CN113655768B - Assembly yield control method, equipment and computer readable storage medium - Google Patents
Assembly yield control method, equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN113655768B CN113655768B CN202111224221.3A CN202111224221A CN113655768B CN 113655768 B CN113655768 B CN 113655768B CN 202111224221 A CN202111224221 A CN 202111224221A CN 113655768 B CN113655768 B CN 113655768B
- Authority
- CN
- China
- Prior art keywords
- assembly
- training
- yield
- assembly yield
- prediction model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS], computer integrated manufacturing [CIM]
- G05B19/41875—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS], computer integrated manufacturing [CIM] characterised by quality surveillance of production
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/32—Operator till task planning
- G05B2219/32204—Performance assurance; assure certain level of non-defective products
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The application discloses an assembly yield control method, equipment and a computer readable storage medium, wherein the method comprises the following steps: obtaining an assembly yield prediction model; inputting target material information of a target material into the assembly yield prediction model to obtain a target arrangement of the target material and a target assembly yield corresponding to the target arrangement; the assembly yield prediction model is obtained according to training material information, a training scheduling combination and a training assembly yield training initial prediction model; the target material information of the target material to be assembled is input into the assembly yield prediction model by acquiring the assembly yield prediction model, the assembly yield prediction model can automatically output the arrangement mode, namely the target arrangement, and meanwhile, the output assembly yield corresponding to the target arrangement is also output, when the assembly yield is maximum, the corresponding target arrangement is the optimal arrangement mode, so that the automatic arrangement is realized on the basis of controlling the assembly yield, and the efficiency of material arrangement is improved.
Description
Technical Field
The present disclosure relates to the field of manufacturing and production technologies, and in particular, to an assembly yield control method, an assembly yield control apparatus, and a computer-readable storage medium.
Background
The yield problem of assembly lines is always widely existed in factories, and an intelligent optimal matching algorithm based on the assembly of two or more materials is not applied. The prior art still stays in the traditional passive problem solving method, an experienced assembly master determines batch-based production scheduling of two sizes of materials based on the sizes of historical measurement materials through a trial and error rule, and the method has the characteristics of large time lag, poor yield and incapability of forming knowledge rules through experience, so that the problems of low scheduling efficiency and unstable scheduling yield of the conventional material assembly are caused.
Disclosure of Invention
The present application mainly aims to provide an assembly yield control method, an assembly yield control device, and a computer-readable storage medium, and aims to solve the problems of low assembly efficiency and unstable assembly yield in material assembly.
To achieve the above object, the present application provides an assembly yield control method, including:
obtaining an assembly yield prediction model;
inputting target material information of a target material into the assembly yield prediction model to obtain a target arrangement of the target material and a target assembly yield corresponding to the target arrangement;
the assembly yield prediction model is obtained according to training material information, a training scheduling combination and a training assembly yield training initial prediction model.
Optionally, before the step of obtaining the assembly yield prediction model, the method further includes:
establishing an initial prediction model;
acquiring a training data set, wherein the training data set comprises training material information, a training scheduling combination and a training assembly yield;
and training the initial prediction model according to the training data set to obtain an assembly yield prediction model.
Optionally, the step of establishing an initial prediction model includes:
acquiring historical related information of the assembly materials;
determining key factors influencing assembly precision in the assembly materials according to the historical relevant information;
and establishing a relation model between the key factors and the predicted yield, and determining an initial prediction model.
Optionally, the history related information includes information of a drawing position size, a measurement size, and a production model number of the assembly materials of different batches, and the step of determining key factors affecting assembly accuracy in the assembly materials according to the history related information includes:
obtaining key point positions influencing assembly precision in the assembly materials;
calculating size data according to the figure position size and the measurement size of the key point location to obtain a production process level numerical value corresponding to the key point location;
calculating to obtain the overall error levels of different key point positions of the assembly material according to the production process level numerical value and a preset calculation model;
comparing the overall error level with a preset error level;
and if the overall error level is greater than the preset error level, determining the key point position corresponding to the overall error level as a key factor.
Optionally, the step of acquiring a training data set includes:
establishing an assembly failure mode library according to the historical relevant information and the historical assembly yield;
performing feature screening on the historical relevant information according to the key factors to obtain training feature data;
acquiring a production process level numerical value and a historical assembly yield corresponding to the training characteristic data, and respectively setting the production process level numerical value and the historical assembly yield as a training level numerical value and a training yield;
establishing the training data set based on the training feature data, the training level numerical value, and the training yield.
Optionally, after the step of establishing an assembly failure pattern library according to the historical related information and the historical assembly yield, the method includes:
comparing the historical assembly yield with a yield reaching standard value;
if the historical assembly yield is smaller than the yield reaching standard value, analyzing historical relevant information corresponding to the historical assembly yield to obtain an influence factor influencing the assembly yield;
and adjusting the size of the map bit in the historical related information according to the influence factor, and updating the training data set.
