CN112836293A - Automobile product design scheme selection method based on PSO information granulation - Google Patents

Automobile product design scheme selection method based on PSO information granulation Download PDF

Info

Publication number
CN112836293A
CN112836293A CN202110063426.1A CN202110063426A CN112836293A CN 112836293 A CN112836293 A CN 112836293A CN 202110063426 A CN202110063426 A CN 202110063426A CN 112836293 A CN112836293 A CN 112836293A
Authority
CN
China
Prior art keywords
evaluation
language
language term
term
decision maker
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.)
Granted
Application number
CN202110063426.1A
Other languages
Chinese (zh)
Other versions
CN112836293B (en
Inventor
张强
吴齐
唐孝安
赵爽耀
黄挺
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.)
Hefei University of Technology
Original Assignee
Hefei University of 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 Hefei University of Technology filed Critical Hefei University of Technology
Priority to CN202110063426.1A priority Critical patent/CN112836293B/en
Publication of CN112836293A publication Critical patent/CN112836293A/en
Application granted granted Critical
Publication of CN112836293B publication Critical patent/CN112836293B/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
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/15—Vehicle, aircraft or watercraft design

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Geometry (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Machine Translation (AREA)

Abstract

The invention relates to the technical field of strategy selection, in particular to a method for selecting an automobile product design scheme based on PSO information granulation. The method comprises the following steps: step S1, establishing an alternative solution set X and a solution evaluation language term set L1(ii) a Step S2, establishing a decision maker set E and a decision maker evaluation language term set L2(ii) a Step S3, establishing a language term evaluation matrix D of each decision maker for all alternatives1w(ii) a Step S4, establishing a language term evaluation matrix D for the decision maker2(ii) a Step S5, granulating the language terms to obtain a language term evaluation matrix D1wNumerical evaluation matrix ND of1wAnd a language term evaluation matrix D2Numerical evaluation matrix ND of2(ii) a Step S6, evaluating the matrix ND according to the numeralization1wAnd a numerical evaluation matrix ND2And acquiring the preference degree of each alternative scheme so as to acquire the optimal design scheme. The invention can better convert the preference information of language glossing into the numerical value processing problemAnd further an optimal scheme is obtained.

