CN110348954B - A Mass Customization-Oriented Partitioning Method for Complex Process Modules - Google Patents

A Mass Customization-Oriented Partitioning Method for Complex Process Modules Download PDF

Info

Publication number
CN110348954B
CN110348954B CN201910555601.1A CN201910555601A CN110348954B CN 110348954 B CN110348954 B CN 110348954B CN 201910555601 A CN201910555601 A CN 201910555601A CN 110348954 B CN110348954 B CN 110348954B
Authority
CN
China
Prior art keywords
matrix
module
correlation
fuzzy
represent
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
Application number
CN201910555601.1A
Other languages
Chinese (zh)
Other versions
CN110348954A (en
Inventor
范波
宋晓明
付主木
许惠
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Henan University of Science and Technology
Original Assignee
Henan University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Henan University of Science and Technology filed Critical Henan University of Science and Technology
Priority to CN201910555601.1A priority Critical patent/CN110348954B/en
Publication of CN110348954A publication Critical patent/CN110348954A/en
Application granted granted Critical
Publication of CN110348954B publication Critical patent/CN110348954B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2468—Fuzzy queries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/28—Databases characterised by their database models, e.g. relational or object models
    • G06F16/284—Relational databases
    • G06F16/285—Clustering or classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0633—Workflow analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0621—Item configuration or customization

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Fuzzy Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Software Systems (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Mathematical Physics (AREA)
  • Development Economics (AREA)
  • General Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Operations Research (AREA)
  • Educational Administration (AREA)
  • Game Theory and Decision Science (AREA)
  • Probability & Statistics with Applications (AREA)
  • Computational Linguistics (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本发明公开了一种面向大规模定制的复杂工艺模块划分方法，借助与工件相对应工序之间的特征关系，提出在工艺层面对模型进行划分的方法，通过构建工艺模块之间的模糊关联矩阵，利用模糊聚类分析方法对其进行转换求解，获得传递闭包矩阵，根据不同分区阈值序列形成模块聚类图，通过选取不同的λ值，得到不同的模块划分方案，逻辑紧密，控制精确、高效，以期实现快速高效的生产出适应市场和技术变化的个性化成品以及完善工艺的高层次要求。

Figure 201910555601

The invention discloses a large-scale customization-oriented complex process module division method. With the help of the characteristic relationship between the processes corresponding to the workpiece, a method for dividing the model at the process level is proposed. By constructing a fuzzy correlation matrix between process modules , using the fuzzy clustering analysis method to transform and solve it, obtain the transitive closure matrix, and form a module clustering diagram according to different partition threshold sequences. By selecting different λ values, different module partitioning schemes are obtained. High efficiency, in order to achieve the high-level requirements of fast and efficient production of personalized finished products that adapt to market and technological changes and perfect processes.