Optionally, the step of training the initial prediction model according to the training data set to obtain an assembly yield prediction model includes:
inputting the training data set into the initial prediction model, and performing prediction processing on the training data set based on the initial prediction model to obtain a corresponding training yield;
comparing the training yield with a preset yield to obtain a comparison result;
and adjusting parameters of the initial prediction model based on the comparison result, determining whether a preset training completion condition is met, and if not, returning to the step of inputting the training data set into the initial prediction model until the preset training completion condition is met to obtain an assembly yield prediction model.
Optionally, after the step of inputting the target material information of the target material into the assembly yield prediction model to obtain the target arrangement of the target material and the target assembly yield corresponding to the target arrangement, the method includes:
comparing the target assembly yield with the actual assembly yield to obtain a yield difference;
if the yield difference value is larger than the yield difference threshold value, outputting adjustment information;
and adjusting and optimizing the assembly yield prediction model according to the adjustment information.
In order to achieve the above object, the present application also provides an assembly yield control apparatus, including: the assembly yield control method comprises a memory, a processor and an assembly yield control program stored on the memory and capable of running on the processor, wherein the assembly yield control program realizes the steps of the assembly yield control method when being executed by the processor.
In addition, to achieve the above object, the present application also provides a computer readable storage medium, having a mounting yield control program stored thereon, which when executed by a processor implements the steps of the mounting yield control method as described above.
In the application, an assembly yield prediction model is obtained; inputting target material information of a target material into the assembly yield prediction model to obtain a target arrangement of the target material and a target assembly yield corresponding to the target arrangement; the assembly yield prediction model is obtained according to training material information, a training scheduling combination and a training assembly yield training initial prediction model; through the method, the assembly yield prediction model is obtained, the target material information of the target material to be assembled is input into the assembly yield prediction model, the assembly yield prediction model can automatically output the arrangement mode, namely the target arrangement, and meanwhile, the output assembly yield corresponding to the target arrangement is also output.
Drawings
FIG. 1 is a schematic diagram of a hardware operating environment according to an embodiment of the present disclosure;
FIG. 2 is a flowchart illustrating a first embodiment of an assembly yield control method according to the present application;
FIG. 3 is a block diagram illustrating a functional diagram of an assembly yield control apparatus according to a preferred embodiment of the present invention.
The implementation, functional features and advantages of the objectives of the present application will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
As shown in fig. 1, fig. 1 is a schematic device structure diagram of a hardware operating environment according to an embodiment of the present application.
It should be noted that the assembly yield control device in the embodiment of the present application may be a smart phone, a personal computer, a server, and the like, and is not limited herein.
As shown in fig. 1, the assembly yield control apparatus may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, a communication bus 1002. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (e.g., a magnetic disk memory). The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the device configuration shown in fig. 1 does not constitute a limitation of the assembly yield control device and may include more or fewer components than shown, or some components in combination, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a kind of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and an assembly yield control program. The operating system is a program for managing and controlling hardware and software resources of the equipment, and supports the operation of an assembly yield control program and other software or programs. In the device shown in fig. 1, the user interface 1003 is mainly used for data communication with a client; the network interface 1004 is mainly used for establishing communication connection with a server; the processor 1001 may be configured to call the assembly yield control program stored in the memory 1005, and perform the following operations:
step S10, obtaining an assembly yield prediction model;
step S20, inputting target material information of a target material into the assembly yield prediction model to obtain a target arrangement of the target material and a target assembly yield corresponding to the target arrangement;
the assembly yield prediction model is obtained according to training material information, a training scheduling combination and a training assembly yield training initial prediction model.
Therefore, the target material information of the target material to be assembled is input into the assembly yield prediction model by obtaining the assembly yield prediction model, the assembly yield prediction model can automatically output the arrangement mode, namely the target arrangement, and meanwhile, the output assembly yield corresponding to the target arrangement is also output, when the assembly yield is the maximum, the corresponding target arrangement is the optimal arrangement mode, so that the automatic arrangement is realized on the basis of controlling the assembly yield, and the efficiency of material arrangement is improved.
Referring to fig. 2, a first embodiment based on material arrangement specifically includes the following steps:
step S10, obtaining an assembly yield prediction model;
in this embodiment, it should be noted that the assembly yield control method can be applied to an assembly yield control device belonging to an assembly yield control system belonging to an assembly yield control apparatus.
In this embodiment, the specific application scenarios may be: fixed collocation assembles between the material more than two kinds of the upper reaches production line on certain assembly line, and finished product size level and assembly board yield are directly relevant, and at present, the assembly process lacks supervision and effective utilization to material production size information, can't accomplish the optimal arrangement of material to lead to the unstable problem of assembly yield.
In this embodiment, the assembly yield prediction model is obtained by training the initial prediction model according to the training material information, the training arrangement combination and the training assembly yield, so that the assembly yield of the material to be assembled and the arrangement mode corresponding to the assembly yield can be predicted by inputting the material information of the material to be assembled into the assembly yield prediction model, and the arrangement mode of the material to be assembled can be rapidly obtained on the basis of controlling the assembly yield. It should be noted that the materials to be assembled include at least two materials, and the materials are in a fixed assembly relationship, for convenience of understanding, detailed analysis is performed below by taking two fixed assemblies as an example, specific embodiments of two or more materials are basically the same, and details are not described here. To the assembly of two kinds of materials, two kinds of materials can be housing material and terminal material, and the terminal indicates that the housing that the copper conductor that comes the condensation electric wire used is the socket, and terminal and housing are supporting use, consequently, when the assembly, need consider the cooperation relation between housing and the terminal material and arrange the assembly to control assembly yield.
