A full-process manufacturing optimization method for PCB boards based on the Internet of Things

Abstract

A method for optimizing the entire manufacturing process of a PCB board based on the Internet of Things (IoT) relates to the technical field of the IoT. The method comprises the following steps: constructing three-dimensional models of PCB board manufacturing equipment for each subsequence at each stage, obtaining the vertical coupling relationship between each three-dimensional model and the horizontal coupling relationship between each three-dimensional model, and constructing a model for the entire manufacturing process of the PCB board; obtaining the stability coefficient of each type of indicator of each three-dimensional model; constructing a fuzzy fault early warning comparison database, obtaining a dynamic sign network corresponding to each type of fault of each three-dimensional model and storing it in the fuzzy fault early warning comparison database; performing a fuzzy early warning on each type of indicator of each three-dimensional model within a current monitoring period, generating a fault early warning signal for each three-dimensional model, and visually displaying the fault warning signal through the model for the entire manufacturing process of the PCB board, thereby realizing the prediction of early faults of the PCB board manufacturing equipment before it changes from a normal state to a faulty state.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a full-process manufacturing optimization method for PCB boards based on the Internet of Things. Background Art

[0002] The prior art CN106132098B "A method for manufacturing an LED metal-based circuit board" relates to a method for manufacturing an LED metal-based circuit board, which involves surface treatment of the metal plate → lamination with a thermally conductive insulating film → copper foil lamination → cutting → drilling positioning holes → pre-treatment → circuit pattern forming → etching and film removal → application of white solder mask → exposure → development → character printing and curing → punching → V-cutting → anti-oxidation process → obtaining a finished product. The present invention laminates PP glue on the metal base plate, and the circuit pattern is formed using a printing process or an exposure process. The technical solution of the present invention can be used to mass-produce metal-based circuit boards that integrate a driving power supply and a light source, and the manufactured LEDs and power supplies have good heat dissipation properties.

[0003] The prior art CN115119407B "A three-color LED aluminum-based circuit board and its manufacturing process" includes cutting an aluminum substrate according to a circuit board image completed by the patchwork and grinding the aluminum substrate; placing the aluminum substrate after edge grinding on a detection mechanism for surface detection; opening positioning holes, reserved holes and through-board holes on the aluminum substrate after surface detection; copper plating the aluminum substrate after drilling in a solution tank; irradiating the circuit board image on the aluminum substrate through an automatic exposure machine and cleaning it; tinning the irradiated aluminum substrate in an electroplating tank and cleaning it again; generating a large version of the aluminum-based circuit board and sending it to an AOI scanning device for scanning and detection, thereby improving the control accuracy of important nodes in the production process of aluminum-based circuit boards, thereby improving the yield of aluminum-based circuit boards.

[0004] Existing PCB full-process manufacturing optimization technologies often use threshold monitoring technology for various PCB product indicators. This static analysis method fails to leverage the advantages of online monitoring systems, is difficult to set alarm thresholds, and lacks sufficient evidence. Many factors can influence PCB processing equipment failures, including significant differences in machine performance and the sensitivity of machine production capacity to product mix. For example, the rate at which the same machine processes product A may differ significantly from the rate at which it processes product B. Industrial production processes often exhibit complex characteristics such as multivariate and strong coupling, frequent changes in production boundary conditions, dynamic characteristics that vary with operating conditions, and difficulty describing them using mathematical models, making satisfactory optimization results difficult to achieve. Therefore, the real-time determination of full-process operational indicators to optimize comprehensive enterprise production indicators and the implementation of data- and knowledge-based intelligent dynamic operational optimization of the entire manufacturing process, where process models are difficult to establish, are pressing challenges. Summary of the Invention

[0005] In order to solve the above technical problems, the purpose of the present invention is to provide a full-process manufacturing optimization method for PCB boards based on the Internet of Things, comprising the following steps:

[0006] Step s1: Obtain the manufacturing process information of the PCB board to construct the manufacturing process sequence of the PCB board, and extract the values of various types of indicators of each manufacturing process in the manufacturing process sequence;

[0007] Step s2: Construct a 3D model of the PCB manufacturing equipment for each subsequence at each stage, obtain the vertical coupling relationship between each 3D model and the horizontal coupling relationship between each 3D model, and construct a full-process model of PCB manufacturing;

[0008] Step s3: Obtain the stability coefficients of various indicators of each 3D model in the PCB board manufacturing full process model within several historical monitoring periods, as well as the key indicators corresponding to various types of sudden failures of each 3D model;

[0009] Step s4: constructing a fuzzy fault warning reference database, obtaining the dynamic sign network corresponding to each type of fault in each three-dimensional model and storing it in the fuzzy fault warning reference database;

[0010] Step s5: Perform fuzzy warning on various types of indicators of each three-dimensional model in the current monitoring period, generate fault warning signals for each three-dimensional model, and visualize them through the PCB board manufacturing full process model.

[0011] Furthermore, the manufacturing process information of the PCB board is obtained to construct a manufacturing process sequence of the PCB board, and the process of extracting the values of various types of indicators of each manufacturing process in the manufacturing process sequence includes:

[0012] Acquiring process information of a current PCB manufacturing process, extracting manufacturing unit characteristics based on the manufacturing process information, and dividing the PCB manufacturing process into a plurality of stage subsequences according to the manufacturing unit characteristics;

[0013] Set data monitoring points in each stage subsequence, and use data retrieval to obtain equipment process indicators and product quality indicators at each data monitoring point based on the manufacturing unit characteristics of the corresponding stage subsequence;

[0014] Obtain the planning information of the current PCB board manufacturing process, and obtain the production plan indicators of each data monitoring point based on the planning information of the corresponding stage subsequence;

[0015] The data monitoring points are used to collect various types of indicator values of each stage subsequence and mark the monitoring time, and set the monitoring cycle. The various types of indicator values include equipment process indicator values, product quality indicator values, and production plan indicator values.

[0016] Furthermore, a 3D model of the PCB manufacturing equipment for each subsequence of each stage is constructed, and the vertical coupling relationship between each 3D model and the horizontal coupling relationship between each 3D model are obtained. The process of constructing a full-process model of PCB manufacturing includes:

[0017] Obtaining the physical entities of PCB manufacturing equipment at each stage subsequence in the current PCB manufacturing process and the assembly relationships between the physical entities, performing 3D modeling processing on the physical entities of the PCB manufacturing equipment at each stage subsequence to generate 3D models, and connecting the 3D models according to the assembly relationships to generate an assembly connection layer;

[0018] The numerical values of various indicators of each data monitoring point are obtained for data format preprocessing to generate twin data. The twin data are matched with the three-dimensional model in the assembly connection layer to obtain the vertical coupling relationship of each three-dimensional model and the horizontal coupling relationship between each three-dimensional model. The vertical coupling relationship of each three-dimensional model and the horizontal coupling relationship between each three-dimensional model are superimposed on the assembly connection layer to generate a full-process model of PCB board manufacturing.

[0019] Furthermore, the process of obtaining the vertical coupling relationship of each 3D model includes:

[0020] Obtain the values of various indicators of data monitoring points within several historical monitoring periods and the standard threshold ranges corresponding to the values of various indicators of each type, compare the values of various indicators of each type at various data monitoring points within several historical monitoring periods with the corresponding standard threshold ranges, and mark the indicators whose values are not within the corresponding standard threshold ranges as abnormal;

[0021] Obtain the probability that the product quality indicator and the production plan indicator are in an abnormal state when the equipment process indicator at the current data monitoring point is in an abnormal state within several historical monitoring periods, preset a probability threshold, and if the probability that the product quality indicator is in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the equipment process indicator and the product quality indicator; if the probability that the production plan indicator is in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the equipment process indicator and the production plan indicator;

[0022] By analogy, the vertical coupling relationship and vertical coupling coefficient between various types of indicators at each data monitoring point are obtained.

[0023] Furthermore, the process of obtaining the horizontal coupling relationship between the three-dimensional models includes:

[0024] Obtain a target data monitoring point and other data monitoring points that have an assembly relationship with the target monitoring point, obtain various types of indicator values of the target data monitoring point and other data monitoring points in several historical monitoring periods, obtain the probability that the equipment process indicators, product quality indicators, and production plan indicators of other data monitoring points are in an abnormal state when the equipment process indicators of the target data monitoring point are in an abnormal state in several historical monitoring periods, and if the probability of the equipment process indicators of the other data monitoring points is greater than a probability threshold, generate a horizontal coupling relationship and a horizontal coupling coefficient between the equipment process indicators of the target data monitoring point and the equipment process indicators of the other data monitoring points;

[0025] By analogy, the horizontal coupling relationship and horizontal coupling coefficient between various types of indicators between various data monitoring points are obtained.

[0026] Furthermore, the process of obtaining the stability coefficients of various types of indicators of each three-dimensional model in the PCB board manufacturing full process model within several historical monitoring periods and the key type indicators corresponding to sudden occurrence of several types of failures of each three-dimensional model includes:

[0027] Obtain historical occurrence data of several types of sudden faults of each three-dimensional model in the PCB board manufacturing full-process model within several historical monitoring periods, the historical occurrence data including time series sequences of numerical values of each type of indicator before the sudden occurrence of several types of faults of the three-dimensional model within the historical monitoring period, comparing the time series sequences of numerical values of each type of indicator within the monitoring period with the corresponding standard threshold range, obtaining the frequency of each type of indicator value not being within the corresponding standard threshold range and the fluctuation amplitude of each type of indicator value, obtaining the stability coefficient of each type of indicator based on the frequency and the fluctuation amplitude, and simultaneously obtaining the stability coefficient of each type of indicator when no various types of faults occur in each three-dimensional model, and marking the stability coefficient as the standard stability coefficient;

[0028] Obtain the stability coefficient difference between the stability coefficient of each three-dimensional model when several types of sudden faults occur and the corresponding standard stability coefficient, and screen out the type indicator corresponding to the maximum stability coefficient difference when each three-dimensional model suddenly encounters a certain type of sudden fault as the key type indicator. Similarly, obtain the key type indicator corresponding to each three-dimensional model when several types of sudden faults occur.

[0029] Furthermore, the process of constructing a fuzzy fault warning comparison database, obtaining a dynamic sign network corresponding to each type of fault in each three-dimensional model, and storing the dynamic sign network in the fuzzy fault warning comparison database includes:

[0030] Extract the stability coefficient differences of several types of indicators that have a vertical coupling relationship with the key type indicators corresponding to the type fault of the three-dimensional model, and the stability coefficient differences of several types of indicators of other three-dimensional models that have a horizontal coupling relationship with the key type indicators. According to the stability coefficient differences of the key type indicators corresponding to the type fault, the stability coefficient differences of several types of indicators that have a vertical coupling relationship with the key type indicators, and the stability coefficient differences of several types of indicators of other three-dimensional models that have a horizontal coupling relationship with the key type indicators, construct a dynamic marker network of the type fault. Similarly, obtain the dynamic marker network corresponding to each type of fault of each three-dimensional model and store it in the fuzzy fault warning comparison database.

[0031] Furthermore, a fuzzy early warning is performed on various indicators of each 3D model within the current monitoring period, and a fault early warning signal of each 3D model is generated. The process of visually displaying the PCB board manufacturing full process model includes:

[0032] Obtain the stability coefficient difference of each type of indicator of each three-dimensional model in the current monitoring period, screen out the type indicator corresponding to the maximum stability coefficient difference of each three-dimensional model and mark it as the type indicator to be analyzed, obtain several type indicators that have a vertical coupling relationship with the type indicator to be analyzed and several type indicators of other three-dimensional models that have a horizontal coupling relationship, and construct a marker network to be analyzed based on the stability coefficient difference of the type indicator to be analyzed, the stability coefficient difference of several type indicators that have a vertical coupling relationship with the type indicator to be analyzed, and the stability coefficient difference of several type indicators of other three-dimensional models that have a horizontal coupling relationship with the type indicator to be analyzed;

[0033] Match the sign network to be analyzed with a dynamic sign network in the fuzzy fault warning comparison database, obtain the type index to be analyzed of the sign network to be analyzed and the stability coefficient difference similarity between each type index and the key type index of the dynamic sign network and each type index, use the stability coefficient difference similarity of the type index to be analyzed of the sign network to be analyzed and the stability coefficient difference similarity of each type index as evaluation indexes, and simultaneously obtain the vertical coupling coefficient or horizontal coupling coefficient between the type index to be analyzed and each type index, set the weight index matrix of the evaluation index according to the vertical coupling coefficient or horizontal coupling coefficient between the type index to be analyzed and each type index, preset the warning level, obtain the membership matrix of the evaluation index of each three-dimensional model for the warning level of a dynamic sign network through fuzzy comprehensive evaluation, obtain the warning level of each three-dimensional model for a dynamic sign network according to the membership matrix and the weight index matrix, and so on, obtain the warning level of each three-dimensional model for each dynamic sign network;

[0034] A warning level threshold is preset, and the warning level of each three-dimensional model for each dynamic marker network is compared with the warning level threshold. If there is a three-dimensional model whose warning level for a dynamic network marker is greater than the warning level threshold, the type of fault corresponding to the dynamic network marker is obtained, and a fault warning signal for the three-dimensional model is generated according to the type of fault.

[0035] Compared with the prior art, the present invention has the following beneficial effects: constructing a three-dimensional model of PCB manufacturing equipment for each stage subsequence, obtaining the vertical coupling relationship between each three-dimensional model and the horizontal coupling relationship between each three-dimensional model, constructing a full-process model of PCB manufacturing, obtaining the stability coefficient of each type of indicator of each three-dimensional model in the full-process model of PCB manufacturing within several historical monitoring periods, and the key type indicators corresponding to each type of sudden failure of each three-dimensional model, constructing a fuzzy fault early warning comparison database, obtaining the dynamic sign network corresponding to each type of failure of each three-dimensional model and storing it in the fuzzy fault early warning comparison database, performing fuzzy early warning on each type of indicator of each three-dimensional model within the current monitoring period, generating fault early warning signals for each three-dimensional model, and visually displaying them through the full-process model of PCB manufacturing, thereby realizing the prediction of early-stage failures of PCB manufacturing equipment before it transitions from a normal state to a faulty state, thereby taking preventive measures before the failure of the PCB manufacturing equipment matures. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a schematic diagram of a full-process manufacturing optimization method for PCB boards based on the Internet of Things in an embodiment of the present application. DETAILED DESCRIPTION

[0037] The following is a clear and complete description of the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0038] like Figure 1 As shown, a full-process manufacturing optimization method for PCB boards based on the Internet of Things includes the following steps:

[0039] Step s1: Obtain the manufacturing process information of the PCB board to construct the manufacturing process sequence of the PCB board, and extract the values of various types of indicators of each manufacturing process in the manufacturing process sequence;

[0040] Step s2: Construct a 3D model of the PCB manufacturing equipment for each subsequence at each stage, obtain the vertical coupling relationship between each 3D model and the horizontal coupling relationship between each 3D model, and construct a full-process model of PCB manufacturing;

[0041] Step s3: Obtain the stability coefficients of various indicators of each 3D model in the PCB board manufacturing full process model within several historical monitoring periods, as well as the key indicators corresponding to various types of sudden failures of each 3D model;

[0042] Step s4: constructing a fuzzy fault warning reference database, obtaining the dynamic sign network corresponding to each type of fault in each three-dimensional model and storing it in the fuzzy fault warning reference database;

[0043] Step s5: Perform fuzzy warning on various types of indicators of each three-dimensional model in the current monitoring period, generate fault warning signals for each three-dimensional model, and visualize them through the PCB board manufacturing full process model.

[0044] It should be further explained that, in a specific implementation process, the process of obtaining the manufacturing process information of the PCB board, constructing the manufacturing process sequence of the PCB board, and extracting the values of various types of indicators of each manufacturing process in the manufacturing process sequence includes:

[0045] Acquiring process information of a current PCB manufacturing process, extracting manufacturing unit characteristics based on the manufacturing process information, and dividing the PCB manufacturing process into a plurality of stage subsequences according to the manufacturing unit characteristics;

[0046] Set data monitoring points in each stage subsequence, and use data retrieval to obtain equipment process indicators and product quality indicators at each data monitoring point based on the manufacturing unit characteristics of the corresponding stage subsequence;

[0047] Obtain the planning information of the current PCB board manufacturing process, and obtain the production plan indicators of each data monitoring point based on the planning information of the corresponding stage subsequence;

[0048] The data monitoring points are used to collect the values of various types of indicators in each stage subsequence and mark the monitoring time, and set the monitoring cycle. The various types of indicator values include equipment process indicator values, product quality indicator values, and production plan indicator values.

[0049] It should be further explained that, in the specific implementation process, the manufacturing unit characteristics of each stage subsequence of the PCB board and the equipment process indicators corresponding to the manufacturing unit characteristics include:

[0050] Printing circuit pattern stage: The required circuit pattern is produced by printing a circuit board, using the corresponding photocopy film and photosensitive adhesive, and using exposure and development processes to produce the circuit pattern;

[0051] Exposure time: affects the clarity and quality of the circuit pattern;

[0052] Exposure light intensity: affects the exposure effect and needs to ensure consistency;

[0053] Development time: affects the etching effect of the circuit board, and the appropriate time needs to be controlled;

[0054] Ink viscosity: affects printing quality and needs to be kept stable;

[0055] Chemical etching stage: Place the printed circuit board into the etching solution to dissolve the areas not covered by the copper foil to form the required circuit structure;

[0056] Etching solution temperature: affects etching speed and quality;

[0057] Etching solution concentration: directly affects the etching speed and needs to be controlled within an appropriate range;

[0058] Etchant spraying uniformity: ensure etching uniformity and accuracy;

[0059] Drilling stage: On the etched circuit board, according to the design requirements, CNC drilling machine or laser drilling machine is used to drill holes in the corresponding positions to connect the circuits of different layers;

[0060] Drilling speed: affects drilling quality and accuracy;

[0061] Drill bit wear: affects drilling quality and accuracy;

[0062] Drilling depth: Ensure that the drilling depth meets the design requirements;

[0063] Film stripping stage: Apply film stripping agent to the surface of the circuit board to protect the circuit board from corrosion by exposed copper;

[0064] Film coating uniformity: affects the film effect and protection;

[0065] Temperature of film stripping agent: affects the film stripping effect and needs to be kept stable;

[0066] Soldering stage: by heating, the solder is melted and applied to the circuit board to connect the electronic components and the circuit board;

[0067] Welding temperature and time: affect welding quality and connection firmness;

[0068] Quality inspection stage: The manufactured PCB boards are subjected to multiple inspections such as appearance inspection, steel mesh testing, electrical testing, and functional testing to ensure that the products meet quality requirements.

[0069] It should be further explained that, in the specific implementation process, the product quality indicators of each stage subsequence of the PCB board include:

[0070] Dimensional parameters: PCB thickness: ensure that it meets the design requirements; PCB size: including length, width, etc., ensure that it meets the design specifications;

[0071] Electrical parameters: Resistance value: Ensure that the circuit is powered normally. Continuity test: Test whether the wire is powered normally; Insulation resistance: Test the insulation performance of the circuit board; Welding integrity: Check whether the solder joints are good and there are no short circuits, open circuits, etc.

[0072] Appearance parameters: Surface finish: Check the surface flatness and cleanliness; Hole position accuracy: Check whether the drilling position is accurate; Product appearance defects: such as scratches, bubbles, uneven color, etc.

[0073] Performance parameters: Photoelectric performance: LED brightness, luminous wavelength and other performance indicators; Thermal management performance: whether the heat dissipation effect meets the requirements;

[0074] Environmental parameters: Temperature and humidity adaptability: test the performance of the product in different temperature and humidity environments; Light adaptability: test the performance of the product under different lighting conditions.

[0075] It should be further explained that, in the specific implementation process, the production planning indicators of each stage sub-sequence of the PCB board include product processing time, product processing volume within a unit monitoring cycle, defective rate, etc.

[0076] It should be further explained that, in the specific implementation process, the three-dimensional models of the PCB manufacturing equipment in each stage subsequence are constructed, the vertical coupling relationship between each three-dimensional model and the horizontal coupling relationship between each three-dimensional model are obtained, and the process of constructing the full-process model of PCB manufacturing includes:

[0077] Obtaining the physical entities of PCB manufacturing equipment at each stage subsequence in the current PCB manufacturing process and the assembly relationships between the physical entities, performing 3D modeling processing on the physical entities of the PCB manufacturing equipment at each stage subsequence to generate 3D models, and connecting the 3D models according to the assembly relationships to generate an assembly connection layer;

[0078] The numerical values of various indicators of each data monitoring point are obtained for data format preprocessing to generate twin data. The twin data are matched with the three-dimensional model in the assembly connection layer to obtain the vertical coupling relationship of each three-dimensional model and the horizontal coupling relationship between each three-dimensional model. The vertical coupling relationship of each three-dimensional model and the horizontal coupling relationship between each three-dimensional model are superimposed on the assembly connection layer to generate a full-process model of PCB board manufacturing.

[0079] It should be further explained that, in the specific implementation process, the process of obtaining the vertical coupling relationship of each three-dimensional model includes:

[0080] Obtain the values of various indicators of data monitoring points within several historical monitoring periods and the standard threshold ranges corresponding to the values of various indicators of each type, compare the values of various indicators of each type at various data monitoring points within several historical monitoring periods with the corresponding standard threshold ranges, and mark the indicators whose values are not within the corresponding standard threshold ranges as abnormal;

[0081] Obtain the probability that the product quality indicator and the production plan indicator are in an abnormal state when the equipment process indicator at the current data monitoring point is in an abnormal state within several historical monitoring periods, preset a probability threshold, and if the probability that the product quality indicator is in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the equipment process indicator and the product quality indicator; if the probability that the production plan indicator is in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the equipment process indicator and the production plan indicator;

[0082] Obtain the probability that the equipment process indicators and the production plan indicators are in an abnormal state when the product quality indicators at the data monitoring points are in an abnormal state during several historical monitoring periods, preset a probability threshold, and if the probability that the equipment process indicators are in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the product quality indicators and the equipment process indicators; if the probability that the production plan indicators are in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the product quality indicators and the production plan indicators;

[0083] Obtain the probability that the product quality indicator and the equipment process indicator are in an abnormal state when the production plan indicator of the data monitoring point is in an abnormal state during several historical monitoring periods, preset a probability threshold, and if the probability that the equipment process indicator is in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the production plan indicator and the equipment process indicator; if the probability that the product quality indicator is in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the production plan indicator and the product quality indicator;

[0084] By analogy, the vertical coupling relationship and vertical coupling coefficient between various types of indicators at each data monitoring point are obtained.

[0085] It should be further explained that, in the specific implementation process, the calculation formula for obtaining the vertical coupling coefficient is:

[0086]

[0087]

[0088]

[0089]

[0090]

[0091]

[0092] Among them, p(v i , b j ) represents the equipment process index v at the data monitoring point i and product quality indicators b j The vertical coupling coefficient between i , c k ) represents the equipment process index v at the data monitoring point i and production planning index c k The vertical coupling coefficient between j , v i ) represents the product quality index b at the data monitoring point j and equipment process indicators v i The vertical coupling coefficient between j , c k ) represents the product quality index b at the data monitoring point j and production planning index c k The vertical coupling coefficient between k , b j ) represents the production plan index c of the data monitoring point k and product quality indicators b j The vertical coupling coefficient between k , v i ) represents the production plan index c of the data monitoring point k and equipment process indicators v i The vertical coupling coefficient between failure(v i ) represents the equipment process index v at the data monitoring point i The cumulative number of abnormal states, failure(b j ) represents the product quality index b at the data monitoring point j The cumulative number of abnormal states, failure(c k ) represents the production plan index c of the data monitoring point k The cumulative number of abnormal states, failure(v i , b j ) represents the equipment process index v at the data monitoring point i Under abnormal conditions, product quality index b j The cumulative number of abnormal states, failure(v i , ck ) represents the equipment process index v at the data monitoring point i Under abnormal conditions, the production plan indicator c k The cumulative number of abnormal states, failure(b j , v i ) represents the product quality index b at the data monitoring point j Under abnormal conditions, the equipment process index v i The cumulative number of abnormal states, failure(b j , c k ) represents the product quality index b at the data monitoring point j Under abnormal conditions, the production plan indicator c k The cumulative number of abnormal states, failure(c k , b j ) represents the production plan index c of the data monitoring point k Under abnormal conditions, product quality index b j The cumulative number of abnormal states, failure(c k , v i ) represents the production plan index c of the data monitoring point k Under abnormal conditions, the equipment process index v i The cumulative number of abnormal states, θ represents the coupling conversion coefficient;

[0093] Where v={v1,v2,...,v i ,...,v q}, i = 1, 2, ..., q, v i represents the equipment process indicator of type i, q represents the total number of types of equipment process indicators, such as exposure time, exposure light intensity, development time, ink viscosity in the printing circuit pattern stage, etching solution temperature, etching solution concentration, and etching agent spraying uniformity in the chemical etching stage;

[0094] b={b1,b2,...,b j ,...,b r}, j = 1, 2, ..., r, b j represents the jth type of product quality indicator, and r represents the total number of product quality indicator types. Product quality indicator types include PCB thickness, PCB size, resistance value, soldering integrity, surface finish, hole position accuracy, and product appearance defects such as scratches, bubbles, and color unevenness.

[0095] c={c1,c2,...,c k ,...,cy}, k = 1, 2, ..., y, c k represents the kth type of production plan indicator, and y represents the total number of types of production plan indicators. The types of production plan indicators include product processing time, product processing volume within a unit monitoring cycle, defective rate, etc.

[0096] It should be further explained that, in the specific implementation process, the process of obtaining the horizontal coupling relationship between the three-dimensional models includes:

[0097] Obtain target data monitoring points and other data monitoring points that have an assembly relationship with the target monitoring points, obtain the values of various types of indicators of the target data monitoring points and other data monitoring points in several historical monitoring periods, obtain the probability that the equipment process indicators, product quality indicators and production plan indicators of other data monitoring points are in an abnormal state when the equipment process indicators of the target data monitoring points are in an abnormal state in several historical monitoring periods, and if the probability of the equipment process indicators of other data monitoring points is greater than the probability threshold, then generate the horizontal coupling relationship and horizontal coupling coefficient between the equipment process indicators of the target data monitoring points and the equipment process indicators of other data monitoring points, for example, obtain the equipment process indicators of the target data monitoring points and other data monitoring points.

[0098] ˊ

[0099] The horizontal coupling coefficient p(v i ,(v i )) is calculated as:

[0100]

[0101] ˊ

[0102] Among them, (v i ) represents the equipment process index of type i at other data monitoring points,

[0103] ˊ

[0104] failure(v i ,(v i )) represents the equipment process index v at the target data monitoring point i Under the condition of abnormal state, it indicates the cumulative number of abnormal states of equipment process indicators of type i at other data monitoring points;

[0105] By analogy, the horizontal coupling relationship and horizontal coupling coefficient between the equipment process indicators, product quality indicators and production plan indicators of the target data monitoring point and the equipment process indicators, product quality indicators and production plan indicators of other data monitoring points are obtained.

[0106] It should be further explained that, in the specific implementation process, the process of obtaining the stability coefficients of various types of indicators of each three-dimensional model in the PCB board manufacturing full process model within several historical monitoring periods and the key type indicators corresponding to several types of sudden failures of each three-dimensional model includes:

[0107] Obtain historical occurrence data of several types of sudden faults in each three-dimensional model in the PCB board manufacturing full-process model within several historical monitoring periods, the several types of faults including: circuit board printer failure, chemical etching equipment failure, drilling equipment failure, welding equipment failure, etc. The historical occurrence data includes a time series sequence of the numerical values of each type of indicator before the sudden occurrence of several types of faults in the three-dimensional model within the historical monitoring period, compare the time series sequence of the numerical values of each type of indicator within the monitoring period with the corresponding standard threshold range, obtain the frequency of each type of indicator value not being within the corresponding standard threshold range and the fluctuation amplitude of each type of indicator value, obtain the stability coefficient of each type of indicator based on the frequency and fluctuation amplitude, and simultaneously obtain the stability coefficient of each type of indicator when each three-dimensional model does not have various types of faults, and mark the stability coefficient as the standard stability coefficient;

[0108] It should be further explained that, in the specific implementation process, the calculation formula for obtaining the stability coefficient of each type of indicator based on the frequency and fluctuation amplitude is:

[0109]

[0110] Where JC represents the stability coefficient, θ represents the conversion coefficient, sc represents the frequency, and fb represents the fluctuation amplitude.

[0111] Obtain the stability coefficient difference between the stability coefficient of each three-dimensional model when several types of sudden faults occur and the corresponding standard stability coefficient, and screen out the type indicator corresponding to the maximum stability coefficient difference when each three-dimensional model suddenly encounters a certain type of sudden fault as the key type indicator. Similarly, obtain the key type indicator corresponding to each three-dimensional model when several types of sudden faults occur.

[0112] It should be further explained that, in the specific implementation process, the process of constructing a fuzzy fault warning reference database, obtaining a dynamic sign network corresponding to each type of fault in each three-dimensional model, and storing it in the fuzzy fault warning reference database includes:

[0113] Extract the stability coefficient differences of several types of indicators that have a vertical coupling relationship with the key type indicators corresponding to the type fault of the three-dimensional model, and the stability coefficient differences of several types of indicators of other three-dimensional models that have a horizontal coupling relationship with the key type indicators. According to the stability coefficient differences of the key type indicators corresponding to the type fault, the stability coefficient differences of several types of indicators that have a vertical coupling relationship with the key type indicators, and the stability coefficient differences of several types of indicators of other three-dimensional models that have a horizontal coupling relationship with the key type indicators, construct a dynamic marker network of the type fault. Similarly, obtain the dynamic marker network corresponding to each type of fault of each three-dimensional model and store it in the fuzzy fault warning comparison database.

[0114] It should be further explained that, in the specific implementation process, fuzzy early warning is performed on various types of indicators of each 3D model in the current monitoring period, and fault warning signals of each 3D model are generated. The process of visual display through the PCB board manufacturing full process model includes:

[0115] Obtain the stability coefficient difference of each type of indicator of each three-dimensional model in the current monitoring period, screen out the type indicator corresponding to the maximum stability coefficient difference of each three-dimensional model and mark it as the type indicator to be analyzed, obtain several type indicators that have a vertical coupling relationship with the type indicator to be analyzed and several type indicators of other three-dimensional models that have a horizontal coupling relationship, and construct a marker network to be analyzed based on the stability coefficient difference of the type indicator to be analyzed, the stability coefficient difference of several type indicators that have a vertical coupling relationship with the type indicator to be analyzed, and the stability coefficient difference of several type indicators of other three-dimensional models that have a horizontal coupling relationship with the type indicator to be analyzed;

[0116] The marker network to be analyzed is matched with a dynamic marker network in the fuzzy fault warning comparison database to obtain the similarity between the type index to be analyzed of the marker network and each type index and the key type index of the dynamic marker network and the stability coefficient difference of each type index. The similarity between the type index to be analyzed of the marker network and each type index and the key type index of the dynamic marker network and the stability coefficient difference of each type index is specifically as follows: the similarity between the stability coefficient difference between the type index to be analyzed of the marker network and the key type index of the dynamic marker network, the similarity between the stability coefficient difference between the marker network to be analyzed and each corresponding same type index in the dynamic marker network. The similarity of the difference in stability coefficients of the type of indicators to be analyzed of the sign network and the similarity of the difference in stability coefficients of the indicators of each type are used as evaluation indicators. At the same time, the vertical coupling coefficient or horizontal coupling coefficient between the type of indicators to be analyzed and the indicators of each type is obtained. The weight index matrix of the evaluation index is set according to the vertical coupling coefficient or horizontal coupling coefficient between the type of indicators to be analyzed and the indicators of each type. The warning level is preset. The membership matrix of the evaluation index of each three-dimensional model for the warning level of a certain dynamic sign network is obtained through fuzzy comprehensive evaluation. The warning level of each three-dimensional model for a certain dynamic sign network is obtained according to the membership matrix and the weight index matrix. By analogy, the warning level of each three-dimensional model for each dynamic sign network is obtained.

[0117] A warning level threshold is preset, and the warning level of each three-dimensional model for each dynamic marker network is compared with the warning level threshold. If there is a three-dimensional model whose warning level for a dynamic network marker is greater than the warning level threshold, the type of fault corresponding to the dynamic network marker is obtained, and a fault warning signal for the three-dimensional model is generated according to the type of fault.

[0118] It should be further explained that, in the specific implementation process, the calculation formula for obtaining the similarity of the stability coefficient difference between each type indicator of the marker network to be analyzed and the corresponding same type indicator in the dynamic marker network is:

[0119] S (i，j) =JC i -JC j ;

[0120] Among them, S (i，j) Indicates the similarity of the stability coefficient difference between the analyzed sign network and the corresponding indicators of the same type in the dynamic sign network, JC i Indicates the stability coefficient difference of the type index of the marker network to be analyzed, JC j Indicates the stability coefficient difference of the type index of the dynamic sign network.

[0121] It should be further explained that, in the specific implementation process, the process of obtaining the warning level of each three-dimensional model for a certain dynamic sign network according to the membership matrix and the weight index matrix includes:

[0122] The indicator weight matrix and the membership matrix of the evaluation indicators are integrated through a formula to obtain a fuzzy comprehensive evaluation matrix of the evaluation indicators. The membership of each three-dimensional model to different warning levels is obtained according to the fuzzy comprehensive evaluation matrix. The warning level with the highest membership corresponding to each three-dimensional model is screened out, and the warning level with the highest membership corresponding to each three-dimensional model is used as the warning level of each three-dimensional model.

[0123] Wherein, the formula is:

[0124] M = αM1 × βM2;

[0125] Among them, M is the fuzzy comprehensive evaluation matrix of the evaluation index, M1 is the indicator weight matrix of the evaluation index, M2 is the membership matrix, "×" represents the multiplication of the elements at corresponding positions of the weight matrix of the evaluation index and the membership matrix, and α and β are weighting parameters used to control the balance between the weight matrix and the membership matrix in the fuzzy comprehensive evaluation matrix of the evaluation index.

[0126] It should be further explained that, in the specific implementation process, if a certain type of equipment failure occurs in the subsequence of the current monitoring cycle stage, but the PCB board manufacturing full process model does not perform a visual display of the fault warning signal and the dynamic network mark of the equipment failure of the type in question exists in the fuzzy fault warning comparison database, then the numerical time series of each type of indicator in the subsequence of the current monitoring cycle stage is obtained to construct the dynamic network mark of the equipment failure of the type in question, and the original dynamic network mark of the equipment failure of the type in question is replaced, thereby realizing real-time updating of the fuzzy fault warning comparison database;

[0127] If a certain type of equipment failure occurs in the current monitoring cycle stage subsequence, but the PCB board manufacturing full process model does not perform a visual display of the fault warning signal and the dynamic network mark of the said type of equipment failure does not exist in the fuzzy fault warning comparison database, then the numerical time series sequence of each type of indicator in the current monitoring cycle stage subsequence is obtained to construct the dynamic network mark of the said type of equipment failure, and the dynamic network mark of the said type of equipment failure is added to the fuzzy fault warning comparison database to realize real-time update of the fuzzy fault warning comparison database, thereby ensuring the real-time and accuracy of the fault warning signal of the three-dimensional model.

[0128] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A full-process manufacturing optimization method for PCB boards based on the Internet of Things, characterized in that: The following steps are involved: Step s1: Obtain the manufacturing process information of the PCB board to construct the manufacturing process sequence of the PCB board, and extract the values of various types of indicators of each manufacturing process in the manufacturing process sequence; Step s2: Construct a 3D model of the PCB manufacturing equipment for each subsequence at each stage, obtain the vertical coupling relationship between each 3D model and the horizontal coupling relationship between each 3D model, and construct a full-process model of PCB manufacturing; Step s3: Obtain the stability coefficients of various indicators of each 3D model in the PCB board manufacturing full process model within several historical monitoring periods, as well as the key indicators corresponding to various types of sudden failures of each 3D model; Step s4: constructing a fuzzy fault warning reference database, obtaining the dynamic sign network corresponding to each type of fault in each three-dimensional model and storing it in the fuzzy fault warning reference database; Step s5: Perform fuzzy warning on various indicators of each 3D model in the current monitoring period, generate fault warning signals for each 3D model, and visualize them through the PCB board manufacturing process model; The process of obtaining the vertical coupling relationship of each 3D model includes: Obtain the values of various indicators of data monitoring points within several historical monitoring periods and the standard threshold ranges corresponding to the values of various indicators of each type, compare the values of various indicators of each type at various data monitoring points within several historical monitoring periods with the corresponding standard threshold ranges, and mark the indicators whose values are not within the corresponding standard threshold ranges as abnormal; Obtain the probability that the product quality indicator and the production plan indicator are in an abnormal state when the equipment process indicator at the current data monitoring point is in an abnormal state within several historical monitoring periods, preset a probability threshold, and if the probability that the product quality indicator is in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the equipment process indicator and the product quality indicator; if the probability that the production plan indicator is in an abnormal state is greater than the probability threshold, generate a vertical coupling relationship and a vertical coupling coefficient between the equipment process indicator and the production plan indicator; Similarly, the vertical coupling relationship and vertical coupling coefficient between various types of indicators at each data monitoring point are obtained; The process of obtaining the horizontal coupling relationship between the three-dimensional models includes: Obtain a target data monitoring point and other data monitoring points that have an assembly relationship with the target monitoring point, obtain various types of indicator values of the target data monitoring point and other data monitoring points in several historical monitoring periods, obtain the probability that the equipment process indicators, product quality indicators, and production plan indicators of other data monitoring points are in an abnormal state when the equipment process indicators of the target data monitoring point are in an abnormal state in several historical monitoring periods, and if the probability of the equipment process indicators of the other data monitoring points is greater than a probability threshold, generate a horizontal coupling relationship and a horizontal coupling coefficient between the equipment process indicators of the target data monitoring point and the equipment process indicators of the other data monitoring points; By analogy, the horizontal coupling relationship and horizontal coupling coefficient between various types of indicators between various data monitoring points are obtained.

2. The method for optimizing the whole process of PCB board manufacturing based on the Internet of Things according to claim 1, wherein: The process of obtaining the manufacturing process information of the PCB board, constructing the manufacturing process sequence of the PCB board, and extracting the values of various types of indicators of each manufacturing process in the manufacturing process sequence includes: Acquiring process information of a current PCB manufacturing process, extracting manufacturing unit characteristics based on the manufacturing process information, and dividing the PCB manufacturing process into a plurality of stage subsequences according to the manufacturing unit characteristics; Set data monitoring points in each stage subsequence, and use data retrieval to obtain equipment process indicators and product quality indicators at each data monitoring point based on the manufacturing unit characteristics of the corresponding stage subsequence; Obtain the planning information of the current PCB board manufacturing process, and obtain the production plan indicators of each data monitoring point based on the planning information of the corresponding stage subsequence; The data monitoring points are used to collect various types of indicator values of each stage subsequence and mark the monitoring time, and set the monitoring cycle. The various types of indicator values include equipment process indicator values, product quality indicator values, and production plan indicator values.

3. The method for optimizing the whole process of PCB board manufacturing based on the Internet of Things according to claim 2, wherein: Constructing a 3D model of the PCB manufacturing equipment for each subsequence of each stage, obtaining the vertical coupling relationship between each 3D model and the horizontal coupling relationship between each 3D model, and constructing a full-process model of PCB manufacturing includes the following steps: Obtaining the physical entities of PCB manufacturing equipment at each stage subsequence in the current PCB manufacturing process and the assembly relationships between the physical entities, performing 3D modeling processing on the physical entities of the PCB manufacturing equipment at each stage subsequence to generate 3D models, and connecting the 3D models according to the assembly relationships to generate an assembly connection layer; The numerical values of various indicators of each data monitoring point are obtained for data format preprocessing to generate twin data. The twin data are matched with the three-dimensional model in the assembly connection layer to obtain the vertical coupling relationship of each three-dimensional model and the horizontal coupling relationship between each three-dimensional model. The vertical coupling relationship of each three-dimensional model and the horizontal coupling relationship between each three-dimensional model are superimposed on the assembly connection layer to generate a full-process model of PCB board manufacturing.

4. The method for optimizing the whole process of PCB board manufacturing based on the Internet of Things according to claim 3, wherein: The process of obtaining the stability coefficients of various types of indicators of each three-dimensional model in the PCB board manufacturing full process model within several historical monitoring periods and the key type indicators corresponding to several types of sudden failures of each three-dimensional model includes: Obtain historical occurrence data of several types of sudden faults of each three-dimensional model in the PCB board manufacturing full-process model within several historical monitoring periods, the historical occurrence data including time series sequences of numerical values of each type of indicator before the sudden occurrence of several types of faults of the three-dimensional model within the historical monitoring period, comparing the time series sequences of numerical values of each type of indicator within the monitoring period with the corresponding standard threshold range, obtaining the frequency of each type of indicator value not being within the corresponding standard threshold range and the fluctuation amplitude of each type of indicator value, obtaining the stability coefficient of each type of indicator based on the frequency and the fluctuation amplitude, and simultaneously obtaining the stability coefficient of each type of indicator when no various types of faults occur in each three-dimensional model, and marking the stability coefficient as the standard stability coefficient; Obtain the stability coefficient difference between the stability coefficient of each three-dimensional model when several types of sudden faults occur and the corresponding standard stability coefficient, and screen out the type indicator corresponding to the maximum stability coefficient difference when each three-dimensional model suddenly encounters a certain type of sudden fault as the key type indicator. Similarly, obtain the key type indicator corresponding to each three-dimensional model when several types of sudden faults occur.

5. The method for optimizing the whole process of PCB board manufacturing based on the Internet of Things according to claim 4, wherein: The process of constructing a fuzzy fault warning reference database, obtaining a dynamic sign network corresponding to each type of fault in each three-dimensional model, and storing the dynamic sign network in the fuzzy fault warning reference database includes: Extract the stability coefficient differences of several types of indicators that have a vertical coupling relationship with the key type indicators corresponding to the type fault of the three-dimensional model, and the stability coefficient differences of several types of indicators of other three-dimensional models that have a horizontal coupling relationship with the key type indicators. According to the stability coefficient differences of the key type indicators corresponding to the type fault, the stability coefficient differences of several types of indicators that have a vertical coupling relationship with the key type indicators, and the stability coefficient differences of several types of indicators of other three-dimensional models that have a horizontal coupling relationship with the key type indicators, construct a dynamic marker network of the type fault. Similarly, obtain the dynamic marker network corresponding to each type of fault of each three-dimensional model and store it in the fuzzy fault warning comparison database.

6. The method for optimizing the whole process of PCB board manufacturing based on the Internet of Things according to claim 5, characterized in that: The process of providing fuzzy early warning for various indicators of each 3D model within the current monitoring period, generating fault warning signals for each 3D model, and visually displaying them through the PCB board manufacturing full-process model includes: Obtain the stability coefficient difference of each type of indicator of each three-dimensional model in the current monitoring period, screen out the type indicator corresponding to the maximum stability coefficient difference of each three-dimensional model and mark it as the type indicator to be analyzed, obtain several type indicators that have a vertical coupling relationship with the type indicator to be analyzed and several type indicators of other three-dimensional models that have a horizontal coupling relationship, and construct a marker network to be analyzed based on the stability coefficient difference of the type indicator to be analyzed, the stability coefficient difference of several type indicators that have a vertical coupling relationship with the type indicator to be analyzed, and the stability coefficient difference of several type indicators of other three-dimensional models that have a horizontal coupling relationship with the type indicator to be analyzed; Match the sign network to be analyzed with a dynamic sign network in the fuzzy fault warning comparison database, obtain the type index to be analyzed of the sign network to be analyzed and the stability coefficient difference similarity between each type index and the key type index of the dynamic sign network and each type index, use the stability coefficient difference similarity of the type index to be analyzed of the sign network to be analyzed and the stability coefficient difference similarity of each type index as evaluation indexes, and simultaneously obtain the vertical coupling coefficient or horizontal coupling coefficient between the type index to be analyzed and each type index, set the weight index matrix of the evaluation index according to the vertical coupling coefficient or horizontal coupling coefficient between the type index to be analyzed and each type index, preset the warning level, obtain the membership matrix of the evaluation index of each three-dimensional model for the warning level of a dynamic sign network through fuzzy comprehensive evaluation, obtain the warning level of each three-dimensional model for a dynamic sign network according to the membership matrix and the weight index matrix, and so on, obtain the warning level of each three-dimensional model for each dynamic sign network; A warning level threshold is preset, and the warning level of each three-dimensional model for each dynamic marker network is compared with the warning level threshold. If there is a three-dimensional model whose warning level for a dynamic network marker is greater than the warning level threshold, the type of fault corresponding to the dynamic network marker is obtained, and a fault warning signal for the three-dimensional model is generated according to the type of fault.

Images