parameters, and in the training process, the fitness of the initial PCB electroplating parameter prediction model to the training data is gradually improved based on a GBDT-PL algorithm; and updating the initial PCB electroplating parameter prediction model based on the target PCB electroplating parameter prediction model parameters to obtain a target PCB electroplating parameter prediction model. According to the method, the adaptability of the initial PCB electroplating parameter prediction model to the training data is gradually improved, so that the prediction accuracy of the target PCB electroplating parameter prediction model to the PCB electroplating parameters is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit board electroplating, and in particular to a PCB electroplating parameter prediction model training method, device, equipment and storage medium. Background Art

[0002] PCB electroplating parameters include current density and electroplating time. The accuracy of electroplating parameter settings directly determines the product quality yield, production efficiency and production cost. Estimate the current density according to the PCB electroplating copper thickness requirements, and set the electroplating time according to the upper and lower limits of the current density that the VCP line can withstand and the estimated current density. However, in the prior art, by obtaining the characteristic vector that affects the current density, a LightGBM decision tree model is established to predict parameters. Since the tree model is a constant at the leaf node. When the predicted sample copper thickness requirement exceeds the value range of the training sample copper thickness requirement, the prediction model will output the electroplating parameter prediction value divided into the training sample copper thickness requirement threshold leaf node. The electroplated copper thickness produced using the electroplating parameter prediction value is lower than the copper thickness requirement value or exceeds the copper thickness upper limit, resulting in the need for rework of the produced PCB, thereby increasing production costs, reducing production efficiency and quality yield. Summary of the invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a PCB electroplating parameter prediction model training method, device, equipment and storage medium.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0005] The present invention is achieved through the following technical solutions:

[0006] In a first aspect, this embodiment provides a PCB electroplating parameter prediction model training method, comprising the following steps:

[0007] Get the initial data of PCB electroplating parameters;

[0008] Cleaning the initial data to obtain target data;

[0009] Dividing the target data into training data and test data based on a random number generator;

[0010] Based on the training data, the initial PCB electroplating parameter prediction model is trained to obtain target PCB electroplating parameter prediction model parameters, wherein during the training process, based on the GBDT-PL algorithm, the adaptability of the initial PCB electroplating parameter prediction model to the training data is gradually improved;

[0011] The initial PCB electroplating parameter prediction model is updated based on the target PCB electroplating parameter prediction model parameters to obtain a target PCB electroplating parameter prediction model.

[0012] Furthermore, the step of cleaning the initial data to obtain target data includes:

[0013] Scanning the PCB electroplating process parameters in the initial data based on the abnormal data strategy, eliminating the abnormal data, and obtaining the first cleaning data;

[0014] Performing duplicate data cleaning on the first cleaned data based on a duplicate data strategy to obtain second cleaned data;

[0015] The missing data of the second cleaned data is completed based on the data completion strategy to obtain the target data.

[0016] Furthermore, the step of dividing the target data into training data and test data based on a random number generator includes:

[0017] Obtaining proportion data generated by the random number generator for dividing target data;

[0018] Dividing the target data based on the proportion data to generate a plurality of training sets and a plurality of validation sets;

[0019] Determine whether the data in the plurality of training sets overlap with the data in the plurality of validation sets;

[0020] If the data in the several training sets do not overlap with the data in the several verification sets, the several training sets are used to generate the training data, and the several verification sets are used to generate the verification data.

[0021] Further, the step of training the initial PCB electroplating parameter prediction model based on the training data to obtain the parameters of the target PCB electroplating parameter prediction model includes:

[0022] Determining a linear function based on the GBDT-PL algorithm and the initial PCB electroplating parameter prediction model;

[0023] Based on the training data, determining a gradient value of the linear function;

[0024] Based on the gradient value, construct a target matrix;

[0025] Performing diagonal estimation on the target matrix to obtain parameter increments;

[0026] The parameters of the initial PCB electroplating parameter prediction model are obtained according to the initial parameters of the initial PCB electroplating parameter prediction model and the parameter increment.