Further, the step of obtaining the assembly yield prediction model is preceded by the steps of:
step a, establishing an initial prediction model;
in the embodiment, the initial prediction model is a Neural Network, the Neural Network is an information processing mathematical model similar to a human brain Neural cell synaptic structure, the method is developed based on a DNN (Deep Neural Network), the advantages of the Neural Network are fully utilized, the built Network comprises a plurality of hidden layers, each hidden layer is distinguished through different Neural networks, information data of assembling materials are calculated through the Network, mathematical calculation and transformation are carried out on the data, and the assembling yield is predicted.
Further, the step of establishing an initial prediction model includes:
step a1, acquiring historical relevant information of the assembly materials;
in this embodiment, the history related information refers to related information of a material which has actually been processed, and the history related information includes information of map location sizes, measurement sizes, finished product model numbers and the like of the assembled materials of different batches, where the map location sizes are the assembled map sizes of the material and belong to standard sizes which need to be realized by processing; the measurement size refers to the size after production and processing, and may be different from the size of a figure due to the machining precision; the finished product model number refers to a mould for processing materials, different moulds are correspondingly arranged in the material processing through two modes of forming or stamping, after the finished product model number is obtained, when the assembly yield is reduced due to the occurrence of material processing errors, the finished product model number of the material can be traced back to the processed mould, and then the mould is repaired, and the assembly yield is improved through improving the material processing precision.
Different map sizes may exist in the same assembling material (such as housin or a terminal) in different batches due to the influence of machining precision and the like, so that multiple groups of map sizes exist in historical related information, and key factors influencing the assembling precision can be mined according to statistical analysis of different map sizes.
Step a2, determining key factors influencing assembly precision in the assembly materials according to the historical relevant information;
in this embodiment, after obtaining respective history related information of the houseing material and the terminal material, the history related information includes respective measurement sizes of different batches of the two materials, and the key factors affecting the assembly accuracy are obtained through analysis of data of the measurement sizes of multiple batches. The method for obtaining the key factor can be divided into two modes of manual acquisition and calculation acquisition.
Wherein, the manual acquisition means that after the measurement sizes of the housing and the terminal are obtained, according to the geometric matching relationship between the housing and the terminal, quality testing personnel on an assembly line determine key factors according to inherent quality testing standards.
When the key factor is obtained through a calculation and acquisition mode, the specific steps are as follows:
a21, acquiring key point positions influencing assembly precision in the assembly materials;
in this embodiment, the key point location is a certain location point of a material, and the location point is a point generally matched between two materials, that is, the key point location, for example, when a terminal is matched with a housing, a hole allowing the terminal to pass through is set in the housing, at this time, a hole diameter location point in a hole in the housing is the key point location of the housing, and when the terminal needs to be assembled with the housing in an inserting manner, the location point of the terminal is the key point location of the terminal. Therefore, a plurality of key point positions influencing the assembly precision are preliminarily determined manually, the accuracy of the key point positions is unstable based on manual determination, and therefore, accurate key factors influencing the assembly precision need to be further determined through size data calculation.
A22, calculating size data according to the drawing position size and the measurement size of the key point location to obtain a production process level numerical value corresponding to the key point location;
in this embodiment, when the measurement size of the key point of the material changes, the material assembly accuracy may be affected, and since the measurement size of the same key point of the same material may change during different batches of processes due to differences in processing machines, forming molds, processing molds and the like of the material during the processing of the material, the processing accuracy of the key point, that is, the CPK value (composite Process Capability index, production Process level), may need to be calculated according to data of the drawing size of different batches of the key point of the material, and the processing accuracy of the material may be divided into different levels according to the CPK value.
Specifically, in the process of calculating the CPK value, a CPK value of a key point is obtained by automatically calculating according to a graph and bit size (graph and bit size is also a standard size set by assembly) of a certain key point in a housing material, a size tolerance, and a plurality of measurement sizes (measurement sizes are also actual sizes after production) obtained by different batches of production in combination with a CPK calculation formula or a table containing key point data. Therefore, a plurality of key points of the housing material and the CPK values corresponding to the key points can be obtained, and the key points of the terminal material and the CPK value calculation method corresponding to the key points are basically the same as those of the specific implementation mode of the housing material, and are not repeated here.
Step a23, calculating to obtain the overall error levels of different key point positions of the assembly material according to the production process level numerical value and a preset calculation model;
in this embodiment, if preliminarily select, select 4 key points in the houseing material that affect the assembly accuracy, which are X respectivelyH/S1，XH/S2，XH/S3，XH/S4, then based on the error transfer theory, the preset calculation model of the overall error level of the houseing material is as follows:
wherein, Δ yH/SAs the housing overall error level, yH/SCPK value, Δ x, for houseingiIs the absolute error of the measured dimension.
And (4) calculating to obtain the respective integral error level of the four key point positions through the formula (1).