Description

Automobile product design scheme selection method based on PSO information granulation
Technical Field
The invention relates to the technical field of strategy selection, in particular to a method for selecting an automobile product design scheme based on PSO information granulation.
Background
The automobile industry is one of the important post industries of national economy, and with the development of automobile products becoming more and more rapid, the life style of people is being changed slowly. As a new industry, the demand of automobiles is increasing, and the industrial revolution taking information technology as the leading factor also brings new development opportunities to the automobile industry. The traditional automobile and the new generation information technology are continuously fused, the continuous change of the consumer demand is adapted, and the mode of providing better service for users becomes a new mode of the development of the current automobile products. On one hand, the development process of automobile products is very tedious, and not only needs to know the requirements of users, but also needs to pay attention to quality management. On the other hand, with the development of the internet and big data, technologies such as artificial intelligence and voice recognition provide more ideas and suggestions for the design of automobile products. In order to develop the intelligent networked automobile which is popular with consumers, the automobile enterprises need to consider opinions from various departments, such as a research and development department, a production department, a financial department, a market department and the like, when making research and development decisions, so as to select an automobile product design scheme which best meets the requirements of users and the strategic planning of enterprises.
Because opinions from various departments contain a large number of rather ambiguous language terms during the design of automotive products, it is difficult to make a systematic analysis and decision.
Disclosure of Invention
The present invention provides a method for automotive product design selection based on PSO information pelletization that overcomes some or all of the deficiencies of the prior art.
The invention relates to a method for selecting an automobile product design scheme based on PSO information granulation, which comprises the following steps:
step S1Establishing a candidate set X and a scheme evaluation language term set L for evaluating preference degree of the candidate1(ii) a Wherein X ═ { X ═ Xt|t＝1,2,3,…,i,…,j,…,n}，xtRepresents the tth alternative; wherein L is1＝{lh|h＝-p,-p+1,…,0,…,p-1,p}，lhThe h language term, language term l, representing the evaluation of the scheme-pTo the linguistic term lpThe indicated preference degree is increased;
step S2, establishing a decision maker set E and a decision maker evaluation language term set L for evaluating the preference degree of the decision maker2(ii) a Wherein E ═ { E ═ Ew|w＝1,2,3,…,s…,g,…,m}，ewRepresents the w-th decision maker; wherein L is2＝{lr|r＝-q,-q+1,…,0,…,q-1,q}，lrThe r-th language term, language term l, representing the evaluation of a decision maker-qTo the linguistic term lqThe indicated preference degree is increased;
step S3, each decision maker evaluates the language term set L based on the scheme1The language term in (1) carries out language term evaluation on all the alternatives, and a language term evaluation matrix D of each decision maker on all the alternatives is established1w，D1wRepresenting a matrix obtained after the w-th decision maker evaluates all the alternative schemes; wherein the content of the first and second substances,
Figure BDA0002903224660000021
Figure BDA0002903224660000022
indicating the w-th decision maker compared to alternative xjFor alternative xi(ii) a linguistic term evaluation of (a); wherein the content of the first and second substances,
Figure BDA0002903224660000023
Figure BDA0002903224660000024
representing linguistic term evaluation
Figure BDA0002903224660000025
Chinese language term lhThe probability of occupation; wherein the content of the first and second substances,
Figure BDA0002903224660000026
and is
Figure BDA0002903224660000027
Step S4, evaluating language term set L based on decision maker2The language term in (1) carries out language term evaluation on all decision-makers, and establishes a language term evaluation matrix D for the decision-makers2(ii) a Wherein the content of the first and second substances,
Figure BDA0002903224660000028
Figure BDA0002903224660000029
representing the comparison to decision egFor decision maker es(ii) a linguistic term evaluation of (a); wherein the content of the first and second substances,
Figure BDA00029032246600000210
Figure BDA00029032246600000211
representing linguistic term evaluation
Figure BDA00029032246600000212
Chinese language term lrThe probability of occupation; wherein the content of the first and second substances,
Figure BDA00029032246600000213
and is
Figure BDA00029032246600000214
Step S5, granulating the language terms to obtain a language term evaluation matrix D1wNumerical evaluation matrix ND of1w，
Figure BDA00029032246600000215
Figure BDA00029032246600000216
Indicating the w-th decision maker compared to alternative xjFor alternative xiNumerical evaluation of (1); and a language term evaluation matrix D2Numerical evaluation matrix ND of2，
Figure BDA00029032246600000217
Figure BDA00029032246600000218
Representing the comparison to decision egFor decision maker esNumerical evaluation of (1);
step S6, evaluating the matrix ND according to the numeralization1wAnd a numerical evaluation matrix ND2And acquiring the preference degree of each alternative scheme so as to acquire the optimal design scheme.
By the method, the opinions of enterprise managers and decision makers of all departments can be comprehensively considered, and a decision model is constructed to determine the optimal automobile product design scheme aiming at a plurality of feasible automobile product design schemes, so that the product is closer to the real requirement of a user, and better service is provided for the user. Meanwhile, in the decision making process, considering the ambiguity of different decision makers when using language terms to express preference information, a PSO information granulation method taking consistency and consensus degree as indexes is provided, the language value is converted into a calculable interval particle form, and finally the opinion of each department is comprehensively considered to select the most appropriate scheme.
