Automatic production method and system for multi-layer PCB

Abstract

The invention provides an automatic production method and system for a multi-layer PCB, and relates to the technical field of printed circuit boards, and the method comprises the steps: carrying out the multi-layer alignment processing of a to-be-pressed PCB; according to the basic feature vector of the laminated PCB, multi-node lamination parameter credible mining is carried out; performing multi-node lamination parameter combination decision on the laminated PCB according to the multi-node lamination trusted space; performing multi-dimensional prediction evaluation on the first lamination control domain according to the lamination evaluation multi-model; performing multi-target optimization on the first pressing control domain based on the pressing evaluation map and the pressing quality analysis space; and according to the multi-node lamination credible space, the lamination evaluation multi-model and the lamination quality analysis space, carrying out expansion optimization on the second lamination control domain, and according to a lamination optimization strategy, carrying out lamination stage management and control on the laminated PCB. According to the invention, the technical problem of poor quality of the PCB in the prior art can be solved, and the technical effect of improving the quality of the PCB is achieved.

Description

An automated production method and system for multilayer PCBs Technical Field

[0001] This application relates to the field of printed circuit technology, and in particular to an automated production method and system for multilayer PCBs. Background Technology

[0002] In the process of printed circuit board manufacturing, the lamination process is a key link that determines the quality and performance of the finished product. Its core task is to ensure that each layer of circuit board and insulating medium are firmly bonded and kept flat during the lamination of multi-layer PCBs through multi-node processes such as vacuum preheating, hot lamination and cold pressing.

[0003] Currently, existing methods often set fixed parameters such as temperature, pressure, and time in a single process stage. While this can ensure production stability to some extent, it is difficult to cope with batch differences in materials, variations in the number of layers, and minor structural deviations. This fixed process strategy not only leads to fluctuations in lamination quality but also results in problems such as insufficient bonding strength, bubbles or delamination defects, and board warping.

[0004] In summary, existing technologies suffer from technical problems due to the fixed lamination process parameters and lack of multi-node collaborative optimization. This results in an inability to effectively adapt to batch differences in materials, variations in the number of layers, and structural deviations, further affecting the bonding strength, appearance flatness, and overall quality stability of the finished PCB product. Summary of the Invention

[0005] The purpose of this application is to provide an automated production method and system for multilayer PCBs, in order to solve the technical problems in the prior art where the fixed lamination process parameters and lack of multi-node collaborative optimization lead to an inability to effectively adapt to batch differences in materials, changes in the number of layers, and structural deviations, which further affect the bonding strength, appearance flatness, and overall quality stability of the finished PCB.

[0006] In view of the above problems, this application provides an automated production method and system for multilayer PCBs.

[0007] In a first aspect, this application provides an automated production method for multilayer PCBs, implemented through an automated production system for multilayer PCBs, comprising: performing multilayer alignment processing on the PCBs to be laminated to obtain laminated PCBs that meet a predetermined alignment accuracy; performing multi-node lamination parameter reliability mining based on the basic feature vectors of the laminated PCBs to construct a multi-node lamination reliability space; performing multi-node lamination parameter combination decision-making on the laminated PCBs based on the multi-node lamination reliability space to obtain a first lamination control domain; performing multi-dimensional prediction evaluation on the first lamination control domain based on a lamination evaluation multi-model to construct a lamination evaluation map; constructing a lamination quality analysis space, and performing multi-objective optimization on the first lamination control domain based on the lamination evaluation map and the lamination quality analysis space to obtain a second lamination control domain; performing expansion optimization on the second lamination control domain based on the multi-node lamination reliability space, the lamination evaluation multi-model, and the lamination quality analysis space to obtain a lamination optimization strategy, and performing lamination stage control on the laminated PCBs based on the lamination optimization strategy.

[0008] Preferably, the automated production method for multilayer PCBs further includes: performing a historical retrieval of pressing control based on the basic feature vector to obtain a set of historical pressing control schemes; performing reliable mining of vacuum preheating node parameters based on the set of historical pressing control schemes to construct a reliable vacuum preheating space; performing reliable mining of hot pressing node parameters based on the set of historical pressing control schemes to construct a reliable hot pressing space; performing reliable mining of cold pressing and shaping node parameters based on the set of historical pressing control schemes to construct a reliable cold pressing and shaping space; and generating the multi-node pressing reliable space based on the reliable vacuum preheating space, the reliable hot pressing space, and the reliable cold pressing and shaping space.

[0009] Preferably, the automated production method for multilayer PCBs further includes: classifying vacuum preheating node parameters according to the historical set of lamination control schemes to obtain multiple vacuum preheating parameter regions; parsing trigger intervals according to the multiple vacuum preheating parameter regions to obtain multiple preheating parameter trigger intervals; evaluating the endpoint frequent states of the multiple preheating parameter trigger intervals according to the multiple vacuum preheating parameter regions to obtain multiple endpoint frequent state coefficients; performing endpoint reliability detection on the multiple preheating parameter trigger intervals based on the multiple endpoint frequent state coefficients and endpoint frequent state thresholds to determine an untrusted endpoint set; and performing endpoint correction on the multiple preheating parameter trigger intervals based on the multiple vacuum preheating parameter regions, the untrusted endpoint set, and the endpoint frequent state thresholds to generate the vacuum preheating reliability space.

[0010] Preferably, the automated production method for multilayer PCBs further includes: performing simulated pressing control on the laminated PCBs according to each pressing control scheme within the first pressing control domain to obtain multiple pressing simulation data; the pressing evaluation multi-model includes an adhesive strength evaluation model, a pressing defect evaluation model, and a flatness uniformity evaluation model; inputting the multiple pressing simulation data into the adhesive strength evaluation model to obtain an adhesive strength evaluation sequence; inputting the multiple pressing simulation data into the pressing defect evaluation model to obtain a pressing defect evaluation sequence; inputting the multiple pressing simulation data into the flatness uniformity evaluation model to obtain a flatness uniformity evaluation sequence; mapping the first pressing control domain according to the adhesive strength evaluation sequence, the pressing defect evaluation sequence, and the flatness uniformity evaluation sequence to generate the pressing evaluation map.

[0011] Preferably, the automated production method for multilayer PCBs further includes: generating a bonding evaluation constraint matrix based on bonding strength evaluation constraints, bonding defect evaluation constraints, and flatness uniformity evaluation constraints; performing initial optimization on the first bonding control domain based on the bonding evaluation map and the bonding evaluation constraint matrix to obtain an initial bonding optimization domain; calculating the bonding quality of each bonding control scheme within the initial bonding optimization domain based on the bonding quality analytical space to obtain a bonding quality map; and performing bonding quality optimization on the initial bonding optimization domain based on the bonding quality map and a bonding quality threshold to generate a second bonding control domain.