Step a24, comparing the integral error level with a preset error level;
in this embodiment, the preset error level refers to a preset allowable error level, and the key point location within the preset error level belongs to a normal processing state, that is, the influence of the key point location on the assembly precision between two materials is general and may not be considered temporarily;
step a25, if the overall error level is greater than the preset error level, determining a key point position corresponding to the overall error level as a key factor;
in this embodiment, if the calculation result of the overall error level of one of the key points is greater than the preset error level, it indicates that the key point has low processing precision and is prone to deviation in the processing process, and the change in the measurement size caused by the deviation has a large influence on the material assembly, so that the key point is set as the key factor, and the number of the key factors of one material is at least greater than one. The implementation of the determination process of the key factors of the terminal is the same as that of houseing, and is not described herein again.
Step a3, establishing a relation model between the key factors and the predicted yield, and determining an initial prediction model;
in this embodiment, since the assembly yield is related to the dimension level of the assembly material and the assembly equipment, the dimension level of the assembly material is related to the CPK value of the material processing, and the CPK value of the key factor and the change of the measurement size have an influence on the assembly accuracy, that is, the assembly yield is influenced. Meanwhile, the equipment precision of the assembly equipment is high, when the parameters of the assembly equipment are analyzed, a sensor for obtaining the parameters is difficult to install, and the parameters of the assembly equipment are difficult to obtain, so that when an initial prediction model is established, modeling is performed according to the assembly relation between housing and terminal materials, the assembly yield is improved through complementation of tolerance margins of the two materials, and the assembly relation between the two materials is related to the measurement size of key factors of the materials. Therefore, a model is established according to the relation between the key factors influencing the assembly yield and the predicted yield, the corresponding assembly yield can be obtained through prediction based on the measured size of the key factors, and the relation model between the key factors and the predicted yield is set as an initial prediction model.
The above equation (2) is an initial prediction model, wherein y is the predicted yield, and f (X)H/S，Xpin) The method is characterized in that a mapping relation exists between key factors of housing materials and terminal materials and the prediction yield, u is a model coefficient, when the model coefficient is obtained through training, a finished initial prediction model can be determined, and a model relation between the key factors and the prediction yield is established, so that the method has prediction capability.
B, acquiring a training data set, wherein the training data set comprises training material information, a training scheduling combination and a training assembly yield;
in this embodiment, the training data set refers to a data set for training the initial prediction model, the training data set includes at least one of training material information, training scheduling combination, and training assembly yield, and when the initial prediction model is trained, corresponding data is required to be trained, tested, and the like.
Further, the step of obtaining a training data set comprises:
b1, establishing a poor assembly mode library according to the historical relevant information and the historical assembly yield;
in this embodiment, the historical related information at least includes information of the map-bit size, the measurement size, and the production model number of the assembly materials of different batches, and the historical assembly yield refers to the assembly yield obtained after the production of the arrangement manner between two materials of different batches, so the information in the assembly failure mode library includes the arrangement manner between the housing and the terminal of different batches, the measurement size of the housing and the terminal of different batches, and the assembly yield corresponding to the arrangement manner, and these three data constitute a set of data samples, and the assembly failure mode library is constructed by combining a plurality of sets of samples.
B2, performing feature screening on the historical relevant information according to the key factors to obtain training feature data;
in this embodiment, the measurement sizes of different batches of housing and terminals have measurement size information corresponding to a plurality of key points, and these key points are not all capable of affecting the assembly yield, so in order to improve the training efficiency and the data accuracy, data elimination that does not affect the assembly yield prediction in the history related information is required. Through obtaining the key factor of housing and terminal, screen out the data relevant with the key factor through the characteristic engineering, obtain training characteristic data.
It should be noted that, in addition to screening out the training characteristic data according to the key factors of the material, analysis can be performed according to the data in the assembly failure mode library, specifically, the assembly failure mode library includes the information of the historical assembly yield, the information of the historical assembly yield is analyzed, the assembly failure distribution is established, and the change condition of the historical assembly yield is recorded. Firstly, comparing the historical assembly yield with a yield standard value; the yield reaching standard value is preset by personnel and is used as a reference, and when the historical assembly yield is greater than the yield reaching standard value, the historical assembly yield is within the control range and belongs to the required assembly yield; if the historical assembly yield is smaller than the yield reaching standard value, analyzing historical relevant information corresponding to the historical assembly yield to obtain an influence factor influencing the assembly yield; the method can be understood that historical relevant information corresponding to the lowered historical assembly yield is screened out, a group of data samples corresponding to the lowered historical assembly yield is obtained, dimension measurement data of key points of materials in the data samples are analyzed, fluctuation of measurement size change of the plurality of key points of the housing and the terminal materials is counted, the key points with high fluctuation serve as influence factors influencing the assembly yield, meanwhile, data corresponding to the key points are output, reference is provided for equipment die repairing of materials processed by an assembly line upstream material manufacturer, and the dimension level of the materials is improved; further, the size of the map in the historical relevant information is adjusted according to the influence factors, and the training data set is updated, wherein the specific adjustment mode can be that the measurement data in the two materials are weighted according to the influence factors to obtain updated size data, so that the accuracy of a prediction result is improved; the adjustment mode can also be that the key factors in the historical relevant information in the two materials are increased, deleted and modified by adopting a characteristic engineering and similarity measurement theory so as to increase the comprehensiveness and the referential performance of the data.