[0027] Furthermore, the step of determining a linear function based on the GBDT-PL algorithm and the initial PCB electroplating parameter prediction model includes:

[0028] Constructing a low-rank parameter matrix based on the parameters of the GBDT-PL algorithm and the initial PCB electroplating parameter prediction model;

[0029] Splitting the low-rank parameter matrix;

[0030] Perform singular value decomposition on the split low-rank parameter submatrix to obtain spatial data;

[0031] Based on the spatial data, determining a tensor;

[0032] Based on the tensor, a linear function is determined.

[0033] Furthermore, the step of determining the gradient value of the linear function based on the target training set includes:

[0034] Splitting the training set into multiple training samples;

[0035] Performing reverse calculation on the training sample to obtain a gradient prediction value;

[0036] Based on the training sample, performing a differential operation on the linear function to obtain a true value of the gradient;

[0037] Determine a gradient loss value according to the gradient prediction value and the gradient true value;

[0038] According to the gradient loss value, a gradient value of the linear function is determined.

[0039] Furthermore, the step of performing diagonal estimation on the target matrix to obtain parameter increments includes:

[0040] Extracting diagonal elements of the target matrix;

[0041] constructing an intermediate matrix based on the diagonal elements;

[0042] The parameter increment is obtained based on the intermediate matrix.

[0043] In a second aspect, the present embodiment provides a PCB electroplating parameter prediction model training device, including: an acquisition unit, a cleaning unit, a sorting unit, a training unit and a loading unit;

[0044] The acquisition unit is used to acquire initial data of PCB electroplating parameters;

[0045] The cleaning unit is used to clean the initial data to obtain target data;

[0046] The collating unit is used to divide the target data into training data and test data based on a random number generator;

[0047] The training unit is used to train the initial PCB electroplating parameter prediction model based on the training data to obtain target PCB electroplating parameter prediction model parameters;

[0048] The loading unit is used to update the initial PCB electroplating parameter prediction model based on the target PCB electroplating parameter prediction model parameters to obtain a target PCB electroplating parameter prediction model.

[0049] In a third aspect, the present embodiment provides a computer device, which includes a memory and a processor, wherein a computer program is stored on the processor, and when the processor executes the computer program, the PCB electroplating parameter prediction model training method as described in any one of claims 1-7 is implemented.

[0050] In a fourth aspect, the present embodiment provides a storage medium, wherein the storage medium stores a computer program, wherein the computer program includes program instructions, and wherein the program instructions, when executed by a processor, can implement the PCB electroplating parameter prediction model training method as described in any one of claims 1-7.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows: when the initial PCB electroplating parameter prediction model is trained based on the GBDT-PL algorithm, the adaptability of the initial PCB electroplating parameter prediction model to the training data is gradually improved, the accuracy of the target PCB electroplating parameter prediction model in predicting the PCB electroplating parameters is improved, the defective rate caused by improper electroplating parameter setting is reduced, the production cost is reduced, the production efficiency is improved, and at the same time, the consistency of PCB copper thickness is ensured, and the yield rate of PCB electroplated copper is improved.

[0052] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the following preferred embodiments are specifically cited and described in detail as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic diagram of a flow chart of a PCB electroplating parameter prediction model training method provided by an embodiment of the present invention;

[0054] Figure 2 A schematic diagram of a process for executing step S2 in the PCB electroplating parameter prediction model training method provided in an embodiment of the present invention;

[0055] Figure 3 A schematic diagram of a flow chart for executing step S3 in the PCB electroplating parameter prediction model training method provided in an embodiment of the present invention;

[0056] Figure 4A schematic diagram of a flow chart of executing step S4 in the PCB electroplating parameter prediction model training method provided in an embodiment of the present invention;

[0057] Figure 5 A schematic diagram of a flow chart of executing step S41 in the PCB electroplating parameter prediction model training method provided in an embodiment of the present invention;

[0058] Figure 6 A schematic diagram of a flow chart of executing step S42 in the PCB electroplating parameter prediction model training method provided in an embodiment of the present invention;

[0059] Figure 7 A schematic diagram of a flow chart for executing step S44 in the PCB electroplating parameter prediction model training method provided in an embodiment of the present invention;

[0060] Figure 8 A schematic block diagram of a PCB electroplating parameter prediction model training device provided in an embodiment of the present invention;

[0061] Fig. 9 A schematic block diagram of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0064] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0065] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0066] See also Figure 1In the specific embodiment shown, the present invention discloses a PCB electroplating parameter prediction model training method, comprising the following steps:

[0067] Step S1, obtaining initial data of PCB electroplating parameters;

[0068] It is understandable that the initial data includes theoretical current density and FA (Factory Automation) data. FA data includes current density, electroplating time, actual value of electroplated copper thickness, required value of electroplated copper thickness, plate thickness, minimum aperture, aspect ratio and other related information. Therefore, the theoretical current density in the initial data is calculated by the required value of electroplated copper thickness in FA data / actual value of electroplated copper thickness*current density. Through the above steps, strong support is provided for the initial PCB electroplating parameter prediction model training. Based on these initial data, the generated target PCB electroplating parameter prediction model can optimize the electroplating process and ensure the consistency of product quality.

[0069] Step S2, cleaning the initial data to obtain target data;

[0070] It is understandable that the initial data may contain various duplicate values, outliers, missing values ​​or inconsistent formats, which will make the training of the initial PCB electroplating parameter prediction model invalid. Therefore, the initial data cleaning ensures that the electroplating parameter data set used for the initial PCB electroplating parameter prediction model is accurate, complete and consistent, thereby improving the prediction accuracy and generalization ability of the target PCB electroplating parameter prediction model.

[0071] Through the above steps, the initial data is cleaned to improve the quality of training data and improve the accuracy and stability of PCB electroplating parameter prediction model training.

[0072] Step S3, dividing the target data into training data and test data based on a random number generator;

[0073] It is understandable that the random number generator randomly divides the target data into training data and test data to ensure the randomness of the training data and prevent the deviation between the training data and the test data from being too large. At the same time, when dividing the data, the distribution consistency between the training data and the test data must be maintained. This can be achieved through stratified sampling, that is, dividing the data according to the proportion of a certain important feature (such as label distribution).

[0074] Through the above steps, based on the random number generator, it is ensured that the results of each division of training data and test data are consistent, thereby improving the reproducibility and comparability of the target PCB electroplating parameter prediction model, which is conducive to maintaining the consistency of training data under different experimental conditions and avoiding overfitting or underfitting of the target PCB electroplating parameter prediction model.

[0075] Step S4, training the initial PCB electroplating parameter prediction model based on the training data to obtain target PCB electroplating parameter prediction model parameters, wherein during the training process, based on the GBDT-PL algorithm, the adaptability of the initial PCB electroplating parameter prediction model to the training data is gradually improved;

[0076] It can be understood that the GBDT-PL algorithm is a combination of the Gradient Boosting Decision Tree (GBDT) and the Partial Least Squares (PLS). GBDT is used for feature selection and nonlinear relationship fitting of training data, while PLS is used to process multicollinearity and dimensionality reduction of training data. Through step-by-step iteration, the parameters of the initial PCB electroplating parameter prediction model are continuously adjusted so that the initial PCB electroplating parameter prediction model better fits the training data. Through the above steps, the GBDT-PL algorithm enables the initial PCB electroplating parameter prediction model to capture complex nonlinear relationships and improve the accuracy of the prediction. At the same time, the fitness of the PCB electroplating parameter prediction model to the training data is gradually improved to avoid overfitting.

[0077] Step S5, updating the initial PCB electroplating parameter prediction model based on the parameters of the target PCB electroplating parameter prediction model to obtain a target PCB electroplating parameter prediction model.