[0012] Preferably, the automated production method for multilayer PCBs further includes: randomly mutating the second pressing control domain according to the multi-node pressing confidence space to obtain a first pressing control variation domain; performing pressing evaluation optimization on the first pressing control variation domain based on the pressing evaluation constraint matrix and the pressing evaluation multi-model to obtain a second pressing control variation domain; performing pressing quality optimization on the second pressing control variation domain based on the pressing quality analytical space and the pressing quality threshold to obtain a third pressing control variation domain; expanding the second pressing control domain according to the third pressing control variation domain to obtain a third pressing control domain, and performing iterative pressing quality optimization based on the third pressing control domain to generate the pressing optimization strategy.

[0013] Preferably, the automated production method for multilayer PCBs further includes: constructing a lamination quality evaluation coordinate system, the lamination quality evaluation coordinate system including an adhesive strength axis, a lamination defect axis, a flatness uniformity axis, and a lamination quality axis; configuring vector weight relationships for the lamination quality evaluation coordinate system by multiple lamination quality evaluation experts to obtain multiple evaluation weight relationship vectors; performing reliable weight calculations for each dimension based on the multiple evaluation weight relationship vectors to establish reliable weight relationship vectors; and performing weighted optimization on the lamination quality evaluation coordinate system based on the reliable weight relationship vectors to generate the lamination quality analytical space.

[0014] Preferably, the automated production method for multilayer PCBs further includes: obtaining PCB design information of the PCB to be laminated; performing multilayer alignment processing on the PCB to be laminated according to the PCB design information to obtain an initial alignment board; constructing a standard alignment model according to the PCB design information, and evaluating the alignment accuracy of the initial alignment board according to the standard alignment model to obtain an alignment accuracy coefficient; if the alignment accuracy coefficient meets the predetermined alignment accuracy, identifying the initial alignment board as the laminated PCB.

[0015] Preferably, the automated production method for multilayer PCBs further includes: if the alignment accuracy coefficient does not meet the predetermined alignment accuracy, correcting the alignment deviation of the initial alignment board according to the standard alignment model to obtain the laminated PCB.

[0016] Secondly, this application also provides an automated production system for multilayer PCBs, used to execute an automated production method for multilayer PCBs as described in the first aspect, comprising: a multilayer alignment processing module for performing multilayer alignment processing on the PCBs to be laminated to obtain laminated PCBs that meet a predetermined alignment accuracy; a lamination parameter reliability mining module for performing multi-node lamination parameter reliability mining based on the basic feature vectors of the laminated PCBs to construct a multi-node lamination reliability space; and a combination decision module for performing multi-node lamination parameter combination decisions on the laminated PCBs based on the multi-node lamination reliability space to obtain a first lamination control. The system comprises: a control domain; a multi-dimensional prediction and evaluation module, used to perform multi-dimensional prediction and evaluation of the first pressing control domain based on the pressing evaluation multi-model, and construct a pressing evaluation map; a multi-objective optimization module, used to construct a pressing quality analysis space, and perform multi-objective optimization of the first pressing control domain based on the pressing evaluation map and the pressing quality analysis space, to obtain a second pressing control domain; and an expansion optimization module, used to expand and optimize the second pressing control domain based on the multi-node pressing confidence space, the pressing evaluation multi-model, and the pressing quality analysis space, to obtain a pressing optimization strategy, and to perform pressing stage control of the laminated PCB board based on the pressing optimization strategy.

[0017] The technical solution provided in this application has at least the following technical effects or advantages: by realizing the technical goal of intelligent optimization and adaptive control of dynamic multi-node parameters in the pressing process, it achieves the technical effects of improving pressing consistency, reducing defect rate and significantly improving the overall quality of finished products.

[0018] The above description is merely an overview of the technical solution of this application. To enable a clearer understanding of the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0020] Figure 1 is a flowchart illustrating an automated production method for multilayer PCBs according to this application.

[0021] Figure 2 is a schematic diagram of an automated production system for multilayer PCB boards according to this application.

[0022] Figure labeling: 1. Multi-layer alignment processing module; 2. Confidential compression parameter mining module; 3. Combined decision-making module; 4. Multi-dimensional prediction and evaluation module; 5. Multi-objective optimization module; 6. Expansion optimization module. Detailed Implementation

[0023] This application provides an automated production method and system for multilayer PCBs, solving the technical problems in existing technologies where fixed lamination process parameters and a lack of multi-node collaborative optimization prevent effective adaptation to batch differences in materials, variations in the number of layers, and structural deviations, further affecting the bonding strength, appearance flatness, and overall quality stability of the finished PCB. The method achieves the technical goal of dynamic multi-node parameter intelligent optimization and adaptive control in the lamination process, resulting in improved lamination consistency, reduced defect rates, and significantly enhanced overall product quality.

[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0025] Example 1, please refer to Figure 1. This application provides an automated production method for multilayer PCBs, which is applied to an automated production system for multilayer PCBs. Specifically, it includes the following steps: S1: Perform multilayer alignment processing on the PCB to be laminated to obtain a laminated PCB that meets the predetermined alignment accuracy.

[0026] Furthermore, this application also includes: obtaining PCB design information of the PCB board to be laminated; performing multi-layer alignment processing on the PCB board to be laminated according to the PCB design information to obtain an initial alignment board; constructing a standard alignment model according to the PCB design information, and evaluating the alignment accuracy of the initial alignment board according to the standard alignment model to obtain an alignment accuracy coefficient; if the alignment accuracy coefficient meets the predetermined alignment accuracy, identifying the initial alignment board as the laminated PCB board.

[0027] Furthermore, this application also includes: if the alignment accuracy coefficient does not meet the predetermined alignment accuracy, the initial alignment board is corrected for alignment deviation according to the standard alignment model to obtain the laminated PCB board.

[0028] Specifically, obtaining the PCB design information of the PCB to be laminated refers to extracting the electronic design files or drawings of the PCB to be laminated before laminating the multilayer printed circuit boards. This includes the circuit diagrams of each layer, alignment marks, the location of interlayer interconnects, and the number of layers. A PCB is a printed circuit board, and its design information comes from Gerber files exported from electronic design automation software, drilling data, etc. This design information determines how different layers should be stacked to ensure correct circuit conduction. For example, for an 8-layer multilayer PCB, the design information would include the circuit layout and reference positioning holes for each layer from layer 1 to layer 8.