B3, acquiring a production process level value and a historical assembly yield corresponding to the training characteristic data, and respectively setting the production process level value and the historical assembly yield as a training level value and a training yield;
in this embodiment, since the size level of the material may affect the assembly accuracy of the material, and the size level of the material may be predicted and determined by the production process level value, when the training data is obtained, the production process level value (CPK value) needs to be added to the measured size data as an influencing factor to consider the situation that the size level of the assembly material in the assembly production has an error, so as to improve the accuracy of the prediction. And setting the CPK value corresponding to the training characteristic data as a training level value, and setting the historical assembly yield corresponding to the training characteristic data as a training yield, so that a training data set can be conveniently established subsequently.
The information in the assembly failure mode library comprises arrangement modes between the housing and the terminals of different batches, measurement sizes of the housing and the terminals of different batches and assembly yield corresponding to the arrangement modes, the three data form a group of data samples,
step b4, establishing the training data set based on the training characteristic data, the training level value and the training yield.
In this embodiment, a training data set is created according to training feature data, training level values, and training yields, where the training feature data and the training level values are used as independent variable feature columns, and the training yields are used as dependent variable result columns, and are combined into a training data set.
And c, training the initial prediction model according to the training data set to obtain an assembly yield prediction model.
In this embodiment, the training data set is input into the initial prediction model, and the training of the initial prediction model requires multiple sets of training data sets to perform multiple tests and verifications, so as to obtain the assembly yield prediction model.
Further, the step of training the initial prediction model according to the training data set to obtain an assembly yield prediction model includes:
step c1, inputting the training data set into the initial prediction model, and performing prediction processing on the training data set based on the initial prediction model to obtain a corresponding training yield;
in this embodiment, the training data set at least includes an independent variable feature sequence composed of training feature data and training level values, and a plurality of training data sets for training are input to the initial prediction model to perform polynomial fitting processing, so as to obtain model coefficients. And (3) updating the formula (2) by the obtained model coefficient, so that a complete assembly yield prediction model is obtained through training optimization of the initial prediction model. And inputting a group of data samples of independent variable characteristic columns in a plurality of groups of training data sets for verification into the assembly yield prediction model to obtain the corresponding training yield.
Step c2, comparing the training yield with a preset yield to obtain a comparison result;
in this embodiment, the training data set further includes a dependent variable result column formed by training yields corresponding to the training characteristic data and the training level value, and a yield result in the dependent variable result column corresponds to one of the data samples, and the yield result is a predicted yield.
And c3, adjusting parameters of the initial prediction model based on the comparison result, determining whether a preset training completion condition is met, and if not, returning to the step of inputting the training data set into the initial prediction model until the preset training completion condition is met to obtain the assembly yield prediction model.
In the embodiment, an error threshold is set, and the error threshold is used as an allowable error range, and in the range, the prediction accuracy obtained by the assembly yield prediction model training is up to the standard, a scheduling combination can be obtained through material information, and the training assembly yield corresponding to the scheduling combination is obtained; when the error value between the training yield and the preset yield during training is larger than the error threshold, the prediction is abnormal, corresponding adjustment needs to be made during training, for example, the step length, the convergence, the training duration, the training stopping time and the like of the training are adjusted, the next training is guided to be carried out in the correct direction, and the prediction accuracy of the assembly yield prediction model is ensured.
It should be noted that the preset training completion condition refers to at least one of an error value within an error threshold and an accuracy rate above a low accuracy rate value, where the low accuracy rate value refers to a preset minimum accuracy rate, that is, a ratio obtained by calculating a training yield rate obtained according to the initial prediction model and the preset yield rate needs to be higher than the minimum accuracy rate, so that the training completion condition can be met, and the training is completed to obtain the assembly yield rate prediction model.
Step S20, inputting target material information of a target material into the assembly yield prediction model to obtain a target arrangement of the target material and a target assembly yield corresponding to the target arrangement;
in this embodiment, the target material refers to a housing and a terminal material to be assembled, the target material information refers to measurement size information of the housing and the terminal material, the measurement size information includes existing inventory and measurement size information of the housing and the terminal material of all the current batches, and the measurement size information is input into the assembly prediction model updated by the formula (2), so as to obtain at least one target arrangement of the target material and a target assembly yield corresponding to the target arrangement. And selecting a target arrangement corresponding to the maximum target assembly yield according to the plurality of target assembly yields, wherein the target arrangement is an optimal arrangement mode.
Further, after the step of inputting the target material information of the target material into the assembly yield prediction model to obtain the target arrangement of the target material and the target assembly yield corresponding to the target arrangement, the method includes:
step d, comparing the target assembly yield with the actual assembly yield to obtain a yield difference;
in this embodiment, the work order data is automatically generated according to the target scheduling result, the work order data is stored in an MES System (Manufacturing Execution System) for assembly, and the actual assembly yield is obtained by the ratio of the number of the qualified products to the number of the unqualified products. And comparing the target assembly yield with the actual assembly yield to obtain a yield difference between the target assembly yield and the actual assembly yield.