Preferably, step S5 includes a scenario evaluation language term set L1And the decision maker evaluates the language term set L2In the form of message granules, each language term is expressed as an interval [0, 1%]A subset of (a). Therefore, the language information can be preferably converted into numerical information, and the subsequent operation is further facilitated.
Preferably, in step S5, the language term set L is evaluated for the decider2When granulation is carried out, the method comprises the following steps:
step S511, setting language termslrCorresponding subinterval is [ b ]r-1,br]And set b-q-1＝0，bq＝1；
Step S512, calculating intercept point vector [ b ] based on PSO algorithm-q,b-q+1,…,b0,…bq-2,bq-1]And the decision maker evaluates the language term set L2Set of values W ofL2，WL2＝{WLr-q +1, …,0, …, q-1, q }; wherein, WLrDenotes the language term lrBy the term l in the languagerContinuously and uniformly sampling and obtaining in the corresponding subinterval;
in this step, the matrix ND is evaluated numerically2Index Q of multiplicative consistency2As an optimization index and by a multiplicative consistency index Q2Mean value of
Figure BDA0002903224660000031
Establishing an optimization model as a fitness function, wherein
Figure BDA0002903224660000032
Figure BDA0002903224660000033
Figure BDA0002903224660000034
Representing a numerical evaluation matrix ND established based on values obtained by continuous uniformly distributed sampling in the kth operation2A multiplicative consistency indicator of;
the optimization model is established as follows,
MaxQ2＝cND2
Figure BDA0002903224660000035
wherein, "cND2"indicates the numerical evaluation matrix ND obtained for each operation2A multiplicative consistency indicator of; constructed fitness function f2In order to realize the purpose,
Figure BDA0002903224660000041
step S513, constraint condition based on optimization model and fitness function f2Get the intercept point vector [ b ]-q,b-q+1,…,b0,…bq-2,bq-1]And a numerical evaluation matrix ND2(ii) a Wherein the content of the first and second substances,
Figure BDA0002903224660000042
through steps S511 to S513, the linguistic term matrix is preferably digitized, and the weight of each decision maker can be preferably obtained.
Preferably, the method further comprises calculating a set of decision maker values W, { W ═ Ww|w＝1,2,3,…,s…,g,…,m}，WwA weight representing the w-th decision maker; wwThe calculation formula is as follows,
Figure BDA0002903224660000043
so that the weight for each decision maker can be preferably obtained.
Preferably, in step S5, the language term set L is evaluated for the scheme1When granulation is carried out, the method comprises the following steps:
step S521, setting language term lhCorresponding subinterval is [ a ]h-1,ah]And set a-p-1＝0，ap＝1；
Step S522, calculating intercept point vector [ a ] based on PSO algorithm-p,a-p+1,…,a0,…ap-2,ap-1]And a scheme evaluation language term set L1Set of values W ofL1，WL1＝WLh-p, -p +1, …,0, …, p-1, p }; wherein, WLhDenotes the language term lhBy the term l in the languagehCorresponding subintervals are all continuousUniformly sampling and obtaining;
in the step of the method,
the optimization model is established as follows,
Figure BDA0002903224660000044
Figure BDA0002903224660000051
constructed fitness function f1In order to realize the purpose,
Figure BDA0002903224660000052
wherein the content of the first and second substances,
Figure BDA0002903224660000053
cND1wrepresents a numerical evaluation matrix ND1wThe consistency index of (a); o is2Is corresponding to O1The index of the consistency ratio of (1), delta is a balance factor, and delta belongs to [0,1]]；
Step S523, constraint condition based on optimization model and fitness function f1Obtaining an intercept point vector [ a ]-p,a-p+1,…,a0,…ap-2,ap-1]And a numerical evaluation matrix ND1w(ii) a Wherein the content of the first and second substances,
Figure BDA0002903224660000054
through steps S521-S523, the numeralization of the case language terms is preferably achieved.
Preferably, in step S6, a total evaluation matrix ND is acquiredc，
Figure BDA0002903224660000055
Wherein the content of the first and second substances,
Figure BDA0002903224660000056
thereby can preferably followThe weight of each decision maker integrates all the evaluation matrices.
Preferably, in step S6, the total evaluation matrix ND is subjected to averaging based on an averaging operatorcCalculating to obtain a scheme evaluation matrix X ', X ═ X't)1×n，
Figure BDA0002903224660000057
So that the evaluation value of each scheme can be preferably acquired.
Preferably, in step S6, the optimal solution is obtained by sorting based on the solution evaluation matrix X'. So that the optimum scheme can be preferably obtained.
Drawings
FIG. 1 is a schematic view of the method in example 1;
FIG. 2 is a fitness function f in example 12A graph of the iterative change of the numerical values of (1);
FIG. 3 is a fitness function f in example 11The value of (2) is iteratively changed.
Detailed Description
For a further understanding of the invention, reference should be made to the following detailed description taken in conjunction with the accompanying drawings and examples. It is to be understood that the examples are illustrative of the invention and not limiting.
Example 1
Referring to fig. 1, the present embodiment provides a method for selecting an automotive product design scheme based on PSO information granulation, which includes the following steps:
step S1, establishing a candidate set X and a scheme evaluation language term set L for evaluating the preference degree of the candidate1(ii) a Wherein X ═ { X ═ Xt|t＝1,2,3,…,i,…,j,…,n}，xtRepresents the tth alternative; wherein L is1＝{lh|h＝-p,-p+1,…,0,…,p-1,p}，lhThe h language term, language term l, representing the evaluation of the scheme-pTo the linguistic term lpThe indicated preference degree is increased;
step S2, establishing a decision maker set E and evaluating the bias to the decision makerGood degree decision maker evaluates language term set L2(ii) a Wherein E ═ { E ═ Ew|w＝1,2,3,…,s…,g,…,m}，ewRepresents the w-th decision maker; wherein L is2＝{lr|r＝-q,-q+1,…,0,…,q-1,q}，lrThe r-th language term, language term l, representing the evaluation of a decision maker-qTo the linguistic term lqThe indicated preference degree is increased;
step S3, each decision maker evaluates the language term set L based on the scheme1The language term in (1) carries out language term evaluation on all the alternatives, and a language term evaluation matrix D of each decision maker on all the alternatives is established1w，D1wRepresenting a matrix obtained after the w-th decision maker evaluates all the alternative schemes; wherein the content of the first and second substances,
Figure BDA0002903224660000061
Figure BDA0002903224660000062
indicating the w-th decision maker compared to alternative xjFor alternative xi(ii) a linguistic term evaluation of (a); wherein the content of the first and second substances,
Figure BDA0002903224660000063
Figure BDA0002903224660000064
representing linguistic term evaluation
Figure BDA0002903224660000065
Chinese language term lhThe probability of occupation; wherein the content of the first and second substances,
Figure BDA0002903224660000066
and is
Figure BDA0002903224660000067
Step S4, evaluating language term set L based on decision maker2The language term in (1) evaluates the language term of all decision makers and establishes the language of the decision makersTerm evaluation matrix D2(ii) a Wherein the content of the first and second substances,
Figure BDA0002903224660000068
Figure BDA0002903224660000069
representing the comparison to decision egFor decision maker es(ii) a linguistic term evaluation of (a); wherein the content of the first and second substances,
Figure BDA00029032246600000610
Figure BDA00029032246600000611
representing linguistic term evaluation
Figure BDA00029032246600000612
Chinese language term lrThe probability of occupation; wherein the content of the first and second substances,
Figure BDA00029032246600000613
and is
Figure BDA00029032246600000614
Step S5, granulating the language terms to obtain a language term evaluation matrix D1wNumerical evaluation matrix ND of1w，
Figure BDA0002903224660000071
Figure BDA0002903224660000072
Indicating the w-th decision maker compared to alternative xjFor alternative xiNumerical evaluation of (1); and a language term evaluation matrix D2Numerical evaluation matrix ND of2，
Figure BDA0002903224660000073
Figure BDA0002903224660000074
Representing a comparison to a decisionE isgFor decision maker esNumerical evaluation of (1);
step S6, evaluating the matrix ND according to the numeralization1wAnd a numerical evaluation matrix ND2And acquiring the preference degree of each alternative scheme so as to acquire the optimal design scheme.
By the method in the embodiment, the opinions of enterprise managers and decision makers of all departments can be comprehensively considered, and a decision model is constructed to determine the optimal automobile product design scheme aiming at a plurality of feasible automobile product design schemes, so that the product is closer to the real requirement of a user, and better service is provided for the user. Meanwhile, in the decision making process, considering the ambiguity of different decision makers when using language terms to express preference information, a PSO information granulation method taking consistency and consensus degree as indexes is provided, the language value is converted into a calculable interval particle form, and finally the opinion of each department is comprehensively considered to select the most appropriate scheme.
In the invention, the language term set is a limited language expression set with the number of elements not being 1, and the PSO algorithm is a particle swarm algorithm.
This embodiment is an embodiment of the present invention, and the present invention is explained with a specific application.
In this embodiment, the alternative set X includes 4 elements, each of which is X1、x2、x3And x4I.e., n ═ 4; x is the number of1、x2、x3And x4The different vehicle designs in fig. 4 are shown.
In this example, the scheme evaluation language term set L1The 5 elements involved are "Very Low (VL)", "low (L)", "same (M)", "high (H)" and "Very High (VH)", respectively, i.e. p ═ 2. "Very Low (VL)", "low (L)", "same (M)", "high (H)" and "Very High (VH)" are feasibility assessment terms that may be used when comparing two different protocols.
In this embodiment, the decision maker set E comprises 4 elements E1、e2、e3And e4I.e., m is 4; e.g. of the type1、e2、e3And e4Representing different departments of a company, such as the "product department", "research and development department", "finance department", and "marketing department".
In this example, L2The inclusion of 7 elements is "very unimportant (WI)", "Less Important (LI)", "unimportant (NI)", "Equally Important (EI)", "important (I)", "more important (VI)" "very important (SI)", i.e. q ═ 3, respectively; "very important (WI)", "Less Important (LI)", "Not Important (NI)", "Equally Important (EI)", "important (I)", "more important (VI)", "very important (SI)" are importance evaluation terms that are available when comparing opinions of different departments.
As shown in tables 1-4, the results of all the evaluation of the solutions for different decision makers, namely D1w(ii) a As shown in Table 5, the results of the evaluation of the importance of the company managers to the different departments, namely D, were obtained2。
TABLE 1 department of products (e)1) Language term evaluation matrix D obtained after all schemes are evaluated1w
Figure BDA0002903224660000081
TABLE 2 research and development part (e)2) Language term evaluation matrix D obtained after all schemes are evaluated12
Figure BDA0002903224660000082
TABLE 3 finance department (e)3) Language term evaluation matrix D obtained after all schemes are evaluated13
Figure BDA0002903224660000083
TABLE 4 market sector (e)4) Linguistics acquired after evaluation of all solutionsLanguage evaluation matrix D14
Figure BDA0002903224660000091
TABLE 5 language term evaluation matrix D obtained after evaluation of all departments by the company manager2
Figure BDA0002903224660000092
As can be seen from tables 1 to 5, it is difficult to perform a systematic and effective evaluation on the language-glossed evaluation matrix during the selection of the actual automobile product design. This makes it difficult to efficiently process the language decision information of each department during the design of the actual product, i.e., to select the optimal evaluation from the language-glossed evaluation decisions.
By the method in the embodiment, the decision of each department and the importance of each department opinion can be comprehensively considered, the fuzzified evaluation of language terminology can be converted into a numerical matrix through a PSO algorithm, and an optimal scheme can be further obtained through a series of processing on numerical values.
In this embodiment, the language term set L is evaluated by the distributed language preference relationship construction scheme1And the decision maker evaluates the language term set L2Therefore, the information in the language glossing evaluation can be retained to the maximum extent.
In this embodiment, step S5 includes a scheme evaluation language term set L1And the decision maker evaluates the language term set L2In the form of message granules, each language term is expressed as an interval [0, 1%]A subset of (a). Therefore, the language information can be preferably converted into numerical information, and the subsequent operation is further facilitated.
In the embodiment, in consideration of the fact that the language information itself is not operable, in order to obtain an intuitive result, the language terms are firstly converted into numerical values, and an information granulation process is introduced, so that the data processing can be preferably facilitated by mapping the language terms into the form of information particles convenient for calculation. In this embodiment, the interval can be selected as the form of the information particle, that is, each language term corresponds to a sub-interval of [0,1] interval; by this, the processing of the language term sets L1 and L2 can be converted into the processing of the upper and lower limits of the interval for each language term, that is, the problem of the granulation of the language terms can be converted into the problem of the allocation of the intercept points of the interval, thereby preferably facilitating the data processing.
In this embodiment, in step S5, the language term set L is evaluated for the decider2When granulation is carried out, the method comprises the following steps:
step S511, setting language term lrCorresponding subinterval is [ b ]r-1,br]And set b-q-1＝0，bq＝1；
Step S512, calculating intercept point vector [ b ] based on PSO algorithm-q,b-q+1,…,b0,…bq-2,bq-1]And the decision maker evaluates the language term set L2Set of values W ofL2，WL2＝{WLr-q +1, …,0, …, q-1, q }; wherein, WLrDenotes the language term lrBy the term l in the languagerContinuously and uniformly sampling and obtaining in the corresponding subinterval;
in this step, the matrix ND is evaluated numerically2Index Q of multiplicative consistency2As an optimization index and by a multiplicative consistency index Q2Mean value of
Figure BDA0002903224660000101
Establishing an optimization model as a fitness function, wherein
Figure BDA0002903224660000102
Figure BDA0002903224660000103
Figure BDA0002903224660000104
Representing a numerical evaluation matrix ND established based on values obtained by continuous uniformly distributed sampling in the kth operation2A multiplicative consistency indicator of;
the optimization model is established as follows,
MaxQ2＝cND2
Figure BDA0002903224660000105
wherein, "cND2"indicates the numerical evaluation matrix ND obtained for each operation2A multiplicative consistency indicator of;
constructed fitness function f2In order to realize the purpose,
Figure BDA0002903224660000111
step S513, constraint condition based on optimization model and fitness function f2Get the intercept point vector [ b ]-q,b-q+1,…,b0,…bq-2,bq-1]And a numerical evaluation matrix ND2Wherein; wherein the content of the first and second substances,
Figure BDA0002903224660000112
through steps S511 to S513, the linguistic term matrix is preferably digitized, and the weight of each decision maker can be preferably obtained.
In this embodiment, the distribution problem of the interval intercept points is actually an optimization process, and the numeralization can be preferably realized by constructing an optimization model and solving the optimization model. The model established in the embodiment can better realize the solution of the model by optimizing the consistency index.
In this embodiment, the parameters of the PSO algorithm can be set as: the number of particles is 100, the maximum number of iterations is 500, the inertia factor is 0.7, the learning factors c1 and c2 are both 2, and the sampling number K is 500.
Through solving the optimization model, the fitness function f in the embodiment can be obtained2Has an optimal value of 0.985 and a fitness function f2The graph of the iterative change of the numerical values of (a) is shown in fig. 2.
In this embodiment, the language term set L is obtained2Has a intercept point vector of [0.15,0.28,0.34,0.50,0.58, 0.84%]I.e. a set of language terms L2Of (1), the interval corresponding to the very important (WI) is [0,0.15 ], the interval corresponding to the Less Important (LI) is [0.15,0.28 ], the interval corresponding to the unimportant (NI) is [0.28,0.34 ], the interval corresponding to the Equally Important (EI) is [0.34,0.50 ], the interval corresponding to the important (I) is [0.50,0.58 ], the interval corresponding to the more important (VI) is [0.58,0.84 ], the interval corresponding to the very important (SI) is [0.84,1]。
At the same time, under the optimal solution, the evaluation matrix ND is quantified2The following were used:
Figure BDA0002903224660000113
in this embodiment, the numerical evaluation matrix ND is acquired2Then, the decision value set W can be calculated preferably, where W is { W ═ Ww|w＝1,2,3,…,s…,g,…,m}，WwA weight representing the w-th decision maker; wwThe calculation formula is as follows,
Figure BDA0002903224660000121
in this embodiment, the obtained decision maker value set W is: w ═ 0.209,0.359,0.276,0.155 }.
In step S5 of the present embodiment, the language term set L is evaluated for the pattern1When granulation is carried out, the method comprises the following steps:
step S521, setting language term lhCorresponding subinterval is [ a ]h-1,ah]And set a-p-1＝0，ap＝1；
Step S522 based on PSO algorithmFinding the intercept point vector [ a-p,a-p+1,…,a0,…ap-2,ap-1]And a scheme evaluation language term set L1Set of values W ofL1，WL1＝WLh-p, -p +1, …,0, …, p-1, p }; wherein, WLhDenotes the language term lhBy the term l in the languagehContinuously and uniformly sampling and obtaining in the corresponding subinterval;
in the step of the method,
the optimization model is established as follows,
Figure BDA0002903224660000122
Figure BDA0002903224660000123
constructed fitness function f1In order to realize the purpose,
Figure BDA0002903224660000124
wherein the content of the first and second substances,
Figure BDA0002903224660000125
cND1wrepresents a numerical evaluation matrix ND1wThe consistency index of (a); o is2Is corresponding to O1The index of the consistency ratio of (1), delta is a balance factor, and delta belongs to [0,1]]；
Step S523, constraint condition based on optimization model and fitness function f1Obtaining an intercept point vector [ a ]-p,a-p+1,…,a0,…ap-2,ap-1]And a numerical evaluation matrix ND1w(ii) a Wherein the content of the first and second substances,
Figure BDA0002903224660000131
through steps S521-S523, the numeralization of the case language terms is preferably achieved.
In this embodiment, the parameter δ is a parameter that balances the overall degree of identity and consistency, and when δ is smaller, it indicates that the optimization result is more concerned with the overall degree of identity; and conversely, the optimization result is more concerned about the consistency of the decision-making individuals. In the present embodiment, δ is set to 0.25, so that a higher degree of consensus can be preferably achieved.
Through solving the optimization model, the fitness function f in the embodiment can be obtained1Has an optimal value of 0.922 and a fitness function f1The graph of the iterative change of the numerical values of (a) is shown in fig. 3.
In this embodiment, the language term set L is obtained1Has an intercept point vector of [0.39,0.58,0.60,0.61 ]]I.e. the set of language terms L1Of the ranges, Very Low (VL) corresponds to [0,0.39 ], low (L) corresponds to [0.39,0.58 ], the same (M) corresponds to [0.58,0.60 ], high (H) corresponds to [0.60,0.61), and Very High (VH) corresponds to [0.61,1]。
At the same time, under the optimal solution, the evaluation matrix ND is quantified11To ND12The following were used:
Figure BDA0002903224660000132
Figure BDA0002903224660000133
Figure BDA0002903224660000134
Figure BDA0002903224660000135
in step S6 of the present embodiment, a total evaluation matrix ND is acquiredc，
Figure BDA0002903224660000141
Wherein the content of the first and second substances,
Figure BDA0002903224660000142
so that all evaluation matrices will preferably be integrated according to the weight of each decision maker.
In this embodiment, the obtained total evaluation matrix NDcComprises the following steps:
Figure BDA0002903224660000143
in this embodiment, in step S6, the total evaluation matrix ND is determined based on the average operatorcCalculating to obtain a scheme evaluation matrix X ', X ═ X't)1×n，
Figure BDA0002903224660000144
So that the evaluation value of each scheme can be preferably acquired.
In this embodiment, the obtained scheme evaluation matrix X' is {0.589,0.603,0.433,0.376 }.
In this embodiment, in step S6, the optimal solution may be obtained by sorting based on the solution evaluation matrix X'. So that the optimum scheme can be preferably obtained.
In this example, the scheme evaluation matrix X' is paired with the scheme X1、x2、x3And x4Sorting may be performed to obtain, x2>x1>x3>x4. Therefore, should choose scheme x2As a left-right design scheme.
By the method in the embodiment, when the alternative automobile design scheme is aimed at, the opinions of all departments of a company can be comprehensively considered, and a scheme which is closest to the market demand and can be accepted by consumers most can be selected.
In consideration of the self-limited rationality of each department decision maker and the uncertainty of selecting the design scheme of the automobile product, the preference relationship of each department decision maker is expressed based on the distributed language preference relationship in the embodiment. Therefore, under the condition that the weight of each decision maker is unknown, the relative importance of each decision maker can be expressed by distributed language information expression, a corresponding relative importance matrix is further established, the absolute importance of each decision maker can be obtained by processing the relative importance matrix, and the weight of each decision maker can be better obtained by normalizing the absolute importance; then, based on the situation of each decision maker which is characterized by the distributed language information, a relevant granulation model can be preferably established based on the consistency of language preference and/or group preference consensus, and further, the language term problem can be preferably converted into a numerical processing problem, so that an optimal design scheme can be preferably obtained.
The present invention and its embodiments have been described above schematically, without limitation, and what is shown in the drawings is only one of the embodiments of the present invention, and the actual structure is not limited thereto. Therefore, if the person skilled in the art receives the teaching, without departing from the spirit of the invention, the person skilled in the art shall not inventively design the similar structural modes and embodiments to the technical solution, but shall fall within the scope of the invention.

Claims (8)

1. A method for selecting an automobile product design scheme based on PSO information granulation comprises the following steps:
step S1, establishing a candidate set X and a scheme evaluation language term set L for evaluating the preference degree of the candidate1(ii) a Wherein X ═ { X ═ Xt|t＝1,2,3,…,i,…,j,…,n}，xtRepresents the tth alternative; wherein L is1＝{lh|h＝-p,-p+1,…,0,…,p-1,p}，lhThe h language term, language term l, representing the evaluation of the scheme-pTo the linguistic term lpThe indicated preference degree is increased;
step S2, establishing a decision maker set E and a decision maker evaluation language term set L for evaluating the preference degree of the decision maker2(ii) a Wherein E ═ { E ═ Ew|w＝1,2,3,…,s…,g,…,m}，ewRepresents the w-th decision maker; wherein L is2＝{lr|r＝-q,-q+1,…,0,…,q-1,q}，lrThe r-th language term, language term l, representing the evaluation of a decision maker-qTo the linguistic term lqThe indicated preference degree is increased;
step S3, each decision maker evaluates the language term set L based on the scheme1The language term in (1) carries out language term evaluation on all the alternatives, and a language term evaluation matrix D of each decision maker on all the alternatives is established1w，D1wRepresenting a matrix obtained after the w-th decision maker evaluates all the alternative schemes; wherein the content of the first and second substances,
Figure FDA0002903224650000011
Figure FDA0002903224650000012
indicating the w-th decision maker compared to alternative xjFor alternative xi(ii) a linguistic term evaluation of (a); wherein the content of the first and second substances,
Figure FDA0002903224650000013
Figure FDA0002903224650000014
representing linguistic term evaluation
Figure FDA0002903224650000015
Chinese language term lhThe probability of occupation; wherein the content of the first and second substances,
Figure FDA0002903224650000016
and is
Figure FDA0002903224650000017
Step S4, evaluating language term set L based on decision maker2The language term in (1) carries out language term evaluation on all decision-makers, and establishes a language term evaluation matrix D for the decision-makers2(ii) a Wherein the content of the first and second substances,
Figure FDA0002903224650000018
Figure FDA0002903224650000019
representing the comparison to decision egFor decision maker es(ii) a linguistic term evaluation of (a); wherein the content of the first and second substances,
Figure FDA00029032246500000110
Figure FDA00029032246500000111
representing linguistic term evaluation
Figure FDA00029032246500000112
Chinese language term lrThe probability of occupation; wherein the content of the first and second substances,
Figure FDA00029032246500000113
and is
Figure FDA00029032246500000114
Step S5, granulating the language terms to obtain a language term evaluation matrix D1wNumerical evaluation matrix ND of1w，
Figure FDA00029032246500000115
Figure FDA00029032246500000116
Indicating the w-th decision maker compared to alternative xjFor alternative xiNumerical evaluation of (1); and a language term evaluation matrix D2Numerical evaluation matrix ND of2，
Figure FDA00029032246500000117
Figure FDA00029032246500000118
Representing the comparison to decision egFor decision maker esNumerical evaluation of (1);
step S6, evaluating the matrix ND according to the numeralization1wAnd a numerical evaluation matrix ND2And acquiring the preference degree of each alternative scheme so as to acquire the optimal design scheme.
2. The method for selecting an automotive product design solution based on PSO information pelletization of claim 1, wherein: step S5 includes scheme evaluation language term set L1And the decision maker evaluates the language term set L2In the form of message granules, each language term is expressed as an interval [0, 1%]A subset of (a).
3. The method for selecting an automotive product design solution based on PSO information pelletization of claim 2, wherein: in step S5, the language term set L is evaluated for the decider2When granulation is carried out, the method comprises the following steps:
step S511, setting language term lrCorresponding subinterval is [ b ]r-1,br]And setting b-q-1＝0，bq＝1；
Step S512, calculating the intercept point vector [ b-q,b-q+1,…,b0,…bq-2,bq-1]And the decision maker evaluates the language term set L2Set of values W ofL2，WL2＝{WLr-q +1, …,0, …, q-1, q }; wherein, WLrDenotes the language term lrBy the term l in the languagerContinuously and uniformly sampling and obtaining in the corresponding subinterval;
in this step, the matrix ND is evaluated numerically2Index Q of multiplicative consistency2As an optimization index and by a multiplicative consistency index Q2Mean value of
Figure FDA0002903224650000021
Establishing an optimization model as a fitness function, wherein
Figure FDA0002903224650000022
Figure FDA0002903224650000023
Figure FDA0002903224650000024
Representing a numerical evaluation matrix ND established based on values obtained by continuous uniformly distributed sampling in the kth operation2A multiplicative consistency indicator of;
the optimization model is established as follows,
MaxQ2＝cND2
Figure FDA0002903224650000025
wherein, "cND2"indicates the numerical evaluation matrix ND obtained for each operation2A multiplicative consistency indicator of;
constructed fitness function f2In order to realize the purpose,
Figure FDA0002903224650000026
step S513, constraint condition based on optimization model and fitness function f2Get the intercept point vector [ b ]-q,b-q+1,…,b0,…bq-2,bq-1]And a numerical evaluation matrix ND2(ii) a Wherein the content of the first and second substances,
Figure FDA0002903224650000031
4. the PSO information granulation-based automotive product design choice method as claimed in claim 3, whereinIn the following steps: further comprising the calculation of a decision maker weight matrix W, { W ═ Ww1, | w ═ 1,2,3, …, s …, g, …, m, Ww denote the weight of the w-th decision maker; the calculation formula of Ww is as follows,
Figure FDA0002903224650000032
5. the method of claim 4, wherein the PSO information based vehicle product design selection process comprises:
in step S5, the language term set L is evaluated for the scheme1When granulation is carried out, the method comprises the following steps:
step S521, setting language term lhCorresponding subinterval is [ a ]h-1,ah]And set a-p-1＝0，ap＝1；
Step S522, calculating intercept point vector [ a ] based on PSO algorithm-p,a-p+1,…,a0,…ap-2,ap-1]And a scheme evaluation language term set L1Set of values W ofL1，WL1＝WLh-p, -p +1, …,0, …, p-1, p }; wherein, WLhDenotes the language term lhBy the term l in the languagehContinuously and uniformly sampling and obtaining in the corresponding subinterval;
in the step of the method,
the optimization model is established as follows,
Figure FDA0002903224650000033
Figure FDA0002903224650000034
constructed fitness function f1In order to realize the purpose,
Figure FDA0002903224650000035
wherein the content of the first and second substances,
Figure FDA0002903224650000036
cND1wrepresents a numerical evaluation matrix ND1wThe consistency index of (a); o is2Is corresponding to O1The index of the consistency ratio of (1), delta is a balance factor, and delta belongs to [0,1]]；
Step S523, constraint condition based on optimization model and fitness function f1Obtaining an intercept point vector [ a ]-p,a-p+1,…,a0,…ap-2,ap-1]And a numerical evaluation matrix ND1w(ii) a Wherein the content of the first and second substances,
Figure FDA0002903224650000041
6. the method of claim 5, wherein the PSO information based vehicle product design selection process comprises: in step S6, a total evaluation matrix ND is acquiredc，
Figure FDA0002903224650000042
Wherein the content of the first and second substances,
Figure FDA0002903224650000043
7. the method of claim 6, wherein the PSO information based vehicle product design selection process comprises: in step S6, the total evaluation matrix ND is subjected to averaging based on the average operatorcCalculating to obtain a scheme evaluation matrix X ', X ═ X't)1×n，
Figure FDA0002903224650000044
8. The method of claim 7, wherein the PSO information based automotive product design selection process comprises: in step S6, the optimal solution can be obtained by sorting based on the solution evaluation matrix X'.
CN202110063426.1A 2021-01-18 2021-01-18 Automobile product design scheme selection method based on PSO information granulation Active CN112836293B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110063426.1A CN112836293B (en) 2021-01-18 2021-01-18 Automobile product design scheme selection method based on PSO information granulation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110063426.1A CN112836293B (en) 2021-01-18 2021-01-18 Automobile product design scheme selection method based on PSO information granulation

Publications (2)

Publication Number Publication Date
CN112836293A true CN112836293A (en) 2021-05-25
CN112836293B CN112836293B (en) 2022-09-30

Family

ID=75928572

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110063426.1A Active CN112836293B (en) 2021-01-18 2021-01-18 Automobile product design scheme selection method based on PSO information granulation

Country Status (1)

Country Link
CN (1) CN112836293B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113705965A (en) * 2021-07-19 2021-11-26 合肥工业大学 Satellite observation scheme screening method and system based on intuitive language preference relation particles

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004118552A (en) * 2002-09-26 2004-04-15 Toshiba Corp Method for supporting multipurpose decision making process and its program
CN102693287A (en) * 2012-05-11 2012-09-26 北京航空航天大学 Automatic group decision-making consensus degree adjustment method based on clustering method and particle swarm optimization method
CN104657778A (en) * 2015-03-20 2015-05-27 华东交通大学 Interval binary semantic information axiom based uncertain language group decision method
WO2019061187A1 (en) * 2017-09-28 2019-04-04 深圳乐信软件技术有限公司 Credit evaluation method and apparatus and gradient boosting decision tree parameter adjustment method and apparatus
CN110009250A (en) * 2019-04-13 2019-07-12 郑州航空工业管理学院 A kind of twin workshop quality evaluating method of number based on hesitation fuzzy language
CN110837966A (en) * 2019-11-07 2020-02-25 延安大学 SNA-based building equipment optimization method based on mixed multi-attribute group decision
CN111784109A (en) * 2020-06-01 2020-10-16 上海大学 Supplier selection method based on Bidagolas fuzzy set and VIKOR

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004118552A (en) * 2002-09-26 2004-04-15 Toshiba Corp Method for supporting multipurpose decision making process and its program
CN102693287A (en) * 2012-05-11 2012-09-26 北京航空航天大学 Automatic group decision-making consensus degree adjustment method based on clustering method and particle swarm optimization method
CN104657778A (en) * 2015-03-20 2015-05-27 华东交通大学 Interval binary semantic information axiom based uncertain language group decision method
WO2019061187A1 (en) * 2017-09-28 2019-04-04 深圳乐信软件技术有限公司 Credit evaluation method and apparatus and gradient boosting decision tree parameter adjustment method and apparatus
CN110009250A (en) * 2019-04-13 2019-07-12 郑州航空工业管理学院 A kind of twin workshop quality evaluating method of number based on hesitation fuzzy language
CN110837966A (en) * 2019-11-07 2020-02-25 延安大学 SNA-based building equipment optimization method based on mixed multi-attribute group decision
CN111784109A (en) * 2020-06-01 2020-10-16 上海大学 Supplier selection method based on Bidagolas fuzzy set and VIKOR

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Y SHEN等: "A Data-Driven Fuzzy Information Granulation Approach for Freight Volume Forecasting", 《IEEE XPLORE》 *
张世涛: "异质语言信息群体共识决策模型及其应用研究", 《万方学位论文》 *
徐新清等: "构建不确定语言型多属性决策的投影模型", 《数学的实践与认识》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113705965A (en) * 2021-07-19 2021-11-26 合肥工业大学 Satellite observation scheme screening method and system based on intuitive language preference relation particles

Also Published As

Publication number Publication date
CN112836293B (en) 2022-09-30

Similar Documents

Publication Publication Date Title
Rao A decision making methodology for material selection using an improved compromise ranking method
CN110135630B (en) Short-term load demand prediction method based on random forest regression and multi-step optimization
CN110990461A (en) Big data analysis model algorithm model selection method and device, electronic equipment and medium
CN109741120B (en) Product deep personalized customization method for user requirements
Cheng et al. Artificial intelligence approaches to achieve strategic control over project cash flows
Peng et al. Picture fuzzy large-scale group decision-making in a trust-relationship-based social network environment
Yan Research on green suppliers' evaluation based on AHP & genetic algorithm
Wu et al. Fuzzy support vector regression machine with penalizing Gaussian noises on triangular fuzzy number space
Sarkar et al. A multi‐criteria decision making approach for strategy formulation using Pythagorean fuzzy logic
CN112836293B (en) Automobile product design scheme selection method based on PSO information granulation
Tembusai et al. K-nearest neighbor with K-fold cross validation and analytic hierarchy process on data classification
Liu et al. Three-way decisions with single-valued neutrosophic decision theory rough sets based on grey relational analysis
Fu et al. DEA cross‐efficiency aggregation based on preference structure and acceptability analysis
Comesaña-Campos et al. The Value Index as a decision support tool applied to a new system for evaluating and selecting design alternatives
Li et al. Application of fuzzy-based hybrid taguchi method for multiobjective optimization of product form design
CN115994778A (en) Behavior fine portrait method for multiple users
CN113792984B (en) Cloud model-based anti-air defense anti-pilot command control model capability assessment method
CN112270047B (en) Urban vehicle path optimization method based on data-driven group intelligent calculation
CN109345274B (en) Neighbor user selection method based on BP neural network scoring prediction error
CN108364136B (en) Water resource shortage risk analysis method and system based on evidence reasoning
Mohammadi et al. A novel approach to fuzzy multi-attribute group decision making based on interval-valued intuitionistic fuzzy best-worst method
Zeyang Research on intelligent acceleration algorithm for big data mining in communication network based on support vector machine
Petrovsky et al. Multi-stage technique'PAKS'for multiple criteria decision aiding
CN110717545B (en) Personalized customization method based on improved interactive artificial immune algorithm
Pan Financial Risk Assessment Model Based on Stochastic Forest and Decision Tree Hybrid Algorithm

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