Figure 201910555601

Description

Complex process module dividing method for large-scale customization
Technical Field
The invention belongs to the technical field of complex process planning, and relates to a complex process module dividing method for large-scale customization.
Background
At present, rapid economic development and continuous technical progress promote the personalized customization process to gradually gain the favor of the economic market, provide customized service for customers, comprehensively improve the satisfaction degree of the customers, and become a competitive trend among modern enterprises.
The advantages of the large-scale customized production mode and the mass production are combined, the personalized requirements of customers are met, meanwhile, the production task can be guaranteed to be completed within a short production period at low cost, and the unique advantages of the large-scale customized production mode attract wide attention of manufacturing enterprises. One of the large-scale customization core technologies is a modularization method, which is used for partitioning a general module, a customization module and a personalized module which have relatively independent functions, structures and performances on the basis of structural analysis of different functional characteristics of a product family or different geometric characteristics and performance characteristics of the same function, and can quickly and efficiently partition the modules of parts of a complex product through matching combination among the modules to produce a process product and a service which meet the personalized requirements of customers.
At present, a modular design method is widely applied to system module division of a product family, but the conventional performance division of the product is considered more in the existing division method, the idea of individual requirements is not integrated, and the method is not suitable for creating a process module with complex product. Compared with the establishment of a product component module, the establishment of a process module needs to incorporate process information of a product, and in order to solve the problem, the invention designs a complex process module dividing method oriented to large-scale customization, and intelligently divides the processing process of the product by taking the characteristic attribute of the process as a criterion, thereby realizing the personalized requirements of customers.
Disclosure of Invention
In view of the above, to solve the above-mentioned deficiencies of the prior art, the present invention aims to provide a method for partitioning a complex process module oriented to large-scale customization, which has compact logic, precise control and high efficiency, and provides a method for partitioning a model at a process level by means of a characteristic relationship between processes corresponding to a workpiece.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a large-scale customization-oriented complex process module dividing method comprises the following steps:
s1: establishing module characteristic attributes by a modular mapping mechanism, establishing the relevance among process modules by the attributes, and describing the relevance as replaceable relevance and influence degree relevance;
s2: the correlation degree between the process modules is related to the substitutability, the influence degree and the like, and is represented by a mathematical model of a fuzzy correlation matrix R: r ═ ωfF+ωpP, where F, P denotes the defined influence correlation matrix and the alternative correlation matrix, ω, respectivelyf、ωpRepresenting the weight coefficient, ω, to which the characteristic correspondsf、ωp∈[0，1]And satisfies the following conditions: omegaf+ωp＝1；
S3: converting and solving the fuzzy incidence matrix model by using a fuzzy clustering analysis method to obtain a transfer closure matrix;
s31: determining the correlation among all the working procedures and determining the correlation degree among all the process modules;
s32: and (3) data standardization treatment: according to different original formats, expression formats and level quantization modes, translation and range transformation are adopted to carry out data standardization processing, and each data of the fuzzy association matrix is compressed to [0, 1 ];
s33: establishing a fuzzy similarity matrix R': the fuzzy incidence matrix R and the fuzzy similarity matrix R' are in equivalent relation, and a direct Euclidean distance method is applied, wherein R is(i,j)＝1-c×d(xi,xj) Wherein C is a parameter of any selected region, so that r is more than or equal to 0(i,j)≤1，d(xi,xj) Denotes xiAnd xjThe distance between:
Figure GDA0003369678340000031
s34: solving a transitive closure matrix
Figure GDA0003369678340000032
: according to the theorem, the fuzzy equivalent matrix is obtained by using the fuzzy incidence matrix R by a quadratic method
Figure GDA0003369678340000033
I.e. transitive closure matrix
Figure GDA0003369678340000034
(ii) a And, make
Figure GDA0003369678340000035
S35: solving the truncation matrix Rλ＝λ(i,j): the method comprises the following steps that lambda is used as a metric value to represent a truncation matrix coefficient, a fuzzy relation matrix between modules is truncated, elements which are larger than or equal to lambda in a fuzzy association matrix R are provided with the numerical value of 1, the numerical value which is smaller than the lambda element is provided with 0, the elements with the numerical value of 1 in the same row or column are gathered into the same module, the rest elements independently become the modules, and the more the lambda value is, the more detailed the module division is; different process module clustering division results can be obtained by selecting different lambda values;
s36: with the help of Matlab tool and the application of fuzzy clustering analysis method, the standardized matrix, fuzzy similar matrix, transfer closure matrix can be calculated in turn according to the original incidence matrix, and finally the fuzzy incidence matrix R is cut by using lambda-cut matrix to form a module clustering graph:
Figure GDA0003369678340000036
s4: and forming an integral module clustering graph according to different partition threshold sequences, and obtaining different module division schemes by selecting different lambda values.
Further, in step S2, the method for establishing the fuzzy association matrix specifically includes the following steps:
a1: and (3) taking the attribute value domain of the output parameter as the correlation for judging each processing procedure, describing the correlation as the correlation of replaceable correlation and influence degree, and expressing a fuzzy correlation matrix R by using a mathematical model: r isi,j＝ωffi,j+ωppi,jWherein r isi,jRepresenting the overall correlation coefficient between module i and module j, fi,j、pi,jRespectively representing the number of subphases, ω, between module i and module jf、ωpRepresenting a weight coefficient corresponding to the characteristic;
a2: the fuzzy correlation matrix R is also called an original matrix R, and R satisfies the self-reflexivity: r (i, j) is not less than 0 and not more than 1, and r (i, i) is r (j, j) is 1; and R satisfies symmetry: r (i, j) ═ R (j, i), the mathematical model of the fuzzy correlation matrix R in step a1 is expressed as:
Figure GDA0003369678340000041
further, in the step A1, ri,jRepresenting the overall correlation coefficient between module i and module j, fi,jRepresenting the correlation coefficient, p, between block i and block ji,jDenotes the correlation coefficient, ω, between block j and block if、ωpRepresenting a weight coefficient corresponding to the characteristic; omegaf、ωp∈[0，1]And satisfies the following conditions: omegaf+ωp＝1。