Step e, if the yield difference value is larger than the yield difference threshold value, outputting adjustment information;
in this embodiment, the yield difference threshold is preset for determining whether the actual assembly yield meets the standard, and when the yield difference is greater than the yield difference threshold, it indicates that the error between the target assembly yield and the actual assembly yield is large, and at this time, adjustment information needs to be output to improve the assembly yield prediction model. The specific adjustment mode can be that target material information, target arrangement and actual assembly yield in the assembly are stored in an assembly failure mode library, and the determination of an influence factor influencing the assembly yield in the assembly failure mode library and historical related information used for training the model are adjusted to increase the number of samples and the accuracy of the samples, so that the model is subjected to parameter optimization in the process of training the model; the method can also be used for measuring and calculating key factors through an assembly graph and position mechanism between two materials and adding the key factors into an assembly yield prediction model.
When the yield difference is smaller than or equal to the yield difference threshold, it indicates that the error between the target assembly yield and the actual assembly yield belongs to a normal allowable range, that is, the target arrangement and the target assembly yield obtained by the assembly yield prediction model are accurate, so that the assembly line can continuously use the assembly yield prediction model to optimally arrange the materials coming from the upstream and predict the corresponding assembly yield, and assembly production is performed according to the optimal arrangement mode, thereby achieving the effects of high arrangement efficiency and stable arrangement yield.
And f, adjusting and optimizing the assembly yield prediction model according to the adjustment information.
In this embodiment, corresponding operations are performed according to the adjustment mode in the adjustment information, so as to adjust and optimize the assembly yield prediction model.
The application provides an assembly yield control method, which comprises the following steps: obtaining an assembly yield prediction model; inputting target material information of a target material into the assembly yield prediction model to obtain a target arrangement of the target material and a target assembly yield corresponding to the target arrangement; the assembly yield prediction model is obtained according to training material information, a training scheduling combination and a training assembly yield training initial prediction model; the target material information of the target material to be assembled is input into the assembly yield prediction model by acquiring the assembly yield prediction model, the assembly yield prediction model can automatically output the arrangement mode, namely the target arrangement, and meanwhile, the output assembly yield corresponding to the target arrangement is also output, when the assembly yield is maximum, the corresponding target arrangement is the optimal arrangement mode, so that the automatic arrangement is realized on the basis of controlling the assembly yield, and the efficiency of material arrangement is improved.
In addition, an assembly yield control apparatus is further provided in an embodiment of the present application, and with reference to fig. 3, the apparatus includes:
a first obtaining module 10, obtaining an assembly yield prediction model;
the first determining module 20 inputs target material information of a target material into the assembly yield prediction model to obtain a target arrangement of the target material and a target assembly yield corresponding to the target arrangement;
further, the assembly yield control apparatus further includes:
the modeling module is used for establishing an initial prediction model;
the second acquisition module is used for acquiring a training data set, wherein the training data set comprises training material information, a training arrangement combination and a training assembly yield;
and the second determining module trains the initial prediction model according to the training data set to obtain an assembly yield prediction model.
Further, the modeling module includes:
the first acquisition submodule acquires historical related information of the assembly material;
the first determining submodule determines key factors influencing assembly precision in the assembly materials according to the historical relevant information;
and the modeling submodule is used for establishing a relation model between the key factors and the predicted yield and determining an initial prediction model.
Further, the first determination submodule includes:
the obtaining subunit is used for obtaining key point positions influencing the assembling precision in the assembling materials;
the first determining subunit calculates size data according to the figure position size and the measurement size of the key point location to obtain a production process level numerical value corresponding to the key point location;
the self-calculation unit is used for calculating and obtaining the overall error level of different key point positions of the assembly material according to the production process level numerical value and a preset calculation model;
the comparison subunit compares the overall error level with a preset error level;
and the second determining subunit determines the key point position corresponding to the overall error level as a key factor if the overall error level is greater than the preset error level.
Further, the second obtaining module includes:
the database building submodule builds a poor assembly mode database according to the historical relevant information and the historical assembly yield;
the second determining submodule is used for carrying out feature screening on the historical relevant information according to the key factors to obtain training feature data;
the second obtaining submodule is used for obtaining a production process level numerical value and a historical assembly yield corresponding to the training characteristic data and setting the production process level numerical value and the historical assembly yield as a training level numerical value and a training yield respectively;
and the set building submodule is used for building the training data set based on the training characteristic data, the training level numerical value and the training yield.
Further, the second obtaining module further includes:
the first comparison sub-module is used for comparing the historical assembly yield with a yield standard value;
a third determining submodule, configured to, if the historical assembly yield is smaller than the yield reaching standard value, analyze historical relevant information corresponding to the historical assembly yield to obtain an influence factor influencing the assembly yield;
and the optimization submodule adjusts the size of the graph in the historical relevant information according to the influence factors and updates the training data set.