[0078] It can be understood that after the initial PCB electroplating parameter prediction model is trained, a set of optimized parameters will be obtained. These parameters represent the best fitting state of the initial PCB electroplating parameter prediction model on the training data. These parameters are used to replace the default parameters in the initial model to obtain the target PCB electroplating parameter prediction model. At the same time, after updating the initial PCB electroplating parameter prediction model, test data is also needed to verify the generalization ability of the target PCB electroplating parameter prediction model to ensure that the target PCB electroplating parameter prediction model not only performs well on the training data, but also can give accurate predictions on unseen data.

[0079] Through the above steps, the target PCB electroplating parameter prediction model improves the accuracy of predicting PCB electroplating parameters and reduces prediction errors.

[0080] Through steps S1 to S5, when the initial PCB electroplating parameter prediction model is trained by the GBDT-PL algorithm, the adaptability of the initial PCB electroplating parameter prediction model to the training data is gradually improved, the accuracy of the target PCB electroplating parameter prediction model in predicting the PCB electroplating parameters is improved, the defective rate caused by improper electroplating parameter settings is reduced, the production cost is reduced, the production efficiency is improved, and the consistency of PCB copper thickness is ensured, and the yield rate of PCB electroplated copper is improved.

[0081] See also Figure 2 , the step of cleaning the initial data to obtain target data includes:

[0082] S21, scanning the PCB electroplating process parameters in the initial data based on the abnormal data strategy, eliminating the abnormal data, and obtaining the first cleaning data;

[0083] S22, performing duplicate data cleaning on the first cleaned data based on a duplicate data strategy to obtain second cleaned data;

[0084] S23: completing missing data on the second cleaned data based on a data completion strategy to obtain the target data.

[0085] Through steps S21 to S23, outliers, duplicate records and missing values ​​are eliminated, which significantly improves data quality, improves the accuracy and stability of the initial PCB electroplating parameter prediction model training, improves production efficiency, and reduces the overfitting phenomenon of the target PCB parameter prediction model.

[0086] In one embodiment, see Figure 3 , the step of dividing the target data into training data and test data based on a random number generator comprises:

[0087] S31, obtaining the ratio data generated by the random number generator for dividing the target data;

[0088] S32, dividing the target data based on the proportion data to generate a plurality of training sets and a plurality of validation sets;

[0089] S33, determining whether the data in the plurality of training sets overlap with the data in the plurality of validation sets;

[0090] S34, if the data in the several training sets do not overlap with the data in the several verification sets, the several training sets are used to generate the training data, and the several verification sets are used to generate the verification data.

[0091] Through steps S31 to S34, based on the seed of the random number generator, it is ensured that the same data division result is obtained each time the division is performed, so as to facilitate the reproduction and comparison of the training results of the initial PCB parameter prediction model. It is checked whether the divided training set and the validation set contain the same data, so as to ensure that the training set and the validation set do not overlap with each other, and improve the generalization ability of the target PCB parameter prediction model on unseen data.

[0092] In one embodiment, see Figure 4 The step of training the initial PCB electroplating parameter prediction model based on the training data to obtain the parameters of the target PCB electroplating parameter prediction model includes:

[0093] S41, determining a linear function based on the GBDT-PL algorithm and the initial PCB electroplating parameter prediction model;

[0094] S42, determining a gradient value of the linear function based on the training data;

[0095] S43, constructing a target matrix based on the gradient value;

[0096] S44, performing diagonal estimation on the target matrix to obtain parameter increments;

[0097] The parameters of the initial PCB electroplating parameter prediction model are obtained according to the initial parameters of the initial PCB electroplating parameter prediction model and the parameter increment.

[0098] Through step S41 to step S44, the fitting degree of the initial PCB electroplating parameter prediction model to the training data is gradually increased, and the generalization ability of the model on unseen data is enhanced. By constructing target matrix operations and diagonal estimation, the training process of the target PCB electroplating parameter prediction model is accelerated, and the training time and computing cost are reduced. The incremental parameter update strategy helps the model to remain stable during the training process and avoids the occurrence of training instability or difficulty in convergence due to large parameter jumps.

[0099] In one embodiment, see Figure 5 , the step of determining the linear function based on the GBDT-PL algorithm and the initial PCB electroplating parameter prediction model comprises:

[0100] S411, constructing a low-rank parameter matrix based on the parameters of the GBDT-PL algorithm and the initial PCB electroplating parameter prediction model;

[0101] S412, splitting the low-rank parameter matrix;

[0102] S413, performing singular value decomposition processing on the split low-rank parameter submatrix to obtain spatial data;

[0103] S414, determining a tensor based on the spatial data;

[0104] S415, determining a linear function based on the tensor.

[0105] Through steps S411 to S414, the low-rank parameter matrix is ​​constructed, split, SVD processed, and the loss function is determined based on the tensor, which achieves efficient compression and acceleration of the model and improves the generalization ability of the model.

[0106] In one embodiment, see Figure 6 , the step of determining the gradient value of the linear function based on the target training set comprises:

[0107] S421, dividing the training set into multiple training samples;

[0108] S422, performing reverse calculation on the training sample to obtain a gradient prediction value;

[0109] S423, performing a differential operation on the linear function based on the training sample to obtain a true value of the gradient;

[0110] S424, determining a gradient loss value according to the gradient prediction value and the gradient true value;

[0111] S425: Determine the gradient value of the linear function according to the gradient loss value.

[0112] Through steps S421 to S425, by dividing the training set into multiple training samples and performing inverse calculation and differential operation, accurate and efficient gradient calculation is achieved, thereby improving the training efficiency of the PCB electroplating parameter prediction model.

[0113] In one embodiment, see Figure 7 , the step of performing diagonal estimation on the target matrix to obtain parameter increments comprises:

[0114] S441, extracting diagonal elements of the target matrix;

[0115] S442, constructing an intermediate matrix based on the diagonal elements;

[0116] S443, obtaining the parameter increment based on the intermediate matrix

[0117] Through steps S441 to S443, accurate capture and effective utilization of key information of the target matrix are achieved, thereby efficiently calculating and updating the parameter increments of the initial PCB electroplating parameter prediction model, and enhancing the adjustment and optimization capabilities of the model.

[0118] See also Figure 8, the present invention also discloses a PCB electroplating parameter prediction model training device, including: an acquisition unit 10, a cleaning unit 20, a sorting unit 30, a training unit 40 and a loading unit 50;

[0119] The acquisition unit 10 is used to acquire initial data of PCB electroplating parameters;

[0120] The cleaning unit 20 is used to clean the initial data to obtain target data;

[0121] The arranging unit 30 is used to divide the target data into training data and test data based on a random number generator;

[0122] The training unit 40 is used to train the initial PCB electroplating parameter prediction model based on the training data to obtain target PCB electroplating parameter prediction model parameters;

[0123] The loading unit 50 is used to update the initial PCB electroplating parameter prediction model based on the target PCB electroplating parameter prediction model parameters to obtain a target PCB electroplating parameter prediction model.

[0124] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned PCB electroplating parameter prediction model training device and each unit can refer to the corresponding description in the aforementioned method embodiment, and for the convenience and simplicity of description, it will not be repeated here.

[0125] The above can be implemented in the form of a computer program, which can be used in Fig. 9 Runs on the computer device shown.

[0126] See also Fig. 9 , Fig. 9 is a schematic block diagram of a computer device provided in an embodiment of the present application; the computer device 500 may be a terminal or a server, wherein the terminal may be an electronic device with communication functions such as a smart phone, a tablet computer, a laptop computer, a desktop computer, a personal digital assistant, and a wearable device. The server may be an independent server or a server cluster composed of multiple servers.

[0127] See also Fig. 9 The computer device 500 includes a processor 502 , a memory and a network interface 505 connected via a system bus 501 , wherein the memory may include a non-volatile storage medium 503 and an internal memory 504 .