Further, in the step a1, the output parameters are grain size, yield strength, elongation, residual stress, corrosion resistance, electrical conductivity, hardness, and tensile strength.
Further, in step S32, the normalization formula used in the data normalization process is:
Figure GDA0003369678340000051
wherein, i is 1,2(′i,j)Representing the original data; x is the number of('i,j)man、x('i,j)minRespectively representing the maximum value and the minimum value of the original data; x is the number of(i,j)The data obtained after normalization.
Further, in step S34, a transitive closure matrix is obtained
Figure GDA0003369678340000052
The method specifically comprises the following steps:
a1: set up R0Is R, wherein R0Representing the original incidence matrix of the module;
a2: let R beiKnowing the result, start to compare RiAnd
Figure GDA0003369678340000053
the corresponding element; if it is not
Figure GDA0003369678340000054
Then R isiIs that
Figure GDA0003369678340000055
The transitive closure matrix of (2), at which point computation may be stopped; if not satisfied
Figure GDA0003369678340000056
If yes, continuing to execute the next step;
a3: by algorithmic calculation
Figure GDA0003369678340000057
And setting the result to Ri+1Executing the step A2;
a4: repeating the above steps until finally solving the transfer closure matrix to ensure that
Figure GDA0003369678340000058
The invention has the beneficial effects that:
a method for dividing a complex process module oriented to large-scale customization is compact in logic, accurate in control and high-efficiency, a method for dividing a model on a process layer is provided by means of characteristic relation between processes corresponding to workpieces, a fuzzy association matrix between process modules is constructed, the fuzzy association matrix is converted and solved by using a fuzzy clustering analysis method to obtain a transfer closure matrix, a module clustering diagram is formed according to different partition threshold sequences, different module division schemes are obtained by selecting different lambda values, and therefore personalized finished products adapting to market and technical changes and high-level requirements of the process are rapidly and efficiently produced;
compared with the traditional product family module division method, the method is mainly based on the analysis of mature products of enterprises and processing technologies of products with market demands, combines the process information extraction, the process creative design and the process modularization to construct a modular process design platform, modularizes the serialized product processes, matches the similarity of the modules by combining the module configuration technology and the order requirements, and finally carries out the deformation design of the corresponding degree on the process requirements by the deformation design module to realize the functions of different product requirements.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a functional block diagram of a modular process domain mapping mechanism of the present invention;
FIG. 2 is a process flow diagram of the production of the aluminum/copper sheet strip in the example;
FIG. 3 is a block clustering diagram of the process sequence of the example.
Detailed Description
The following specific examples are given to further clarify, complete and detailed the technical solution of the present invention. The present embodiment is a preferred embodiment based on the technical solution of the present invention, but the scope of the present invention is not limited to the following embodiments.
A large-scale customization-oriented complex process module dividing method comprises the following steps:
s1: establishing module characteristic attributes by a modular mapping mechanism, establishing the relevance among process modules by the attributes, and describing the relevance as replaceable relevance and influence degree relevance;
s2: the correlation degree between the process modules is related to the substitutability, the influence degree and the like, and is represented by a mathematical model of a fuzzy correlation matrix R: r ═ ωfF+ωpP；
S3: converting and solving the fuzzy incidence matrix model by using a fuzzy clustering analysis method to obtain a transfer closure matrix;
s31: determining the correlation among all the working procedures and determining the correlation degree among all the process modules;
s32: and (3) data standardization treatment: according to different original formats, expression formats and level quantization modes, translation and range transformation are adopted to carry out data standardization processing, and each data of the fuzzy association matrix is compressed to [0, 1 ];
s33: establishing a fuzzy similarity matrix R': the fuzzy incidence matrix R and the fuzzy similarity matrix R' are in equivalent relation, and a direct Euclidean distance method is applied, wherein R is(i,j)＝1-c×d(xi,xj) Wherein C is a parameter of any selected region, so that r is more than or equal to 0(i,j)≤1，d(xi,xj) Denotes xiAnd xjThe distance between:
Figure GDA0003369678340000071
s34: solving a transitive closure matrix
Figure GDA0003369678340000072
: according to the theorem, the fuzzy equivalent matrix is obtained by using the fuzzy incidence matrix R by a quadratic method
Figure GDA0003369678340000073
I.e. transitive closure matrix
Figure GDA0003369678340000074
(ii) a And, make
Figure GDA0003369678340000075
S35: solving the truncation matrix Rλ＝λ(i,j): the method comprises the following steps that lambda is used as a metric value to represent a truncation matrix coefficient, a fuzzy relation matrix between modules is truncated, elements which are larger than or equal to lambda in a fuzzy association matrix R are provided with the numerical value of 1, the numerical value which is smaller than the lambda element is provided with 0, the elements with the numerical value of 1 in the same row or column are gathered into the same module, the rest elements independently become the modules, and the more the lambda value is, the more detailed the module division is; by selectingObtaining different process module clustering division results by taking different lambda values;
s36: with the help of Matlab tool and the application of fuzzy clustering analysis method, the standardized matrix, fuzzy similar matrix, transfer closure matrix can be calculated in turn according to the original incidence matrix, and finally the fuzzy incidence matrix R is cut by using lambda-cut matrix to form a module clustering graph:
Figure GDA0003369678340000081
s4: and forming an integral module clustering graph according to different partition threshold sequences, and obtaining different module division schemes by selecting different lambda values.
Further, in step S2, the method for establishing the fuzzy association matrix specifically includes the following steps:
a1: and (3) taking the attribute value domain of the output parameter as the correlation for judging each processing procedure, describing the correlation as the correlation of replaceable correlation and influence degree, and expressing a fuzzy correlation matrix R by using a mathematical model: r isi,j＝ωffi,j+ωppi,jWherein r isi,jRepresenting the overall correlation coefficient between module i and module j, fi,j、pi,jRespectively representing the number of subphases, ω, between module i and module jf、ωpRepresenting a weight coefficient corresponding to the characteristic;
a2: the fuzzy correlation matrix R is also called an original matrix R, and R satisfies the self-reflexivity: r (i, j) is not less than 0 and not more than 1, and r (i, i) is r (j, j) is 1; and R satisfies symmetry: r (i, j) ═ R (j, i), the mathematical model of the fuzzy correlation matrix R in step a1 is expressed as:
Figure GDA0003369678340000082