[0029] Next, based on the PCB design information, multi-layer alignment processing is performed on the PCB boards to be laminated to obtain the initial alignment board. This means using the extracted PCB drawing data to control alignment equipment (such as optical alignment system, positioning platform, etc.) to stack multiple unlaminated single-layer PCB boards, so that the lines, holes and pads between the layers overlap as much as possible, in order to prevent conduction errors or performance defects in the subsequent lamination process.

[0030] Then, based on the PCB design information, a standard alignment model is constructed, and the alignment accuracy of the initial alignment board is evaluated based on the standard alignment model to obtain the alignment accuracy coefficient. This means that by analyzing the standard alignment points of each layer in the design drawings, a theoretical standard alignment reference frame is generated. Then, the actual interlayer alignment of the initial alignment board is compared with the standard model, and an index value reflecting the alignment accuracy is calculated, namely the alignment accuracy coefficient, which reflects the degree of offset of each key alignment point.

[0031] Finally, if the alignment accuracy coefficient meets the predetermined alignment accuracy, the initial alignment board is marked as a laminated PCB board. This means that if the evaluated alignment accuracy coefficient is within the set acceptable threshold range, such as less than 75 micrometers, the initial alignment board is considered to meet the requirements and can be used as a standard laminated board for subsequent lamination processing; otherwise, the alignment needs to be readjusted. A laminated PCB board signifies that the multilayer board assembly has completed high-precision alignment and entered the lamination preparation stage, and the alignment degree of its layer circuits and conductive vias has met the process requirements.

[0032] If the alignment accuracy coefficient does not meet the predetermined alignment accuracy, it means that during the alignment process of a multilayer PCB board, the alignment accuracy coefficient obtained through evaluation fails to meet the pre-set accuracy requirements, indicating that there is a deviation in the alignment between the board layers. The alignment accuracy coefficient is a numerical value used to quantify the alignment accuracy, representing the actual degree of offset between each alignment point; while the predetermined alignment accuracy is the upper limit tolerance defined in the manufacturing process, representing the maximum acceptable alignment error without affecting the lamination quality and electrical performance.

[0033] Next, the initial alignment board is corrected for alignment deviations based on the standard alignment model. This involves using the standard alignment model built in the original design to compare and calculate the current alignment status, identify the offsets between each layer relative to the standard position, and perform specific correction operations accordingly. The standard alignment model is an idealized multi-layer alignment reference model, established based on key alignment points, reference hole positions, and inter-layer structures defined in the PCB design file, representing a theoretically perfectly aligned state. In the actual correction process, alignment adjustment systems such as high-precision positioning platforms or machine vision systems are used to fine-tune the translation and rotation positions of each layer to achieve an arrangement as close as possible to the standard model.

[0034] Finally, after deviation correction, a laminated PCB board is obtained. The initial alignment board has achieved the required alignment accuracy and meets the interlayer alignment standard before lamination, thus it can be used as the formal lamination input. A laminated PCB board refers to a multi-layer board assembly that has undergone qualified alignment treatment, and its structure is stable, meeting the conditions for entering the hot pressing process stage.

[0035] S2: Based on the basic feature vector of the laminated PCB board, perform reliable mining of multi-node lamination parameters to construct a reliable space for multi-node lamination.

[0036] Furthermore, this application also includes: performing a historical retrieval of pressing control based on the basic feature vector to obtain a set of historical pressing control schemes; performing reliable mining of vacuum preheating node parameters based on the set of historical pressing control schemes to construct a reliable vacuum preheating space; performing reliable mining of hot pressing node parameters based on the set of historical pressing control schemes to construct a reliable hot pressing space; performing reliable mining of cold pressing and shaping node parameters based on the set of historical pressing control schemes to construct a reliable cold pressing and shaping space; and generating the multi-node pressing reliable space based on the reliable vacuum preheating space, the reliable hot pressing space, and the reliable cold pressing and shaping space.

[0037] Furthermore, this application also includes: classifying vacuum preheating node parameters according to the set of historical compression control schemes to obtain multiple vacuum preheating parameter regions; parsing trigger intervals according to the multiple vacuum preheating parameter regions to obtain multiple preheating parameter trigger intervals; evaluating the endpoint frequent states of the multiple preheating parameter trigger intervals according to the multiple vacuum preheating parameter regions to obtain multiple endpoint frequent state coefficients; performing endpoint reliability detection on the multiple preheating parameter trigger intervals based on the multiple endpoint frequent state coefficients and endpoint frequent state thresholds to determine an untrusted endpoint set; and performing endpoint correction on the multiple preheating parameter trigger intervals based on the multiple vacuum preheating parameter regions, the untrusted endpoint set, and the endpoint frequent state thresholds to generate the vacuum preheating reliability space.

[0038] Specifically, retrieving historical lamination control schemes based on fundamental feature vectors involves using the feature vector information of the current PCB to be laminated to search for historical production records with similar parameters or structures in an existing production database. The fundamental feature vector is a numerical representation of multiple attributes of the PCB before lamination, such as structure, material, number of layers, thickness, and tolerances, used to describe the typical properties of the PCB. For example, if the current PCB has an 8-layer structure, a total thickness of 1.6 mm, and contains mixed dielectric materials, retrieving all past production records with these parameters will create a historical lamination control scheme set—a collection of lamination parameter configurations actually used in similar products in the past.

[0039] Next, based on the historical set of lamination control schemes, the reliability of vacuum preheating node parameters is mined to construct a reliable vacuum preheating space. This involves extracting key parameters related to the vacuum preheating process from historical data, evaluating and screening their reliability, and ultimately forming a set of vacuum preheating parameters with high reliability suitable for the current PCB board. The vacuum preheating node is the first stage in the lamination process, typically responsible for expelling air between boards and gradually increasing the temperature. Its parameters may include the heating rate, preheating temperature, and vacuum time.

[0040] Then, based on the historical set of pressing control schemes, reliable parameters for hot pressing nodes are mined to construct a reliable hot pressing space. This means further extracting parameter data related to the hot pressing stage from historical schemes and using algorithms to identify configurations with high repeatability and stable effects, thus constructing a reliable set of hot pressing parameters. A hot pressing node refers to the process node that formally enters the pressurization and heating stage after preheating, and its key parameters include hot pressing temperature, pressure value, and holding time.

[0041] Subsequently, during the cooling and shaping process of lamination, reliable parameters of the cold-pressing shaping nodes are mined based on the historical lamination control scheme set to construct a reliable cold-pressing shaping space. This involves extracting stable and usable cold-pressing parameters based on historical data. The cold-pressing shaping nodes are used to cool the PCB board after lamination and maintain its shape to prevent warping. Typical parameters include cooling rate, cold pressing time, and cold pressing pressure.