[0128] The non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions, which, when executed, can enable the processor 502 to execute a PCB electroplating parameter prediction model training method: obtaining initial data of PCB electroplating parameters; cleaning the initial data to obtain target data; dividing the target data into training data and test data based on a random number generator; training the initial PCB electroplating parameter prediction model based on the training data to obtain target PCB electroplating parameter prediction model parameters, wherein, during the training process, based on the GBDT-PL algorithm, the fitness of the initial PCB electroplating parameter prediction model to the training data is gradually improved; and the initial PCB electroplating parameter prediction model is updated based on the target PCB electroplating parameter prediction model parameters to obtain a target PCB electroplating parameter prediction model.

[0129] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500 .

[0130] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute an SSD garbage collection acceleration method.

[0131] The network interface 505 is used to communicate with other devices over the network. Fig. 9 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0132] The processor 502 is used to run the computer program 5032 stored in the memory to implement the following steps:

[0133] Step S1, obtaining initial data of PCB electroplating parameters;

[0134] Step S2, cleaning the initial data to obtain target data;

[0135] Step S3, dividing the target data into training data and test data based on a random number generator;

[0136] Step S4, training the initial PCB electroplating parameter prediction model based on the training data to obtain target PCB electroplating parameter prediction model parameters, wherein during the training process, based on the GBDT-PL algorithm, the adaptability of the initial PCB electroplating parameter prediction model to the training data is gradually improved;

[0137] Step S5, updating the initial PCB electroplating parameter prediction model based on the target PCB electroplating parameter prediction model parameters to obtain a target PCB electroplating parameter prediction model.

[0138] It should be understood that in the embodiment of the present application, the processor 502 may be a central processing unit (CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0139] It can be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiment can be completed by instructing the relevant hardware through a computer program. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiment of the above method.

[0140] Therefore, the present invention also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program, wherein the computer program includes program instructions, and the program instructions, when executed by a processor, can implement the above-mentioned PCB electroplating parameter prediction model training method. The storage medium stores a computer program, wherein the computer program includes program instructions, and the program instructions, when executed by a processor, can implement the above-mentioned method. The program instructions include the following steps:

[0141] Step S1, obtaining initial data of PCB electroplating parameters;

[0142] Step S2, cleaning the initial data to obtain target data;

[0143] Step S3, dividing the target data into training data and test data based on a random number generator;

[0144] Step S4, training the initial PCB electroplating parameter prediction model based on the training data to obtain target PCB electroplating parameter prediction model parameters, wherein during the training process, based on the GBDT-PL algorithm, the adaptability of the initial PCB electroplating parameter prediction model to the training data is gradually improved;

[0145] Step S5, updating the initial PCB electroplating parameter prediction model based on the target PCB electroplating parameter prediction model parameters to obtain a target PCB electroplating parameter prediction model.

[0146] The storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk, etc., which are computer-readable storage media that can store program codes.

[0147] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0148] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each unit is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0149] The steps in the method of the embodiment of the present invention can be adjusted in order, combined and deleted according to actual needs. The units in the device of the embodiment of the present invention can be combined, divided and deleted according to actual needs. In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0150] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, terminal, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention.

[0151] The above embodiments are preferred implementation schemes of the present invention. In addition, the present invention may also be implemented in other ways. Any obvious replacement without departing from the concept of the present technical solution is within the protection scope of the present invention.

Claims

1. PCB electroplating parameter prediction model training method, characterized in that: The following steps are involved: Get the initial data of PCB electroplating parameters; Cleaning the initial data to obtain target data; Dividing the target data into training data and test data based on a random number generator; Based on the training data, the initial PCB electroplating parameter prediction model is trained to obtain target PCB electroplating parameter prediction model parameters, wherein during the training process, based on the GBDT-PL algorithm, the adaptability of the initial PCB electroplating parameter prediction model to the training data is gradually improved; The initial PCB electroplating parameter prediction model is updated based on the target PCB electroplating parameter prediction model parameters to obtain a target PCB electroplating parameter prediction model.