further, in the step A1, ri,jRepresenting the overall correlation coefficient between module i and module j, fi,jRepresenting the correlation coefficient, p, between block i and block ji,jDenotes the correlation coefficient, ω, between block j and block if、ωpRepresenting a weight coefficient corresponding to the characteristic; omegaf、ωp∈[0，1]And satisfies the following conditions: omegaf+ωp＝1。
Further, in the step a1, the output parameters are grain size, yield strength, elongation, residual stress, corrosion resistance, electrical conductivity, hardness, and tensile strength.
Further, in step S32, the normalization formula used in the data normalization process is:
Figure GDA0003369678340000091
wherein, i is 1,2(′i,j)Representing the original data; x is the number of('i,j)man、x('i,j)minRespectively representing the maximum value and the minimum value of the original data; x is the number of(i,j)The data obtained after normalization.
Further, in step S34, a transitive closure matrix is obtained
Figure GDA0003369678340000098
The method specifically comprises the following steps:
a1: set up R0Is R, wherein R0Representing the original incidence matrix of the module;
a2: let R beiKnowing the result, start to compare RiAnd
Figure GDA0003369678340000092
the corresponding element; if it is not
Figure GDA0003369678340000093
Then R isiIs that
Figure GDA0003369678340000094
The transitive closure matrix of (2), at which point computation may be stopped; if not satisfied
Figure GDA0003369678340000095
Continuing to execute the next step;
a3: by algorithmic calculation
Figure GDA0003369678340000096
And setting the result to Ri+1Executing the step A2;
a4: repeating the above steps until finally solving the transfer closure matrix to ensure that
Figure GDA0003369678340000097
The specific working process of this embodiment is as follows:
1. as shown in fig. 1, each process module has a model representing its own characteristics, each characteristic model may be represented by a plurality of attributes, when the attributes of the module have a plurality of attribute values, the attribute values may be collectively referred to as variables, a common attribute set capable of distinguishing instance modules is defined as a variable set p (x) of process-like modules, and the variable set is used to distinguish differences between the models of the same type, as shown in table 1. The reason for personalized customization is that module models are predefined in a process model library in the process of configuring process modules according to customer requirements, the models are independent of each other and have a plurality of attributes and attribute values, and different modules are called to be combined according to different requirements of customers. P (x) ═ parti|1≤i≤n}parti＝{vali,j|1≤i≤n,1≤j≤m}，PartiRepresenting configuration constraints between the inner classes;
TABLE 1 Process Module Attribute variables and value ranges
Figure GDA0003369678340000101
2. Modular design and fuzzy incidence matrix model establishment
(1) Establishing a fuzzy incidence matrix R
The correlation among the complex product process procedures refers to the degree of correlation in the aspects of researching the characteristics, performance, structure and the like among the configuration modules. Combined upperThe value ranges of the attribute variables are described, and output parameters such as grain size, yield strength, elongation, residual stress, corrosion resistance, electrical conductivity, hardness, tensile strength and the like are used as the correlation between the various processing procedures, and the correlation is described as a replaceable correlation and an influence degree correlation, as shown in tables 2 and 3. Expressed by a mathematical model: r isi,j＝ωffi,j+ωppi,jWherein r isi,jRepresenting the correlation coefficient between block i and block j, fi,j、pi,jRespectively representing the correlation coefficient, ω, between module i and module jf、ωpRepresenting the weight coefficient, ω, to which the characteristic correspondsf、ωp∈[0，1]And satisfies the following conditions: omegaf+ω p1, thereby forming a fuzzy incidence matrix R of the relation between the process modules;
TABLE 2 alternative correlations
Figure GDA0003369678340000111
TABLE 3 correlation of influence
Figure GDA0003369678340000112
Taking an aluminum/copper plate strip series product produced by a certain enterprise as an example, the production process of the plate strip can be divided into the following steps by carrying out on-site research on the enterprise and integrating the related production processes provided by personnel in the production department: casting, hot rolling, cold rolling and heat treatment. The specific process flow of casting is as follows: batching, feeding, melting, stirring, slagging off, analyzing and sampling, adding alloy to adjust components, stirring, refining, standing, guiding to a furnace, casting and the like; the hot rolling process flow comprises the following steps: ingot casting, surface milling and edge milling, heating, hot rolling (cogging rolling), straightening and surface milling; the cold rolling process flow comprises the following steps: rough rolling, annealing, pickling, annealing, finish rolling, and the final heat treatment is finished product annealing, which is divided into complete annealing and low-temperature annealing, as shown in table 4. Soft, hard or semi-hard products in different states are obtained by controlling the annealing temperature and the heat preservation time, so as to stabilize the size, shape and performance of the material. Analyzing the correlation degree between the whole production process and the whole production process by combining the whole production process and the whole production process, and constructing a correlation matrix between the whole production process and the whole production process;
table 4 full process flow module based on process route
Serial number Name of procedure Serial number Name of procedure
1 Casting blank 11 Intermediate annealing
2 Solution heat treatment 12 Acid pickling
3 Hot rolling 13 Finish rolling
4 Shearing 14 Annealing the finished product
5 Straightening milled surface 15 Cleaning finished product
6 Homogenizing annealing 16 Cutting and slitting
7 Rough rolling 17 Acceptance inspection
8 Intermediate annealing 18 Package (I)
9 Acid pickling 19 Put in storage
10 Pre-finish rolling
(2) Module division method based on fuzzy clustering analysis method
1) And (3) data standardization treatment: the original formats of the data information are different greatly, wherein the data information comprises description type, numerical type, selection type and other information expression formats, and the data grade quantization modes are different, so that the data information cannot be directly applied to operation, the data information needs to be subjected to standardization processing, and each data of the fuzzy matrix is compressed to [0, 1], so that the subsequent algorithm operation is facilitated;
2) establishing a fuzzy similarity matrix R': the fuzzy matrix can be obtained from the characteristic table, the fuzzy matrix R and the fuzzy similar matrix R' are in equivalent relation, and a direct Euclidean distance method is applied: r is(i，j)＝1-c×d(xi，xj) Wherein C is a parameter of any selected region, so that r is more than or equal to 0(i，j)≤1，d(xi，xj) Denotes xiAnd xjThe distance between
Figure GDA0003369678340000121
By using the module division method, an expert evaluates and scores to determine the correlation among all the processes, and if the casting blank and the solution heat treatment process are closely connected two processes and the time interval of the action is small, the influence degree correlation is 0.8, and the alternative correlation is 0.5. The weighting coefficients are taken to be 0.55 and 0.45 respectively according to the formula. The correlation between the two was calculated to be 0.665, and similarly, the influence degree correlation and the alternative correlation between the solution heat treatment and the hot rolling were taken to be 0.4 and 0.1, respectively, and the total correlation was 0.265. The correlation between the other steps was calculated in sequence, as shown in table 5. Forming a fuzzy incidence matrix R after correlation analysis and correlation calculation;
TABLE 5 fuzzy association matrix R of major process blocks
Figure GDA0003369678340000131
3) Solving a transitive closure matrix
Figure GDA0003369678340000132
: according to the fuzzy incidence matrix R obtained by calibration, only one fuzzy similar matrix is obtained
Figure GDA0003369678340000133
As shown in Table 6, has reflexibility and contraindicationSymmetry, but not necessarily transitivity, i.e. R is not necessarily a fuzzy equivalent matrix, and needs to be transformed into a fuzzy equivalent matrix RkAccording to the theorem, the transitive closure matrix is solved by a quadratic method
Figure GDA0003369678340000134
Figure GDA0003369678340000135
Is the fuzzy equivalent matrix R soughtkAs shown in table 7;
TABLE 6 fuzzy similarity matrix R 'for relationships between process modules'
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1 1 0.806 0.622 0.522 0.570 0.575 0.470 0.497 0.428 0.456 0.521 0.448 0.508 0.536 0.463 0.471
2 0.806 1 0.621 0.523 0.557 0.658 0.517 0.561 0.398 0.520 0.596 0.414 0.516 0.610 0.412 0.480
3 0.622 0.621 1 0.726 0.753 0.786 0.558 0.508 0.398 0.529 0.475 0.409 0.517 0.478 0.369 0.528
4 0.522 0.523 0.726 1 0.869 0.635 0.527 0.406 0.383 0.517 0.400 0.372 0.535 0.426 0.410 0.552
5 0.570 0.557 0.753 0.869 1 0.671 0.539 0.453 0.405 0.536 0.445 0.406 0.550 0.455 0.416 0.566
6 0.575 0.658 0.786 0.635 0.671 1 0.625 0.614 0.445 0.596 0.569 0.404 0.544 0.512 0.344 0.526
7 0.470 0.517 0.558 0.527 0.539 0.625 1 0.740 0.657 0.788 0.661 0.559 0.558 0.491 0.422 0.560
8 0.497 0.561 0.508 0.406 0.453 0.614 0.740 1 0.593 0.702 0.791 0.538 0.497 0.543 0.377 0.463
9 0.428 0.398 0.398 0.383 0.405 0.445 0.657 0.593 1 0.661 0.580 0.837 0.494 0 0.441 0.604 0.500
10 0.456 0.520 0.529 0.511 0.536 0.596 0.788 0.702 0.661 1 0.657 0.591 0.606 0.513 0.470 0.615
11 0.521 0.596 0.475 0.400 0.445 0.569 0.661 0.791 0.580 0.657 1 0.545 0.575 0.655 0.450 0.531
12 0.448 0.414 0.409 0.372 0.406 0.404 0.559 0.538 0.837 0.591 0.545 1 0.507 0.491 0.680 0.540
13 0.508 0.516 0.517 0.535 0.550 0.544 0.558 0.497 0.490 0.606 0.575 0.507 1 0.711 0.597 0.777
14 0.536 0.610 0.478 0.426 0.455 0.512 0.491 0.543 0.441 0.513 0.655 0.491 0.711 1 0.583 0.661
15 0.463 0.412 0.369 0.410 0.416 0.344 0.422 0.377 0.604 0.470 0.450 0.680 0.597 0.583 1 0.663
16 0.471 0.480 0.528 0.552 0.566 0.526 0.560 0.463 0.500 0.615 0.531 0.540 0.777 0.661 0.663 1
TABLE 7 transitive closure matrix
Figure GDA0003369678340000141
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1 1 0.80 0.62 0.52 0.57 0.57 0.47 0.49 0.42 0.45 0.52 0.44 0.50 0.53 0.46 0.47
2 0.80 1 0.62 0.52 0.55 0.65 0.51 0.56 0.39 0.52 0.59 0.41 0.51 0.61 0.41 0.48
3 0.62 0.62 1 0.72 0.75 0.78 0.55 0.50 0.39 0.52 0.47 0.40 0.51 0.47 0.36 0.52
4 0.52 0.52 0.72 1 0.86 0.63 0.52 0.40 0.38 0.51 0.40 0.37 0.53 0.42 0.41 0.55
5 0.57 0.55 0.75 0.86 1 0.67 0.53 0.45 0.40 0.53 0.44 0.40 0.55 0.45 0.41 0.56
6 0.57 0.65 0.78 0.63 0.67 1 0.62 0.61 0.44 0.59 0.56 0.40 0.54 0.51 0.34 0.52
7 0.47 0.51 0.55 0.52 0.53 0.62 1 0.74 0.65 0.78 0.66 0.55 0.55 0.49 0.42 0.56
8 0.49 0.56 0.50 0.40 0.45 0.61 0.74 1 0.59 0.70 0.79 0.53 0.49 0.54 0.37 0.46
9 0.42 0.39 0.39 0.38 0.40 0.44 0.65 0.59 1 0.66 0.58 0.83 0.49 0.44 0.60 0.50
10 0.45 0.52 0.52 0.51 0.53 0.59 0.78 0.70 0.66 1 0.65 0.59 0.60 0.51 0.47 0.61
11 0.52 0.59 0.47 0.40 0.44 0.56 0.66 0.79 0.58 0.65 1 0.54 0.57 0.65 0.45 0.53
12 0.44 0.41 0.40 0.37 0.40 0.40 0.55 0.53 0.83 0.59 0.54 1 0.50 0.49 0.68 0.54
13 0.50 0.51 0.51 0.53 0.55 0.54 0.55 0.49 0.49 0.60 0.57 0.50 1 0.71 0.59 0.77
14 0.53 0.61 0.47 0.42 0.45 0.51 0.49 0.54 0.44 0.51 0.65 0.49 0.71 1 0.58 0.66
15 0.46 0.41 0.36 0.41 0.41 0.34 0.42 0.37 0.60 0.47 0.45 0.68 0.59 0.58 1 0.66
16 0.47 0.48 0.52 0.55 0.56 0.52 0.56 0.46 0.50 0.61 0.53 0.54 0.77 0.66 0.66 1
4) Solving the truncation matrix Rλ＝λ(i，j): the method comprises the following steps that lambda is used as a metric value to represent a truncation matrix coefficient, a fuzzy relation matrix among modules is truncated, elements which are larger than or equal to lambda in a fuzzy relation matrix R are provided with the numerical value of 1, the numerical value which is smaller than the lambda element is provided with 0, the elements with the numerical value of 1 in the same row or column are gathered into the same module, the rest elements independently become the modules, and the more the lambda value is, the more detailed the module division is; different process module clustering division results can be obtained by selecting different lambda values; and sequentially calculating a standardized matrix, a fuzzy similar matrix and a transfer closure matrix according to the original incidence matrix by using a Matlab tool and a fuzzy clustering analysis algorithm. Finally, a dynamic clustering module can be formed by cutting R by using the lambda-cut matrix, as shown in FIG. 3. Large to small threshold partition sequence: {10.84770.80140.78230.77090.74740.72740.71670.71210.68930.68190.66250.6490.64310.60920.5865}, the full process sequence is divided into different process categories. When lambda is>0.6625, the number of divided modules is excessive and deviates from the actual situation. When λ is 0.6625, all process modules can be classified into 4 major categories, which are fusion casting module, hot rolling module, cold rolling module, and heat treatment module. The cleaning module as an auxiliary process module can be divided independently and is basically consistent with the actual situation.
In conclusion, the method for dividing the complex process modules for large-scale customization is compact in logic, accurate in control and efficient, a method for dividing the models on the process level is provided by means of characteristic relations between processes corresponding to workpieces, a fuzzy association matrix between the process modules is constructed, the fuzzy association matrix is converted and solved by using a fuzzy clustering analysis method to obtain a transfer closure matrix, a module clustering diagram is formed according to different partition threshold sequences, different module division schemes are obtained by selecting different lambda values, and therefore personalized finished products adapting to market and technical changes can be produced quickly and efficiently, and high-level requirements of the process are met;
compared with the traditional product family module division method, the method is mainly based on the analysis of mature products of enterprises and processing technologies of products with market demands, combines the process information extraction, the process creative design and the process modularization to construct a modular process design platform, modularizes the serialized product processes, matches the similarity of the modules by combining the module configuration technology and the order requirements, and finally carries out the deformation design of the corresponding degree on the process requirements by the deformation design module to realize the functions of different product requirements.
The principal features, principles and advantages of the invention have been shown and described above. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to explain the principles of the invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the invention as expressed in the following claims. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (6)