[0042] Based on the reliable space of vacuum preheating, reliable space of hot pressing and reliable space of cold pressing and shaping, a multi-node pressing reliable space is generated. This means that the reliable parameter sets of each process node are integrated to form a comprehensive and structured set of reliable parameters, which lays the foundation for subsequent process optimization and decision-making.

[0043] Furthermore, based on the historical set of pressing control schemes, the vacuum preheating node parameters are classified to obtain multiple vacuum preheating parameter zones. This involves classifying and summarizing the parameter data related to vacuum preheating in different schemes according to feature similarity or application frequency, dividing them into different parameter zones. Next, trigger interval analysis is performed based on these multiple vacuum preheating parameter zones to obtain multiple preheating parameter trigger intervals. This involves identifying the critical value intervals within each parameter zone that are frequently used in actual production.

[0044] Subsequently, based on multiple vacuum preheating parameter zones, frequent endpoint state evaluations were performed on multiple preheating parameter trigger intervals to obtain multiple frequent endpoint state coefficients. This involves statistically analyzing the frequency of occurrence of the start and end points of each interval to quantify their activity level in historical data. The frequency of the endpoints of each preheating parameter trigger interval appearing within the corresponding vacuum preheating parameter zone is defined as the frequent endpoint state coefficient.

[0045] Based on the frequent state coefficients of multiple endpoints, endpoint reliability detection is performed on the trigger intervals of multiple preheating parameters according to the frequent state threshold of the endpoints to determine the set of untrusted endpoints. This means that the frequent state coefficients of all endpoints are filtered by a threshold. Endpoints with a value lower than the threshold are judged to be unstable and are not included in the reliable parameters. Finally, an untrusted endpoint set is formed. That is, the endpoint with a frequent state coefficient of the endpoint is less than the frequent state threshold of the endpoint. The endpoint corresponding to the frequent state coefficient of the endpoint is an untrusted endpoint.

[0046] Finally, based on multiple vacuum preheating parameter zones, endpoint correction is performed on multiple preheating parameter trigger intervals according to the set of untrusted endpoints and the threshold for frequent endpoint states, generating a reliable vacuum preheating space. This involves using identified reliable endpoints to replace or adjust untrusted intervals, thereby forming a reliable vacuum preheating space with clear data sources and high parameter stability. If the frequency of a given interval exceeds the threshold in subsequent evaluations, the replacement is successful and the interval is included in the reliable space. Table 1 shows the data table for constructing the multi-node pressing reliable space.

[0047] Table 1: Data Table for Constructing a Reliable Space for Multi-Node Pressing | Node Stage | Number of Historical Schemes (units) | Main Parameter Type | Parameter Reliability Range | Example | Reliable Space Construction Method | |---|---|---|---|---| | Vacuum Preheating Node 120 | Temperature, Time, Vacuum Degree | Temperature 120-150°C, Time 5-10 minutes, Vacuum Degree 95%-99% | Constructed based on historical scheme frequency analysis and anomaly elimination | | Hot Pressing Node 150 | Temperature, Pressure, Time | Temperature 160-190°C, Pressure 1.5-2.0 MPa, Time 15-25 minutes | Constructed based on parameter stability and result quality correlation | | Cold Pressing and Shaping Node 100 | Pressure, Cooling Time | Pressure 1.0-1.2 MPa, Cooling Time 8-12 minutes | Constructed based on historical distribution and defect rate back-calculation | | Multi-Node Integration Result 370 | Comprehensive Multi-Node Parameters | Cross-Node Parameter Combination | Meets Process Matching Constraints | Vacuum Preheating, Hot Pressing, and Cold Pressing and Shaping Reliable Space Fusion Generation Table S3: Based on the multi-node pressing confidence space, the multi-node pressing parameter combination decision is made on the laminated PCB board to obtain the first pressing control domain.

[0048] Specifically, multi-node lamination parameter combination decision-making for laminated PCBs based on the multi-node lamination reliability space refers to selecting the most suitable set of parameters for multiple lamination process nodes within the constructed multi-node lamination reliability space, based on the specific structural characteristics of the laminated PCBs. This includes vacuum preheating, hot lamination, and cold pressing shaping. Each node has multiple reliable parameter ranges, and the combination decision-making process selects an optimal set of parameters from these ranges. This ensures that the parameters coordinate and work together throughout the entire lamination process, affecting the current PCBs and guaranteeing that the parameters meet product quality requirements while optimizing energy consumption, efficiency, and equipment load. The parameter combination forms the control parameter range used by the PCBs in the initial lamination stage, obtaining the first lamination control domain. The lamination control domain is a multi-dimensional set of parameters that defines the specific executable temperature, pressure, and time ranges for each process node, determining the control behavior boundaries of the lamination equipment during operation.

[0049] S4: Perform multi-dimensional prediction and evaluation of the first compression control domain based on the compression evaluation multi-model, and construct a compression evaluation map.

[0050] Furthermore, this application also includes: performing simulated pressing control on the laminated PCB board according to each pressing control scheme within the first pressing control domain to obtain multiple pressing simulation data; the pressing evaluation multiple models include an adhesive strength evaluation model, a pressing defect evaluation model, and a flatness uniformity evaluation model; inputting the multiple pressing simulation data into the adhesive strength evaluation model to obtain an adhesive strength evaluation sequence; inputting the multiple pressing simulation data into the pressing defect evaluation model to obtain a pressing defect evaluation sequence; inputting the multiple pressing simulation data into the flatness uniformity evaluation model to obtain a flatness uniformity evaluation sequence; mapping the first pressing control domain according to the adhesive strength evaluation sequence, the pressing defect evaluation sequence, and the flatness uniformity evaluation sequence to generate the pressing evaluation map.

[0051] Specifically, the simulated pressing control of the laminated PCB board is performed according to each pressing control scheme within the first pressing control domain to obtain multiple pressing simulation data. This refers to simulating the pressing process in a virtual environment based on multiple parameter combinations included in the first pressing control domain. Each pressing control scheme is used to predict the possible effects of each scheme in actual pressing. Simulated pressing control uses digital modeling techniques, such as finite element analysis or digital twin systems, to simulate the impact of the real pressing process on the PCB board, thereby obtaining the pressing simulation data corresponding to each scheme.

[0052] Next, the lamination evaluation models include an adhesive strength evaluation model, a lamination defect evaluation model, and a flatness and uniformity evaluation model. The adhesive strength evaluation model predicts the bonding strength between multiple layers of a PCB and typically relies on factors such as stress at the bonding interface and lamination temperature variations. The lamination defect evaluation model identifies potential defects in the simulation, such as bubbles, delamination, and resin voids, determined by analyzing material response and strain fields. The flatness and uniformity evaluation model assesses the flatness of the laminated board surface and the consistency of layer thickness, measured by surface deformation, warpage, or thickness uniformity differences.

[0053] Then, multiple compression simulation data are input into the bond strength evaluation model to obtain the bond strength evaluation sequence. This means that the obtained simulation data are input into the bond strength model one by one, the bond strength value corresponding to each scheme is calculated, and they are arranged into a sequence in order.

[0054] Similarly, multiple pressing simulation data are input into the pressing defect evaluation model to obtain the pressing defect evaluation sequence. This involves using the same data to evaluate the defect risk corresponding to each set of parameters. Each set of solutions will obtain a defect score or defect probability, such as a value from 0 to 1. The higher the value, the greater the defect risk. These values ​​are arranged in order to form the pressing defect evaluation sequence.

[0055] Furthermore, multiple pressing simulation data are input into the flatness and uniformity evaluation model to obtain a flatness and uniformity evaluation sequence. This is achieved by evaluating the surface uniformity and layer thickness consistency of each set of pressing control parameters under the simulation environment, thereby obtaining a numerical evaluation result.

[0056] Finally, the first pressing control domain is mapped according to the bond strength evaluation sequence, pressing defect evaluation sequence, and flatness uniformity evaluation sequence to generate a pressing evaluation map. The three performance indicators are visualized and integrated in the parameter space, and a mapping relationship is established between each set of control parameters and its corresponding performance evaluation results. This allows for display in two-dimensional or three-dimensional form, such as coordinate axes representing strength, defect rate, and flatness, with each point corresponding to a specific control scheme. This facilitates the rapid selection of the optimal or compromise-optimal pressing parameter configuration.

[0057] S5: Construct a pressing quality analysis space, and perform multi-objective optimization on the first pressing control domain based on the pressing evaluation map and the pressing quality analysis space to obtain a second pressing control domain.

[0058] Furthermore, this application also includes: constructing a pressing quality evaluation coordinate system, wherein the pressing quality evaluation coordinate system includes an adhesion strength axis, a pressing defect axis, a flatness uniformity axis, and a pressing quality axis; configuring vector weight relationships for the pressing quality evaluation coordinate system by multiple pressing quality evaluation experts to obtain multiple evaluation weight relationship vectors; performing reliable calculation of the weights of each dimension based on the multiple evaluation weight relationship vectors to establish a reliable weight relationship vector; and performing weighted optimization on the pressing quality evaluation coordinate system based on the reliable weight relationship vectors to generate the pressing quality analytical space.

[0059] Furthermore, this application also includes: generating a pressing evaluation constraint matrix based on bond strength evaluation constraints, pressing defect evaluation constraints, and flatness uniformity evaluation constraints; performing initial optimization on the first pressing control domain based on the pressing evaluation map and the pressing evaluation constraint matrix to obtain an initial pressing optimization domain; calculating the pressing quality of each pressing control scheme within the initial pressing optimization domain based on the pressing quality analytical space to obtain a pressing quality map; and performing pressing quality optimization on the initial pressing optimization domain based on the pressing quality map and a pressing quality threshold to generate a second pressing control domain.

[0060] Specifically, a lamination quality evaluation coordinate system is constructed to comprehensively measure the lamination effect through a multi-dimensional evaluation system, including four evaluation dimensions: bond strength axis, lamination defect axis, flatness and uniformity axis, and lamination quality axis. The bond strength axis reflects the firmness of the bond between the layers of the PCB board, measuring its tensile strength in Newtons; the lamination defect axis represents the potential defects that may occur during the lamination process, such as bubbles, voids, and delamination, expressed as defect probability or defect count; the flatness and uniformity axis represents the overall flatness and thickness consistency of the PCB board surface after lamination, reflecting the product appearance and the feasibility of subsequent processing; the lamination quality axis is a summary indicator combining the first three dimensions, used to uniformly evaluate the overall performance of each lamination scheme, ultimately representing the lamination quality with a quantifiable value.

[0061] Next, multiple bonding quality evaluation experts configured the vector weight relationship of the bonding quality evaluation coordinate system. This means that a team of experts with bonding experience assigned weights to each evaluation dimension in the coordinate system, reflecting the importance of each indicator in different application scenarios. The weights can be positive or negative. The bonding strength axis and the flatness uniformity axis are configured with positive weights, and the bonding defect axis is configured with negative weights. Then, multiple evaluation weight relationship vectors are obtained by using the formula: Bonding Quality Coefficient = Bonding Strength Coefficient × Bonding Strength Weight + Bonding Defect Coefficient × Bonding Defect Weight + Flatness Uniformity Coefficient × Flatness Uniformity Weight. Each evaluation weight relationship vector includes [Bonding Strength Weight, Bonding Defect Weight, Flatness Uniformity Weight].

[0062] Then, based on multiple evaluation weight relationship vectors, the weights of each dimension are reliably calculated to establish a reliable weight relationship vector. This is achieved by statistically analyzing and verifying the consistency of multiple weight vectors provided by experts, such as by calculating the mean, standard deviation, and confidence interval, and removing outliers or samples with large deviations, thereby forming a weight vector with high credibility, namely, a reliable weight relationship vector.

[0063] Subsequently, the coordinate system for evaluating pressing quality is weighted and optimized based on the reliable weight relationship vector, generating an analytical space for pressing quality. This involves using the reliable weight relationship vector to adjust the weights of each axis in the original coordinate system, ensuring that the influence of different dimensions accurately reflects their overall value in subsequent pressing quality calculations. The optimized coordinate system then becomes the analytical space for pressing quality, possessing a clear weight allocation structure, which can be used to uniformly score and rank different pressing schemes.

[0064] Next, based on the bond strength evaluation constraints, pressing defect evaluation constraints, and flatness uniformity evaluation constraints, a pressing evaluation constraint matrix is ​​generated, which mathematically represents the three quality requirements as limitations. The pressing evaluation constraint matrix defines whether each pressing control scheme meets the quality baseline.

[0065] Next, based on the pressing evaluation map, the first pressing control domain is initially optimized according to the pressing evaluation constraint matrix to obtain the initial pressing optimization domain. This means that all schemes in the map are substituted into the constraint matrix one by one for screening, and only pressing schemes that simultaneously meet all quality baseline requirements are retained to form the initial pressing optimization domain, which provides a screening basis for further pressing optimization.