2. The PCB electroplating parameter prediction model training method according to claim 1, characterized in that: The step of cleaning the initial data to obtain target data comprises: Scanning the PCB electroplating process parameters in the initial data based on the abnormal data strategy, eliminating the abnormal data, and obtaining the first cleaning data; Performing duplicate data cleaning on the first cleaned data based on a duplicate data strategy to obtain second cleaned data; The missing data of the second cleaned data is completed based on the data completion strategy to obtain the target data.

3. The PCB electroplating parameter prediction model training method according to claim 1, characterized in that: The step of dividing the target data into training data and test data based on a random number generator comprises: Obtaining proportion data generated by the random number generator for dividing target data; Dividing the target data based on the proportion data to generate a plurality of training sets and a plurality of validation sets; Determine whether the data in the plurality of training sets overlap with the data in the plurality of validation sets; If the data in the several training sets do not overlap with the data in the several verification sets, the several training sets are used to generate the training data, and the several verification sets are used to generate the verification data.

4. The PCB electroplating parameter prediction model training method according to claim 1, characterized in that: The step of training the initial PCB electroplating parameter prediction model based on the training data to obtain the parameters of the target PCB electroplating parameter prediction model comprises: Determining a linear function based on the GBDT-PL algorithm and the initial PCB electroplating parameter prediction model; Based on the training data, determining a gradient value of the linear function; Based on the gradient value, construct a target matrix; Performing diagonal estimation on the target matrix to obtain parameter increments; The parameters of the initial PCB electroplating parameter prediction model are obtained according to the initial parameters of the initial PCB electroplating parameter prediction model and the parameter increment.

5. The PCB electroplating parameter prediction model training method according to claim 4, characterized in that: The step of determining a linear function based on the GBDT-PL algorithm and the initial PCB electroplating parameter prediction model comprises: Constructing a low-rank parameter matrix based on the parameters of the GBDT-PL algorithm and the initial PCB electroplating parameter prediction model; Splitting the low-rank parameter matrix; Perform singular value decomposition on the split low-rank parameter submatrix to obtain spatial data; Based on the spatial data, determining a tensor; Based on the tensor, a linear function is determined.

6. The PCB electroplating parameter prediction model training method according to claim 4, characterized in that: Based on the training data, the step of determining the gradient value of the linear function comprises: Splitting the training data into multiple training samples; Performing reverse calculation on the training sample to obtain a gradient prediction value; Based on the training sample, performing a differential operation on the linear function to obtain a true value of the gradient; Determine a gradient loss value according to the gradient prediction value and the gradient true value; According to the gradient loss value, a gradient value of the linear function is determined.

7. The PCB electroplating parameter prediction model training method according to claim 4, characterized in that: The step of performing diagonal estimation on the target matrix to obtain parameter increments comprises: Extracting diagonal elements of the target matrix; constructing an intermediate matrix based on the diagonal elements; The parameter increment is obtained based on the intermediate matrix.

8. PCB electroplating parameter prediction model training device, characterized in that: include: Acquisition unit, cleaning unit, sorting unit, training unit and loading unit; The acquisition unit is used to acquire initial data of PCB electroplating parameters; The cleaning unit is used to clean the initial data to obtain target data; The acquisition unit is used to acquire initial data of PCB electroplating parameters; The cleaning unit is used to clean the initial data to obtain target data; The collating unit is used to divide the target data into training data and test data based on a random number generator; The training unit is used to train the initial PCB electroplating parameter prediction model based on the training data to obtain target PCB electroplating parameter prediction model parameters; The loading unit is used to update the initial PCB electroplating parameter prediction model based on the target PCB electroplating parameter prediction model parameters to obtain a target PCB electroplating parameter prediction model.

9. A computer device, characterized in that: The computer device includes a memory and a processor, a computer program is stored on the processor, and when the processor executes the computer program, the PCB electroplating parameter prediction model training method as described in any one of claims 1-7 is implemented.

10. A storage medium, characterized in that: The storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor, the PCB electroplating parameter prediction model training method according to any one of claims 1 to 7 can be implemented.

Images