1.一种面向大规模定制的复杂工艺模块划分方法，其特征在于：包括以下步骤：1. a complex process module dividing method for mass customization, is characterized in that: comprise the following steps: S1：由模块化映射机制建立模块特征属性，并由属性确立工艺模块之间的关联性，将关联性描述为可替换相关性和影响度相关性；S1: The module feature attribute is established by the modular mapping mechanism, and the correlation between the process modules is established by the attribute, and the correlation is described as replaceable correlation and influence correlation; S2：工艺模块之间的相互关联程度与可替换性特性、影响度相关性相关，用模糊关联矩阵R的数学模型进行表示：R＝ωfF+ωpP，其中，F、P分别表示所定义的影响度相关性矩阵和可替代相关性矩阵，ωf、ωp表示特性所对应的权重系数，ωf、ωp∈[0，1]，且满足：ωf+ωp＝1；S2: The degree of correlation between process modules is related to the characteristics of replaceability and the degree of influence, which is represented by the mathematical model of the fuzzy correlation matrix R: R=ω f F+ω p P, where F and P represent respectively The defined influence degree correlation matrix and alternative correlation matrix, ω f , ω p represent the weight coefficients corresponding to the characteristics, ω f , ω p ∈ [0, 1], and satisfy: ω f +ω p =1 ; S3：利用模糊聚类分析方法，对模糊关联矩阵模型进行转换求解，获得传递闭包矩阵；S3: Use the fuzzy cluster analysis method to transform and solve the fuzzy correlation matrix model to obtain the transitive closure matrix; S31：确定各工序之间的相关性，确定各工艺模块之间的相互关联程度；S31: Determine the correlation between the various processes, and determine the degree of mutual correlation between the various process modules; S32：数据标准化处理：根据原本格式、表达格式、等级量化方式不同，采用平移、极差变换进行数据标准化处理，将模糊关联矩阵的每一个数据压缩到[0，1]上；S32: Data standardization processing: According to the original format, expression format, and grade quantization method, use translation and range transformation to perform data standardization processing, and compress each data of the fuzzy correlation matrix to [0, 1]; S33：建立模糊相似矩阵R′：模糊关联矩阵R与模糊相似矩阵R′为等效关系，运用直接欧几里得距离法:r(i,j)＝1-c×d(xi,xj)，其中，C为任意选取的参数，使得0≤r(i,j)≤1，d(xi,xj)表示xi与xj之间的距离：
Figure FDA0003369678330000011
S33: Establish a fuzzy similarity matrix R': the fuzzy correlation matrix R and the fuzzy similarity matrix R' are equivalent, and the direct Euclidean distance method is used: r (i,j) =1-c×d(x i ,x j ), where C is an arbitrarily selected parameter such that 0≤r (i,j) ≤1, and d(x i ,x j ) represents the distance between x i and x j :
Figure FDA0003369678330000011
 S34：求传递闭包矩阵
Figure FDA0003369678330000012
根据定理，将模糊关联矩阵R用二次方法求得模糊等价矩阵
Figure FDA0003369678330000021
即传递闭包矩阵
Figure FDA0003369678330000022
并且，使得
Figure FDA0003369678330000023
S34: Find the transitive closure matrix
Figure FDA0003369678330000012
According to the theorem, the fuzzy equivalent matrix is obtained by quadratic method of fuzzy correlation matrix R
Figure FDA0003369678330000021
i.e. transitive closure matrix
Figure FDA0003369678330000022
and, so that
Figure FDA0003369678330000023
 S35：求截矩阵Rλ＝λ(i,j)：其中，λ作为度量值表示截矩阵系数，对模块之间的模糊关系矩阵进行截割，模糊关联矩阵R中大于或者等于λ的元素，其数值取1，小于λ元素的数值取0，在同一行或列中数值为1的元素聚集为同一模块，剩余的元素则单独成为模块，λ值越大，则模块划分越详细；通过选取不同的λ值，便可以得到不同的工艺模块聚类划分结果；S35: Find the truncation matrix R λ = λ (i,j) : where λ is used as a metric value to represent the truncation matrix coefficient, the fuzzy relationship matrix between modules is truncated, and the elements in the fuzzy correlation matrix R greater than or equal to λ, Its value is 1, and the value of elements less than λ is 0. Elements with a value of 1 in the same row or column are aggregated into the same module, and the remaining elements become modules separately. The larger the λ value, the more detailed the module division; by selecting With different λ values, different process module clustering results can be obtained; S36：借助Matlab工具并运用模糊聚类分析方法，依据原始关联矩阵可依次计算出标准化矩阵、模糊相似矩阵、传递闭包矩阵，最后利用λ-截矩阵对模糊关联矩阵R进行切割，形成模块聚类图：
Figure FDA0003369678330000024
S36: With the help of Matlab tools and fuzzy clustering analysis method, the normalization matrix, fuzzy similarity matrix, and transitive closure matrix can be calculated in turn according to the original correlation matrix, and finally the fuzzy correlation matrix R is cut by the λ-intersection matrix to form a modular cluster Class Diagram:
Figure FDA0003369678330000024