[0066] Then, based on the compression quality analytical space, the compression quality of each compression control scheme within the initial optimization domain is calculated to obtain a compression quality map. This involves using the optimized coordinate system to evaluate the comprehensive score of each compression control scheme, forming a new distribution map of all scheme scores in space. Each point in the new distribution map not only meets the basic quality requirements but also provides a comprehensive ranking basis for quality.

[0067] Finally, based on the lamination quality map, the initial lamination optimization domain is optimized according to the lamination quality threshold to generate a second lamination control domain. This second control domain is formed by further filtering out schemes with comprehensive scores higher than the set threshold from the lamination quality map. The parameter combinations in the second lamination control domain are considered to be lamination configurations that simultaneously meet the minimum requirements and the quality optimization criteria, and are suitable for the production of high-end or high-precision PCB boards.

[0068] S6: Expand and optimize the second pressing control domain according to the multi-node pressing confidence space, the pressing evaluation multi-model and the pressing quality analysis space to obtain a pressing optimization strategy, and control the pressing stage of the laminated PCB board according to the pressing optimization strategy.

[0069] Furthermore, this application also includes: randomly mutating the second compression control domain according to the multi-node compression confidence space to obtain a first compression control variation domain; performing compression evaluation optimization on the first compression control variation domain based on the compression evaluation constraint matrix and the compression evaluation multi-model to obtain a second compression control variation domain; performing compression quality optimization on the second compression control variation domain based on the compression quality threshold and the compression quality analytical space to obtain a third compression control variation domain; expanding the second compression control domain according to the third compression control variation domain to obtain a third compression control domain, and performing iterative optimization of compression quality according to the third compression control domain to generate the compression optimization strategy.

[0070] Specifically, the first variation domain of the pressing control is obtained by randomly mutating the second pressing control domain based on the multi-node pressing confidence space. This means that, based on the existing parameters of the second pressing control domain, some pressing parameters, such as vacuum preheating temperature, hot pressing time, and cold pressing pressure, are randomly adjusted within a reasonable value range provided by the multi-node pressing confidence space, thereby generating a new set of parameter combinations. Random mutation introduces diversity while retaining the existing excellent characteristics, in order to explore potentially better pressing schemes. The first variation domain is the set of variation schemes, each of which originates from the second pressing control domain but has subtle differences.

[0071] Next, based on the bonding evaluation constraint matrix, the first variation domain of bonding control is optimized according to the multiple bonding evaluation models to obtain the second variation domain of bonding control. This involves sequentially inputting each variation scheme in the first variation domain into the bond strength evaluation model, the bonding defect evaluation model, and the flatness and uniformity evaluation model for analysis, and then filtering them using the bonding evaluation constraint matrix to ensure that the variation schemes still meet the minimum quality requirements. After this round of filtering, the second variation domain of bonding control is obtained, and its members are more in line with the quality standards than those in the first variation domain.

[0072] Then, based on the pressing quality threshold, the pressing quality is optimized in the second variation domain of pressing control according to the pressing quality analytical space to obtain the third variation domain of pressing control. This involves further substituting the solutions in the second variation domain into the pressing quality analytical space, calculating a comprehensive score through weight optimization, and selecting solutions with scores higher than a set threshold. Solutions in the third variation domain not only meet the basic constraints but also possess high overall quality.

[0073] Subsequently, the second compression control domain is expanded based on the third variation domain of the compression control to obtain the third compression control domain. This involves adding high-quality solutions from the third variation domain to the original second compression control domain, thereby expanding the solution pool to include more potential high-quality combinations. This will help provide more choices in future optimization processes and increase the probability of finding the optimal solution.

[0074] Next, based on the third pressing control domain, iterative optimization of pressing quality is carried out to generate a pressing optimization strategy. This means that, based on the third pressing control domain, the quality of the scheme is continuously improved by repeatedly applying steps such as variation, screening, and quality calculation, and finally a set of pressing optimization strategies is formed. This clarifies the parameter combinations and adjustment methods that should be given priority in different pressing stages, providing a basis for decision-making in actual production.

[0075] Finally, the lamination stage control of the laminated PCB board is carried out according to the lamination optimization strategy. This involves applying the optimization strategy to the actual PCB board lamination process and precisely controlling the process parameters of each stage, such as vacuum preheating, hot pressing, and cold pressing, to ensure the stability and consistency of product quality.

[0076] In summary, the automated production method for multilayer PCBs provided in this application has the following technical effects: by realizing the technical goal of intelligent optimization and adaptive control of dynamic multi-node parameters in the lamination process, it achieves the technical effects of improving lamination consistency, reducing defect rate, and significantly improving the overall quality of finished products.

[0077] Example 2: Based on the same inventive concept as the automated production method for multilayer PCBs described in the previous example, this application also provides an automated production system for multilayer PCBs, as shown in Figure 2. The system includes: a multilayer alignment processing module 1, used to perform multilayer alignment processing on the PCBs to be laminated to obtain a laminated PCB that meets a predetermined alignment accuracy; a lamination parameter reliability mining module 2, used to perform multi-node lamination parameter reliability mining based on the basic feature vectors of the laminated PCBs to construct a multi-node lamination reliability space; and a combination decision module 3, used to perform multi-node lamination parameter combination decisions on the laminated PCBs based on the multi-node lamination reliability space to obtain... The module 4 is a first pressing control domain; a multi-dimensional prediction and evaluation module 4 is used to perform multi-dimensional prediction and evaluation of the first pressing control domain according to the pressing evaluation multi-model, and construct a pressing evaluation map; a multi-objective optimization module 5 is used to construct a pressing quality analysis space, and perform multi-objective optimization of the first pressing control domain based on the pressing evaluation map and the pressing quality analysis space to obtain a second pressing control domain; an expansion optimization module 6 is used to expand and optimize the second pressing control domain according to the multi-node pressing confidence space, the pressing evaluation multi-model and the pressing quality analysis space, obtain a pressing optimization strategy, and perform pressing stage control of the laminated PCB board according to the pressing optimization strategy.

[0078] Furthermore, the automated production system for multilayer PCBs is also used for: performing a historical retrieval of pressing control based on the basic feature vector to obtain a set of historical pressing control schemes; performing reliable mining of vacuum preheating node parameters based on the set of historical pressing control schemes to construct a reliable vacuum preheating space; performing reliable mining of hot pressing node parameters based on the set of historical pressing control schemes to construct a reliable hot pressing space; performing reliable mining of cold pressing and shaping node parameters based on the set of historical pressing control schemes to construct a reliable cold pressing and shaping space; and generating the reliable multi-node pressing space based on the reliable vacuum preheating space, the reliable hot pressing space, and the reliable cold pressing and shaping space.