Further, the second determining module comprises:
the fourth determining submodule inputs the training data set into the initial prediction model, and performs prediction processing on the training data set based on the initial prediction model to obtain a corresponding training yield;
the second comparison submodule compares the training yield with a preset yield to obtain a comparison result;
and the training submodule is used for adjusting parameters of the initial prediction model based on the comparison result, determining whether a preset training completion condition is met, and returning to the step of inputting the training data set into the initial prediction model if the preset training completion condition is not met until the preset training completion condition is met, so that the assembly yield prediction model is obtained.
Further, the assembly yield control apparatus further includes:
the comparison module is used for comparing the target assembly yield with the actual assembly yield to obtain a yield difference;
the output module is used for outputting adjustment information if the yield difference value is larger than the yield difference threshold value;
and the adjusting module is used for adjusting and optimizing the assembly yield prediction model according to the adjusting information.
The specific implementation of the assembly yield control apparatus of the present application is basically the same as the embodiments of the assembly yield control method, and is not described herein again.
In addition, an embodiment of the present application further provides a computer-readable storage medium, where an assembly yield control program is stored on the storage medium, and the assembly yield control program, when executed by a processor, implements the steps of the assembly yield control method as described below.
For the embodiments of the assembly yield control apparatus and the computer-readable storage medium, reference may be made to the embodiments of the assembly yield control method, and details are not repeated here.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The above-mentioned serial numbers of the embodiments of the present application are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present application may be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal device (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present application.
The above description is only a preferred embodiment of the present application, and not intended to limit the scope of the present application, and all modifications of equivalent structures and equivalent processes, which are made by the contents of the specification and the drawings of the present application, or which are directly or indirectly applied to other related technical fields, are included in the scope of the present application.
Claims (7)
1. An assembly yield control method, comprising:
acquiring historical related information of the assembly materials, wherein the historical related information comprises information of the drawing position size, the measurement size and the production model number of the assembly materials of different batches;
obtaining key point positions affecting assembly precision in the assembly materials, calculating size data according to the graph position size and the measurement size of the key point positions to obtain production process level numerical values corresponding to the key point positions, and calculating to obtain the overall error levels of different key point positions of the assembly materials according to the production process level numerical values and a preset calculation model;
comparing the overall error level with a preset error level, and if the overall error level is greater than the preset error level, determining a key point position corresponding to the overall error level as a key factor;
establishing a relation model between the key factors and the prediction yield, determining an initial prediction model, obtaining a training data set, training the initial prediction model according to the training data set, and obtaining an assembly yield prediction model, wherein the training data set comprises training material information, a training scheduling combination and a training assembly yield;
inputting target material information of a target material into the assembly yield prediction model to obtain a target arrangement of the target material and a target assembly yield corresponding to the target arrangement;
the assembly yield prediction model is obtained according to training material information, a training scheduling combination and a training assembly yield training initial prediction model.
2. The assembly yield control method of claim 1, wherein the step of obtaining a training data set comprises:
establishing an assembly failure mode library according to the historical relevant information and the historical assembly yield;
performing feature screening on the historical relevant information according to the key factors to obtain training feature data;
acquiring a production process level numerical value and a historical assembly yield corresponding to the training characteristic data, and respectively setting the production process level numerical value and the historical assembly yield as a training level numerical value and a training yield;
establishing the training data set based on the training feature data, the training level numerical value, and the training yield.
3. The assembly yield control method according to claim 2, wherein after the step of creating an assembly failure pattern library based on the historical information and historical assembly yields, the method comprises:
comparing the historical assembly yield with a yield reaching standard value;
if the historical assembly yield is smaller than the yield reaching standard value, analyzing historical relevant information corresponding to the historical assembly yield to obtain an influence factor influencing the assembly yield;
and adjusting the size of the map bit in the historical related information according to the influence factor, and updating the training data set.
4. The assembly yield control method of claim 1, wherein the step of training the initial prediction model according to the training data set to obtain an assembly yield prediction model comprises:
inputting the training data set into the initial prediction model, and performing prediction processing on the training data set based on the initial prediction model to obtain a corresponding training yield;
comparing the training yield with a preset yield to obtain a comparison result;
and adjusting parameters of the initial prediction model based on the comparison result, determining whether a preset training completion condition is met, and if not, returning to the step of inputting the training data set into the initial prediction model until the preset training completion condition is met to obtain an assembly yield prediction model.
5. The assembly yield control method according to any one of claims 1 to 4, wherein after the step of inputting target material information of the target material into the assembly yield prediction model to obtain the target arrangement of the target material and the target assembly yield corresponding to the target arrangement, the method comprises:
comparing the target assembly yield with the actual assembly yield to obtain a yield difference;
if the yield difference value is larger than the yield difference threshold value, outputting adjustment information;
and adjusting and optimizing the assembly yield prediction model according to the adjustment information.
6. An assembly yield control apparatus, comprising: a memory, a processor and an assembly yield control program stored on the memory and executable on the processor, the assembly yield control program when executed by the processor implementing the steps of the assembly yield control method according to any one of claims 1 to 5.