 S4：根据不同分区阈值序列形成整体的模块聚类图，通过选取不同的λ值，得到不同的模块划分方案。S4: According to different partition threshold sequences, an overall module clustering graph is formed, and different module division schemes are obtained by selecting different λ values. 2.根据权利要求1所述的一种面向大规模定制的复杂工艺模块划分方法，其特征在于：所述步骤S2中，模糊关联矩阵的建立方法，具体包括以下步骤：2. a kind of complex process module dividing method for mass customization according to claim 1, is characterized in that: in described step S2, the establishment method of fuzzy correlation matrix specifically comprises the following steps: A1：以输出参数的属性值域作为评判各个加工工序之间的关联性，将关联性描述为可替换相关性和影响度相关性，用数学模型进行表示模糊关联矩阵R：ri,j＝ωffi,j+ωppi,j，其中，ri,j表示模块i与模块j之间的总相关系数，fi,j 表示模块 i 与模块 j 之间的相关系数， pi,j 表示模块 j 与模块 i 之间的相关系数 ，ωf、ωp表示特性所对应的权重系数；A1: Take the attribute value range of the output parameter as the judgment of the correlation between the various processing procedures, describe the correlation as the replaceable correlation and the influence correlation, and use the mathematical model to represent the fuzzy correlation matrix R: r i,j = ω f f i,j +ω p p i,j , where ri ,j represents the total correlation coefficient between module i and module j, f i,j represents the correlation coefficient between module i and module j, p i,j represent the correlation coefficient between module j and module i, and ω f and ω p represent the weight coefficients corresponding to the characteristics; A2：模糊关联矩阵R也称原始矩阵R，且R满足自反性：0≤r(i,j)≤1，r(i,i)＝r(j,j)＝1；且R满足对称性：r(i,j)＝r(j,i)，则步骤A1中模糊关联矩阵R的数学模型表示为：A2: The fuzzy correlation matrix R is also called the original matrix R, and R satisfies reflexivity: 0≤r(i,j)≤1, r(i,i)=r(j,j)=1; and R satisfies symmetry property: r(i,j)=r(j,i), then the mathematical model of the fuzzy correlation matrix R in step A1 is expressed as:
Figure FDA0003369678330000031
Figure FDA0003369678330000031
 3.根据权利要求2所述的一种面向大规模定制的复杂工艺模块划分方法，其特征在于：所述步骤A1中，ri,j表示模块i与模块j之间的总相关系数，fi,j表示模块i与模块j之间的相关系数，pi,j表示模块j与模块i之间的相关系数，ωf、ωp表示特性所对应的权重系数；ωf、ωp∈[0，1]，且满足：ωf+ωp＝1。3. a kind of complex process module dividing method for mass customization according to claim 2, is characterized in that: in described step A1, r i, j represent the total correlation coefficient between module i and module j, f i,j represents the correlation coefficient between module i and module j, p i,j represents the correlation coefficient between module j and module i, ω f , ω p represent the weight coefficients corresponding to the characteristics; ω f , ω p ∈ [0, 1], and satisfy: ω f +ω p =1. 4.根据权利要求2所述的一种面向大规模定制的复杂工艺模块划分方法，其特征在于：所述步骤A1中，所述输出参数为晶粒度、屈服强度、延伸率、残余应力、耐蚀性、电导率、硬度、抗拉强度。4. The method for dividing complex process modules for mass customization according to claim 2, wherein in the step A1, the output parameters are grain size, yield strength, elongation, residual stress, Corrosion resistance, electrical conductivity, hardness, tensile strength. 5.根据权利要求1所述的一种面向大规模定制的复杂工艺模块划分方法，其特征在于：所述步骤S32中，数据标准化处理运用到的标准化公式为：
Figure FDA0003369678330000032
5. a kind of complex process module division method for mass customization according to claim 1, is characterized in that: in described step S32, the standardized formula that data standardization processing is applied to is:
Figure FDA0003369678330000032
 其中，i＝1,2,...,m，x′(i,j)代表原始数据；x'(i,j)man、x'(i,j)min分别表示原始数据的最大值和最小值；x(i,j)为标准化后得到的数据。Among them, i=1,2,...,m, x' (i,j) represents the original data; x' (i,j)man and x' (i,j)min respectively represent the maximum value and Minimum value; x (i,j) is the data obtained after normalization. 6.根据权利要求1所述的一种面向大规模定制的复杂工艺模块划分方法，其特征在于：所述步骤S34中，求取传递闭包矩阵
Figure FDA0003369678330000041
具体包括以下步骤：6. a kind of complex process module division method for mass customization according to claim 1, is characterized in that: in described step S34, obtain the transitive closure matrix
Figure FDA0003369678330000041
Specifically include the following steps: A1：设置R0为R，其中，R0表示模块原始关联矩阵；A1: Set R 0 to R, where R 0 represents the original correlation matrix of the module; A2：假设Ri已得知结果，开始比较Ri和
Figure FDA0003369678330000042
所对应的元素；如果
Figure FDA0003369678330000043
则Ri是
Figure FDA0003369678330000044
的传递闭包矩阵，此时可停止计算；如果不满足
Figure FDA0003369678330000045
则继续执行下一步；A2: Assuming that Ri has known the result, start comparing Ri and
Figure FDA0003369678330000042
the corresponding element; if
Figure FDA0003369678330000043
Then R i is
Figure FDA0003369678330000044
The transitive closure matrix of , the calculation can be stopped at this time; if it does not satisfy
Figure FDA0003369678330000045
then proceed to the next step; A3：通过算法计算
Figure FDA0003369678330000046
并将结果设定为Ri+1，执行步骤A2；A3: Calculated by algorithm
Figure FDA0003369678330000046
and set the result as R i+1 , and execute step A2; A4：重复上述步骤，直到最终求出传递闭包矩阵，使
Figure FDA0003369678330000047
A4: Repeat the above steps until the transitive closure matrix is finally obtained, so that
Figure FDA0003369678330000047
CN201910555601.1A 2019-06-25 2019-06-25 A Mass Customization-Oriented Partitioning Method for Complex Process Modules Active CN110348954B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910555601.1A CN110348954B (en) 2019-06-25 2019-06-25 A Mass Customization-Oriented Partitioning Method for Complex Process Modules

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910555601.1A CN110348954B (en) 2019-06-25 2019-06-25 A Mass Customization-Oriented Partitioning Method for Complex Process Modules

Publications (2)

Publication Number Publication Date
CN110348954A CN110348954A (en) 2019-10-18
CN110348954B true CN110348954B (en) 2022-02-25

Family

ID=68183029

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910555601.1A Active CN110348954B (en) 2019-06-25 2019-06-25 A Mass Customization-Oriented Partitioning Method for Complex Process Modules

Country Status (1)

Country Link
CN (1) CN110348954B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110929949A (en) * 2019-11-30 2020-03-27 温州大学 Method for obtaining optimal module assembly scheme on garment production line
CN111583055A (en) * 2020-05-09 2020-08-25 电子科技大学 A Product Grouping Method Based on Genetic Algorithm under Multiple Process Paths
CN111915153A (en) * 2020-07-11 2020-11-10 天津大学 Method for dividing reconfigurable manufacturing system workpiece family by considering multiple indexes
CN117676562B (en) * 2024-01-31 2024-05-10 四川省机场集团有限公司成都天府国际机场分公司 Data safety communication method

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101315644A (en) * 2008-05-09 2008-12-03 浙江工业大学 Part Classification Method Based on Extension Clustering
CN103198138A (en) * 2013-04-16 2013-07-10 北京科技大学 Large-scale hot continuous rolling data scheme customizing system based on cloud computing
CN103605843A (en) * 2013-11-13 2014-02-26 西安工业大学 Complex production assembling ability evaluation system and method based on DELMIA
CN104077432A (en) * 2014-05-21 2014-10-01 浙江工业大学 Process-adjustment choosing analysis method based on multidimensional correlation function
CN108536975A (en) * 2018-04-16 2018-09-14 宁夏汇川服装有限公司 A kind of template standardized system based on mass customization production

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4272494B2 (en) * 2003-08-20 2009-06-03 パナソニック株式会社 Manufacturing process development method

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101315644A (en) * 2008-05-09 2008-12-03 浙江工业大学 Part Classification Method Based on Extension Clustering
CN103198138A (en) * 2013-04-16 2013-07-10 北京科技大学 Large-scale hot continuous rolling data scheme customizing system based on cloud computing
CN103605843A (en) * 2013-11-13 2014-02-26 西安工业大学 Complex production assembling ability evaluation system and method based on DELMIA
CN104077432A (en) * 2014-05-21 2014-10-01 浙江工业大学 Process-adjustment choosing analysis method based on multidimensional correlation function
CN108536975A (en) * 2018-04-16 2018-09-14 宁夏汇川服装有限公司 A kind of template standardized system based on mass customization production

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Dynamic supply chain integration optimization in service mass customization;ChangLiu 等;《science direct》;20180630;42-52 *
面向大规模定制的复杂产品模块规划方法研究;李军鹏;《中国博士学位论文全文数据库》;20130315;I138-43 *
面向大规模定制的模块化产品族设计;詹跃跃;《万方数据库》;20140122;全文 *

Also Published As

Publication number Publication date
CN110348954A (en) 2019-10-18

Similar Documents

Publication Publication Date Title
CN110348954B (en) A Mass Customization-Oriented Partitioning Method for Complex Process Modules
CN114358185B (en) Multi-dimensional short-term power load prediction method based on improved K-means clustering CCA-BiLSTM
CN107967542B (en) A prediction method of electricity sales based on long short-term memory network
CN114239733B (en) Machine tool response modeling method, system and response prediction method based on transfer learning
CN110750931B (en) Efficient profile extrusion die design method
CN111539845B (en) A method for judging and judging enterprise environmental protection management and control response based on the subordination score of electricity consumption mode
CN109308063B (en) Stage division method of fermentation process based on score matrix
CN103810306A (en) Efficient design method of profile extrusion mould
CN110404965A (en) Method and model system for hot-rolled medium-thick plate assembly and slab design considering the flexibility of non-cut order specifications
JP2004086897A (en) Model building method and model building system
CN101808339A (en) Telephone traffic subdistrict self-adaptive classification method applying K-MEANS and prior knowledge
CN104050547A (en) Non-linear optimization decision-making method of planning schemes for oilfield development
CN113627685A (en) Wind power generator power prediction method considering wind power on-grid load limitation
CN113221282A (en) Aero-engine service life prediction method based on Bayesian residual convolutional network
CN104077432A (en) Process-adjustment choosing analysis method based on multidimensional correlation function
Wu et al. Optimal shape design of an extrusion die using polynomial networks and genetic algorithms
Zirngibl et al. APPROACH FOR THE AUTOMATED AND DATA-BASED DESIGN OF MECHANICAL JOINTS.
CN110334416B (en) Optimal design method of preform in forging of dual performance disc
CN109117597B (en) Processing key factor grading method based on corresponding analysis method
CN111027612B (en) Energy metering data feature reduction method and device based on weighted entropy FCM
CN106202686A (en) A kind of objective design method of turbine disk isothermal die forging process preform blank
CN114863169A (en) An Image Classification Method Combining Parallel Ensemble Learning and Federated Learning
CN105590141A (en) Genetic algorithm initial population construction method applied to optimized design of complex products
CN113359435A (en) Correction method for dynamic working condition data of thermal power generating unit
CN113762754A (en) A low-entropy adaptive scheduling method for hybrid workshops

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