[0079] Furthermore, the automated production system for multilayer PCBs is further configured to: classify vacuum preheating node parameters according to the historical set of lamination control schemes to obtain multiple vacuum preheating parameter regions; analyze trigger intervals according to the multiple vacuum preheating parameter regions to obtain multiple preheating parameter trigger intervals; evaluate the endpoint frequent states of the multiple preheating parameter trigger intervals according to the multiple vacuum preheating parameter regions to obtain multiple endpoint frequent state coefficients; perform endpoint reliability detection on the multiple preheating parameter trigger intervals based on the multiple endpoint frequent state coefficients and endpoint frequent state thresholds to determine an untrusted endpoint set; and perform endpoint correction on the multiple preheating parameter trigger intervals based on the multiple vacuum preheating parameter regions, the untrusted endpoint set, and the endpoint frequent state thresholds to generate the vacuum preheating reliability space.

[0080] Furthermore, the automated production system for multilayer PCBs is further configured to: perform simulated pressing control on the laminated PCBs according to each pressing control scheme within the first pressing control domain, thereby obtaining multiple pressing simulation data; the pressing evaluation multi-model includes an adhesive strength evaluation model, a pressing defect evaluation model, and a flatness uniformity evaluation model; input the multiple pressing simulation data into the adhesive strength evaluation model to obtain an adhesive strength evaluation sequence; input the multiple pressing simulation data into the pressing defect evaluation model to obtain a pressing defect evaluation sequence; input the multiple pressing simulation data into the flatness uniformity evaluation model to obtain a flatness uniformity evaluation sequence; and map the first pressing control domain according to the adhesive strength evaluation sequence, the pressing defect evaluation sequence, and the flatness uniformity evaluation sequence to generate the pressing evaluation map.

[0081] Furthermore, the automated production system for multilayer PCBs is also used for: generating a bonding evaluation constraint matrix based on bonding strength evaluation constraints, bonding defect evaluation constraints, and flatness uniformity evaluation constraints; performing initial optimization on the first bonding control domain based on the bonding evaluation map and the bonding evaluation constraint matrix to obtain an initial bonding optimization domain; calculating the bonding quality of each bonding control scheme within the initial bonding optimization domain based on the bonding quality analytical space to obtain a bonding quality map; and performing bonding quality optimization on the initial bonding optimization domain based on the bonding quality map and a bonding quality threshold to generate a second bonding control domain.

[0082] Furthermore, the automated production system for multilayer PCBs is further configured to: randomly mutate the second pressing control domain according to the multi-node pressing confidence space to obtain a first pressing control mutation domain; optimize the first pressing control mutation domain according to the pressing evaluation constraint matrix and the pressing evaluation multi-model to obtain a second pressing control mutation domain; optimize the pressing quality of the second pressing control mutation domain according to the pressing quality analytical space based on the pressing quality threshold to obtain a third pressing control mutation domain; expand the second pressing control domain according to the third pressing control mutation domain to obtain a third pressing control domain, and iteratively optimize the pressing quality according to the third pressing control domain to generate the pressing optimization strategy.

[0083] Furthermore, the automated production system for multilayer PCBs also includes: constructing a lamination quality evaluation coordinate system, which includes an adhesion strength axis, a lamination defect axis, a flatness uniformity axis, and a lamination quality axis; configuring vector weight relationships for the lamination quality evaluation coordinate system by multiple lamination quality evaluation experts to obtain multiple evaluation weight relationship vectors; performing reliable weight calculations for each dimension based on the multiple evaluation weight relationship vectors to establish reliable weight relationship vectors; and performing weighted optimization on the lamination quality evaluation coordinate system based on the reliable weight relationship vectors to generate the lamination quality analytical space.

[0084] Furthermore, the automated production system for multilayer PCBs is also used for: obtaining PCB design information of the PCB to be laminated; performing multilayer alignment processing on the PCB to be laminated according to the PCB design information to obtain an initial alignment board; constructing a standard alignment model according to the PCB design information, and evaluating the alignment accuracy of the initial alignment board according to the standard alignment model to obtain an alignment accuracy coefficient; if the alignment accuracy coefficient meets the predetermined alignment accuracy, identifying the initial alignment board as the laminated PCB.

[0085] Furthermore, the automated production system for multilayer PCBs is also used to: if the alignment accuracy coefficient does not meet the predetermined alignment accuracy, correct the alignment deviation of the initial alignment board according to the standard alignment model to obtain the laminated PCB.

[0086] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The automated production method and specific example for multilayer PCBs in the foregoing embodiment one are also applicable to the automated production system for multilayer PCBs in this embodiment. Through the foregoing detailed description of the automated production method for multilayer PCBs, those skilled in the art can clearly understand the automated production system for multilayer PCBs in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0087] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0088] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. An automated production method for multilayer PCBs, characterized in that, The method includes: performing multi-layer alignment processing on the PCB board to be laminated to obtain a laminated PCB board that meets a predetermined alignment accuracy; performing multi-node lamination parameter reliability mining based on the basic feature vector of the laminated PCB board to construct a multi-node lamination reliability space; performing multi-node lamination parameter combination decision-making on the laminated PCB board based on the multi-node lamination reliability space to obtain a first lamination control domain; performing multi-dimensional prediction evaluation on the first lamination control domain based on a lamination evaluation multi-model to construct a lamination evaluation map; constructing a lamination quality analysis space, and performing multi-objective optimization on the first lamination control domain based on the lamination evaluation map and the lamination quality analysis space to obtain a second lamination control domain; performing expansion optimization on the second lamination control domain based on the multi-node lamination reliability space, the lamination evaluation multi-model, and the lamination quality analysis space to obtain a lamination optimization strategy, and performing lamination stage control on the laminated PCB board based on the lamination optimization strategy.

2. The automated production method for multilayer PCBs as described in claim 1, characterized in that, Based on the fundamental feature vectors of the laminated PCB board, a reliable multi-node pressing parameter mining process is performed to construct a reliable multi-node pressing space. This process includes: retrieving pressing control history based on the fundamental feature vectors to obtain a set of historical pressing control schemes; mining reliable vacuum preheating node parameters based on the set of historical pressing control schemes to construct a reliable vacuum preheating space; mining reliable hot pressing node parameters based on the set of historical pressing control schemes to construct a reliable hot pressing space; mining reliable cold pressing and shaping node parameters based on the set of historical pressing control schemes to construct a reliable cold pressing and shaping space; and generating the reliable multi-node pressing space based on the reliable vacuum preheating space, the reliable hot pressing space, and the reliable cold pressing and shaping space.

3. The automated production method for multilayer PCBs as described in claim 2, characterized in that, Based on the historical compression control scheme set, a reliable vacuum preheating node parameter mining process is performed to construct a reliable vacuum preheating space. This process includes: classifying vacuum preheating node parameters according to the historical compression control scheme set to obtain multiple vacuum preheating parameter regions; parsing trigger intervals according to the multiple vacuum preheating parameter regions to obtain multiple preheating parameter trigger intervals; evaluating the endpoint frequent states of the multiple preheating parameter trigger intervals according to the multiple vacuum preheating parameter regions to obtain multiple endpoint frequent state coefficients; performing endpoint reliability detection on the multiple preheating parameter trigger intervals based on the multiple endpoint frequent state coefficients and endpoint frequent state thresholds to determine an untrusted endpoint set; and performing endpoint correction on the multiple preheating parameter trigger intervals based on the multiple vacuum preheating parameter regions, the untrusted endpoint set, and the endpoint frequent state thresholds to generate the reliable vacuum preheating space.

4. The automated production method for multilayer PCBs as described in claim 1, characterized in that, The first pressing control domain is evaluated using a multi-dimensional prediction model based on a pressing evaluation model to construct a pressing evaluation map. This includes: performing simulated pressing control on the laminated PCB board according to various pressing control schemes within the first pressing control domain to obtain multiple pressing simulation data; the pressing evaluation model includes an adhesive strength evaluation model, a pressing defect evaluation model, and a flatness and uniformity evaluation model; inputting the multiple pressing simulation data into the adhesive strength evaluation model to obtain an adhesive strength evaluation sequence; inputting the multiple pressing simulation data into the pressing defect evaluation model to obtain a pressing defect evaluation sequence; inputting the multiple pressing simulation data into the flatness and uniformity evaluation model to obtain a flatness and uniformity evaluation sequence; mapping the first pressing control domain according to the adhesive strength evaluation sequence, the pressing defect evaluation sequence, and the flatness and uniformity evaluation sequence to generate the pressing evaluation map.

5. The automated production method for multilayer PCBs as described in claim 1, characterized in that, Based on the pressing evaluation map and the pressing quality analytical space, a multi-objective optimization is performed on the first pressing control domain to obtain a second pressing control domain. This includes: generating a pressing evaluation constraint matrix based on bond strength evaluation constraints, pressing defect evaluation constraints, and flatness uniformity evaluation constraints; performing initial optimization on the first pressing control domain based on the pressing evaluation map and the pressing evaluation constraint matrix to obtain an initial pressing optimization domain; calculating the pressing quality of each pressing control scheme within the initial pressing optimization domain based on the pressing quality analytical space to obtain a pressing quality map; and performing pressing quality optimization on the initial pressing optimization domain based on the pressing quality map and a pressing quality threshold to generate the second pressing control domain.

6. The automated production method for multilayer PCBs as described in claim 1, characterized in that, The second compression control domain is expanded and optimized based on the multi-node compression confidence space, the compression evaluation multi-model, and the compression quality analytical space to obtain a compression optimization strategy. This includes: randomly mutating the second compression control domain according to the multi-node compression confidence space to obtain a first compression control mutation domain; optimizing the compression evaluation of the first compression control mutation domain based on the compression evaluation constraint matrix and the compression evaluation multi-model to obtain a second compression control mutation domain; optimizing the compression quality of the second compression control mutation domain based on the compression quality threshold and the compression quality analytical space to obtain a third compression control mutation domain; expanding the second compression control domain based on the third compression control mutation domain to obtain a third compression control domain; and iteratively optimizing the compression quality based on the third compression control domain to generate the compression optimization strategy.

7. The automated production method for multilayer PCBs as described in claim 1, characterized in that, Constructing a pressing quality analytical space includes: constructing a pressing quality evaluation coordinate system, which includes an adhesion strength axis, a pressing defect axis, a flatness and uniformity axis, and a pressing quality axis; configuring vector weight relationships for the pressing quality evaluation coordinate system by multiple pressing quality evaluation experts to obtain multiple evaluation weight relationship vectors; reliably calculating the weights of each dimension based on the multiple evaluation weight relationship vectors to establish reliable weight relationship vectors; and optimizing the pressing quality evaluation coordinate system based on the reliable weight relationship vectors to generate the pressing quality analytical space.

8. The automated production method for multilayer PCBs as described in claim 1, characterized in that, Performing multi-layer alignment processing on a PCB board to be laminated to obtain a laminated PCB board that meets a predetermined alignment accuracy includes: obtaining PCB design information of the PCB board to be laminated; performing multi-layer alignment processing on the PCB board to be laminated based on the PCB design information to obtain an initial alignment board; constructing a standard alignment model based on the PCB design information, and evaluating the alignment accuracy of the initial alignment board based on the standard alignment model to obtain an alignment accuracy coefficient; if the alignment accuracy coefficient meets the predetermined alignment accuracy, identifying the initial alignment board as the laminated PCB board.

9. The automated production method for multilayer PCBs as described in claim 8, characterized in that, If the alignment accuracy coefficient does not meet the predetermined alignment accuracy, the initial alignment board is corrected for alignment deviation according to the standard alignment model to obtain the laminated PCB board.

10. An automated production system for multilayer PCBs, characterized in that, The steps of implementing the automated production method for multilayer PCBs according to any one of claims 1 to 9 include: a multilayer alignment processing module for performing multilayer alignment processing on the PCBs to be laminated to obtain a laminated PCB that meets a predetermined alignment accuracy; a lamination parameter reliability mining module for performing multi-node lamination parameter reliability mining based on the basic feature vector of the laminated PCB to construct a multi-node lamination reliability space; a combination decision module for performing multi-node lamination parameter combination decision on the laminated PCB based on the multi-node lamination reliability space to obtain a first lamination control domain; and a multi-dimensional prediction and evaluation module. The system includes a first compression control domain and a second compression control domain. The first compression control domain is evaluated using a multi-dimensional prediction model based on the compression evaluation model, and a compression evaluation map is constructed. The second compression control domain is optimized using a multi-objective optimization module to construct a compression quality analysis space and perform multi-objective optimization based on the compression evaluation map and the compression quality analysis space. The third compression control domain is expanded using a capacity expansion optimization module to obtain a compression optimization strategy and perform compression stage control on the laminated PCB board based on the compression optimization strategy.

Images