7. A computer-readable storage medium, having an assembly yield control program stored thereon, which when executed by a processor, implements the steps of the assembly yield control method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111224221.3A CN113655768B (en) | 2021-10-21 | 2021-10-21 | Assembly yield control method, equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111224221.3A CN113655768B (en) | 2021-10-21 | 2021-10-21 | Assembly yield control method, equipment and computer readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113655768A CN113655768A (en) | 2021-11-16 |
| CN113655768B true CN113655768B (en) | 2022-02-15 |
Family
ID=78484356
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111224221.3A Active CN113655768B (en) | 2021-10-21 | 2021-10-21 | Assembly yield control method, equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113655768B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114749499A (en) * | 2022-06-13 | 2022-07-15 | 深圳市信润富联数字科技有限公司 | Descaling nozzle control method and device, electronic equipment and readable storage medium |
| CN115034098A (en) * | 2022-08-11 | 2022-09-09 | 深圳市信润富联数字科技有限公司 | Wind power algorithm model verification method, device, equipment and storage medium |
| CN116119284B (en) * | 2022-12-16 | 2023-11-24 | 工业富联(杭州)数据科技有限公司 | Material assembling method, device, equipment and medium based on artificial intelligence |
| CN115685949A (en) * | 2022-12-29 | 2023-02-03 | 深圳市信润富联数字科技有限公司 | Method and device for adjusting data sampling frequency in discrete machining production process |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6823287B2 (en) * | 2002-12-17 | 2004-11-23 | Caterpillar Inc | Method for predicting the quality of a product |
| JP2006023238A (en) * | 2004-07-09 | 2006-01-26 | Fujitsu Ten Ltd | Assembly quality evaluation method and system |
| JP6233038B2 (en) * | 2014-01-16 | 2017-11-22 | 富士通株式会社 | Assembly yield prediction apparatus, assembly yield prediction program, and assembly yield prediction method |
| CN111158314B (en) * | 2019-12-31 | 2022-10-25 | 东南大学 | Satellite partial assembly precision control method based on digital twinning technology |
| CN111428373A (en) * | 2020-03-30 | 2020-07-17 | 苏州惟信易量智能科技有限公司 | Product assembly quality detection method, device, equipment and storage medium |
| CN112231966B (en) * | 2020-09-08 | 2023-04-07 | 合肥学院 | Cooperative robot assemblability prediction system and method based on digital twinning |
| CN112686372A (en) * | 2020-12-28 | 2021-04-20 | 哈尔滨工业大学（威海） | Product performance prediction method based on depth residual GRU neural network |
| CN112799369A (en) * | 2021-01-15 | 2021-05-14 | 北京理工大学 | Product assembly process control method and device |
| CN113516285B (en) * | 2021-05-12 | 2024-02-13 | 中船重工鹏力（南京）智能装备系统有限公司 | Product quality analysis and prediction method of automatic assembly detection production line in production |
| CN113504768B (en) * | 2021-08-05 | 2022-06-03 | 东华大学 | High-precision product digital twin computability method for assembly quality prediction |
-
2021
- 2021-10-21 CN CN202111224221.3A patent/CN113655768B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113655768A (en) | 2021-11-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113655768B (en) | Assembly yield control method, equipment and computer readable storage medium | |
| EP3811241B1 (en) | Product performance prediction modeling method and apparatus, computer device, computer-readable storage medium, and product performance prediction method and prediction system | |
| US20140222376A1 (en) | Method for searching, analyzing, and optimizing process parameters and computer program product thereof | |
| US20150371134A1 (en) | Predicting circuit reliability and yield using neural networks | |
| CN113657820B (en) | Production line batching method, device, equipment and readable storage medium | |
| US20150276558A1 (en) | Metrology sampling method and computer program product thereof | |
| CN114819636B (en) | Industrial production data processing method and system based on SPC detection | |
| CN110309052A (en) | A kind of data verification method and relevant device | |
| CN101118422A (en) | Virtual measurement prediction generated by semi-conductor, method for establishing prediction model and system | |
| CN110874685A (en) | Intelligent electric energy meter running state distinguishing method and system based on neural network | |
| CN115373370A (en) | Method and system for monitoring running state of programmable controller | |
| KR102247945B1 (en) | Method of predicting processing fault considering process factor | |
| CN116108642A (en) | Parameter debugging model training method, debugging method and equipment for machine | |
| CN108537249B (en) | Industrial process data clustering method for density peak clustering | |
| US8406904B2 (en) | Two-dimensional multi-products multi-tools advanced process control | |
| CN116382219A (en) | Motor production process optimization method and system based on online measurement technology | |
| CN114219461B (en) | Production control method and device based on multiple production working conditions | |
| CN112257347B (en) | Simulation system of power transformation equipment | |
| CN113627755A (en) | Test method, device, equipment and storage medium for intelligent terminal factory | |
| CN115130737A (en) | Machine learning method based on working condition, related device and medium program product | |
| CN114139643A (en) | Monoglyceride quality detection method and system based on machine vision | |
| CN115839617A (en) | Sintering temperature control method and device | |
| CN114118004A (en) | Method for manufacturing customized chip | |
| CN117634324B (en) | Casting mold temperature rapid prediction method based on convolutional neural network | |
| CN113625669B (en) | Product quality prediction method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |