CN103136402B - The EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure and device - Google Patents
The EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure and device Download PDFInfo
- Publication number
- CN103136402B CN103136402B CN201110396039.6A CN201110396039A CN103136402B CN 103136402 B CN103136402 B CN 103136402B CN 201110396039 A CN201110396039 A CN 201110396039A CN 103136402 B CN103136402 B CN 103136402B
- Authority
- CN
- China
- Prior art keywords
- value
- geometric parameter
- factor
- parameter factor
- geometric
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention discloses EXPERIMENTAL DESIGN list acquirement method and the device of a kind of man-made microstructure, comprise determining that the number of the geometric parameter factor of man-made microstructure to be measured, the value number of each geometric parameter factor and the parameter value of correspondence thereof；Number, the value number of each geometric parameter factor and the parameter value of correspondence thereof according to the geometric parameter factor, and according to uniform Design principle structure initial trial design table；Electromagnetic response delta data when testing according to initial trial design table, obtains active geometric parameter scope in each geometric parameter factor；Active geometric parameter scope in each geometric parameter factor according to Latin Hypercube Sampling method and acquisition, structure obtains EXPERIMENTAL DESIGN table.Use the present invention, experimental procedure can be reduced, it is achieved in the experiment that quantity is few, draw distortionless experimental result in the case of not affecting experimental result.
Description
Technical field
The present invention relates to man-made microstructure test field, particularly relate to the EXPERIMENTAL DESIGN table acquisition side of a kind of man-made microstructure
Method and device.
Background technology
Artificial electromagnetic material is also referred to as Meta Materials, is a kind of synthetic material that electromagnetic wave can produce response, by
Substrate and the man-made microstructure composition being attached on substrate, owing to man-made microstructure is typically by having that conductive material is arranged into
The structure of certain geometrical pattern, therefore, it is possible to produce response to electromagnetic wave, so that Meta Materials entirety embodies and is different from substrate
Electromagnetic property.Man-made microstructure is that the performance of electromagnetic material plays vital effect.
Measure the electromagnetic property of man-made microstructure, be an important step in artificial electromagnetic material design process.In order to
Obtain the electromagnetic response characteristic rule to certain types of man-made microstructure, it is often necessary to do substantial amounts of experiment.At present, this field
" orthogonal design " generally used obtains EXPERIMENTAL DESIGN table, then carries out test according to this EXPERIMENTAL DESIGN table and obtains particular type
The electromagnetic response characteristic rule of man-made microstructure.
As a example by the man-made microstructure of the I-shaped shown in Fig. 1, adopt in prior art, generate test with orthogonal design
The flow process of design table is:
Determine the geometric parameter factor of the man-made microstructure of I-shaped, it is determined here that I-shaped man-made microstructure is several
What parameter factors includes: high a, long b and live width w；
Determine the value number of each geometric parameter factor, and determine parameters numerical value.Can first determine
One numerical range, then in this numerical range, determines each parameter values corresponding to each factor.
Herein, certainty factor w is preset parameter, and parameter values is 0.2, the value of factor a 5, scope be [1.6000,
3.0000], parameter values is a=[1.6000,1.9500,2.3000,2.6500,3.0000], and the value number of factor b is also wrapped
Include 5, in the range from [0.0500,1.4500], each parameter values be respectively b=[0.0500,0.4000,0.7500,
1.1000,1.4500]。
According to orthogonal design, according to factor a and the b of change in value, available following orthogonal table:
Table 1:
| Sequence number | Parameter 1 | Parameter 2 |
| 1 | 1 | 1 |
| 2 | 1 | 2 |
| 3 | 1 | 3 |
| 4 | 1 | 4 |
| 5 | 1 | 5 |
| 6 | 2 | 1 |
| 7 | 2 | 2 |
| 8 | 2 | 3 |
| 9 | 2 | 4 |
| 10 | 2 | 5 |
| 11 | 3 | 1 |
| 12 | 3 | 2 |
| 13 | 3 | 3 |
| 14 | 3 | 4 |
| 15 | 3 | 5 |
| 16 | 4 | 1 |
| 17 | 4 | 2 |
| 18 | 4 | 3 |
| 19 | 4 | 4 |
| 20 | 4 | 5 |
| 21 | 5 | 1 |
| 22 | 5 | 2 |
| 23 | 5 | 3 |
| 24 | 5 | 4 |
| 25 | 5 | 5 |
Factor a is corresponding with described orthogonal table with the parameter values of b, and combine factor w that numerical value is fixing, i.e. can get table
EXPERIMENTAL DESIGN table described in 2.
Table 2:
After obtaining above-mentioned EXPERIMENTAL DESIGN table, according still further to the content of EXPERIMENTAL DESIGN table, make 25 corresponding length, height and line
Then these 25 man-made microstructure are carried out electromagnetism test by the I-shaped man-made microstructure of wide parameter.
From the foregoing, it will be observed that existing orthogonal design needs to make substantial amounts of man-made microstructure, and need to carry out substantial amounts of electricity
Magnetic l-G simulation test, such as, when choosing the parameter value number of the high a of I-shaped man-made microstructure, long b and live width w respectively
When being 30,30,1, need 30*30=900 Electromagnetic Simulation test according to orthogonal design, need 900 relevant parameter numerical value
The man-made microstructure of combination.So many Electromagnetic Simulation is arranged to test and man-made microstructure, the most time-consuming, effort, and
Cost is the highest.
Summary of the invention
Embodiment of the present invention technical problem to be solved is, how to obtain the EXPERIMENTAL DESIGN table of man-made microstructure, can
Make in the case of not affecting result of the test, reduce test number (TN).
In order to solve above-mentioned technical problem, the EXPERIMENTAL DESIGN table embodiments providing a kind of man-made microstructure obtains
Method, including:
Determine the number of the geometric parameter factor of man-made microstructure to be measured, the value number of each geometric parameter factor and
Corresponding parameter value；
Number, the value number of each geometric parameter factor and the parameter of correspondence thereof according to the described geometric parameter factor
Value, and according to uniform Design principle structure initial trial design table；
Electromagnetic response delta data when testing according to described initial trial design table, obtains each geometric parameter factor
In active geometric parameter scope；
Active geometric parameters in each geometric parameter factor according to Latin Hypercube Sampling method and described acquisition
Number scope, structure obtains EXPERIMENTAL DESIGN table.
Wherein, the described number according to the described geometric parameter factor determined, the value number of each geometric parameter factor
And the parameter value of correspondence, and include according to uniform Design principle structure initial trial design table:
According to the number of the described geometric parameter factor arranged, the value number of each geometric parameter factor, and according to all
Even design principle, makes to make good use of grid point method structure uniform designs table；
Dividing the matrix obtaining multiple s row from described uniform designs table, wherein, described s value is less than described uniform Design
Total columns of table；
Multiple matrixes of described acquisition are carried out centralization deviation calculating, obtains the square that centralization deviation therein is minimum
Battle array；
The matrix minimum according to described centralization deviation and the described parameter value determined, obtain initial trial design table.
It is wherein, described according to the number of the described geometric parameter factor arranged, the value number of each geometric parameter factor,
And according to uniform Design principle, make to make good use of grid point method structure uniform designs table and include:
Arranging test number (TN) value n, described test number (TN) value n is the least common multiple of the value number of each geometric parameter factor
Number；
According to described test number (TN) value n, extract less than described test number (TN) value n and become mutual with described test number (TN) value n
All numerical value of matter relation；
According to test number (TN) value n and all numerical value of described extraction, generate each of uniform designs table according to good grid point method
Row, it is thus achieved that uniform designs table.
Wherein, described each geometric parameter factor according to Latin Hypercube Sampling method and described acquisition plays a leading role
Geometric parameter scope, structure obtains EXPERIMENTAL DESIGN table and includes:
According to described test number (TN) value n, use Latin Hypercube Sampling method to calculate and obtain sample value；
According to active geometric parameter scope, structure in calculated sample value and described each geometric parameter factor
Make and obtain EXPERIMENTAL DESIGN table.
Wherein, the number of the described geometric parameter factor determining man-made microstructure to be measured, the taking of each geometric parameter factor
The parameter value of value number and correspondence thereof includes:
Determine the number of the geometric parameter factor of man-made microstructure to be measured, the value number of each geometric parameter factor；
Determine the span of each geometric parameter factor to determine each geometric parameter according to described value number
The parameter value that factor pair is answered.
Correspondingly, the embodiment of the present invention additionally provides the EXPERIMENTAL DESIGN table acquisition device of a kind of man-made microstructure, including:
Parameter determination module, for determining the number of the geometric parameter factor of man-made microstructure to be measured, each geometric parameter
The value number of the factor and the parameter value of correspondence thereof；
First constructing module, for according to the number of the described geometric parameter factor, the value of each geometric parameter factor
Several and corresponding parameter value, and according to uniform Design principle structure initial trial design table；
Parameter chooses module, the electromagnetic response delta data in time testing according to described initial trial design table,
Obtain active geometric parameter scope in each geometric parameter factor；
Second constructing module, rises in each geometric parameter factor according to Latin Hypercube Sampling method and described acquisition
The geometric parameter scope of mastery reaction, structure obtains EXPERIMENTAL DESIGN table.
Wherein, described first constructing module includes:
Structural unit, for according to the number of the described geometric parameter factor arranged, the value of each geometric parameter factor
Number, and according to uniform Design principle, make to make good use of grid point method structure uniform designs table；
Calculate acquiring unit, for dividing the matrix obtaining multiple s row, wherein, described s value from described uniform designs table
Less than total columns of described uniform designs table, multiple matrixes of described acquisition are carried out centralization deviation calculating, obtains therein
The matrix that centralization deviation is minimum, the matrix minimum according to described centralization deviation and the described parameter value determined, obtain initial
EXPERIMENTAL DESIGN table.
Wherein, described structural unit is according to test number (TN) value n preset with less than described test number (TN) value n and with described
Test number (TN) value n becomes all numerical value of relatively prime relation, and generates each row of uniform designs table according to good grid point method, it is thus achieved that uniformly
Design table；
Wherein, described default test number (TN) value n be sample point number corresponding to each geometric parameter that arrange
Little common multiple.
Wherein, described acquisition module includes:
Computing unit, for according to described test number (TN) value n, uses Latin Hypercube Sampling method to calculate and obtains sampling
Value；
Acquiring unit, for according to active several in calculated sample value and described each geometric parameter factor
What parameter area, structure obtains EXPERIMENTAL DESIGN table.
Wherein, described parameter determination module is additionally operable to determine the span of each geometric parameter factor.
Implement the embodiment of the present invention, have the advantages that
In the EXPERIMENTAL DESIGN table acquisition process of man-made microstructure, build uniform designs table according to uniform Design principle, obtain
Take the initial trial design table that parameter to be measured is few, and after testing according to this initial trial design table, determine leading work
Geometric parameter scope, and obtain final examination after carrying out Latin Hypercube Sampling calculating according to the geometric parameter scope determined
Test design table, test according to this experimental design table, contribute in the case of not affecting experimental result, reduce experiment work
Sequence, it is achieved draw distortionless experimental result in the experiment that quantity is few.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
In having technology to describe, the required accompanying drawing used is briefly described, it should be apparent that, the accompanying drawing in describing below is only this
Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, it is also possible to
Other accompanying drawing is obtained according to these accompanying drawings.
Fig. 1 is a kind of I-shaped man-made microstructure to be measured schematic diagram；
Fig. 2 is the first embodiment flow process signal of the EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure of the present invention
Figure；
Fig. 3 is the second embodiment flow process signal of the EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure of the present invention
Figure；
Fig. 4 is the structure composition signal of the EXPERIMENTAL DESIGN table acquisition device of a kind of man-made microstructure of the embodiment of the present invention
Figure.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Describe, it is clear that described embodiment is only a part of embodiment of the present invention rather than whole embodiments wholely.Based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under not making creative work premise
Embodiment, broadly falls into the scope of protection of the invention.
Refer to Fig. 2, be the first embodiment stream of the EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure of the present invention
Journey schematic diagram；The described method of the present embodiment includes:
S101: determine the value of the number of the geometric parameter factor of man-made microstructure to be measured, each geometric parameter factor
Several and corresponding parameter value.
In concrete described S101, each parameter factors can be configured by user, such as required, for shown in Fig. 1
I-shaped man-made microstructure, the geometric parameter factor that user can be arranged is: long a, wide b and live width w；A takes 5 values, and b takes 5
Individual value, and w only takes a value, the concrete numerical value of each factor, then be determined according to the size of required electromagnetic material.
User also can be first according to the size of required Meta Materials, it is first determined go out the span of each parameter factors, from
Some parameter values are extracted so that subsequent construction EXPERIMENTAL DESIGN table in the range of Gai.
Described S101 specifically comprises the steps that the number of the geometric parameter factor determining man-made microstructure to be measured, each geometric parameters
The value number of the number factor；Determine the span of each geometric parameter factor to determine each according to described value number
The parameter value that geometric parameter factor pair is answered.
S102: according to number, the value number of each geometric parameter factor and the correspondence thereof of the described geometric parameter factor
Parameter value, and according to uniform Design principle structure initial trial design table.
Uniform Design principle be based on testing site in whole trial stretch uniformly dispersing from uniformity angle
A kind of test design method, according to number, the value number of each geometric parameter factor of the described geometric parameter factor, uses all
Even design principle can get a design table, and on this table, the parameters factor is dispersed, in this design table, and each list item
Numerical value represent the sequence number of parameter value of the corresponding geometric parameter factor, by corresponding for the parameter value of the parameters factor in this design table
It is entered in corresponding list item, gets final product initial trial design table described in construction complete.
S103: electromagnetic response delta data when testing according to described initial trial design table, obtains each geometric parameters
Active geometric parameter scope in the number factor.
User can design table according to the initial trial of described structure, makes corresponding man-made microstructure according to the parameter value of each row
Model, and these models are carried out respectively electromagnetism test, obtain the change of the electromagnetic response data of each artificial microstructure model
Change situation, and determine parameter area according to changing violent situation, can be further according to changing violent situation making change song
Line chart, then determines parameter area according to curve chart.Such as, for the I-shaped man-made microstructure of Fig. 1, testing
After, it may be determined that go out described geometric parameter factor a be the scope of [1.6000,2.0667], geometric parameter factor b be [0.5167,
During scope 0.9833], change is relatively violent, and changing value is above certain threshold value, concrete, and this threshold value can be according to user to super
The test requirements document of material is configured.
S104: active several in each geometric parameter factor according to Latin Hypercube Sampling method and described acquisition
What parameter area, structure obtains EXPERIMENTAL DESIGN table.
I.e. it is again with Latin Hypercube Sampling method in the range of described, again determine each geometric parameter factor
Value number and new parameter value thereof, thus structure obtains EXPERIMENTAL DESIGN table.
The present invention, in the EXPERIMENTAL DESIGN table acquisition process of man-made microstructure, builds uniform Design according to uniform Design principle
Table, obtains the initial trial design table that parameter to be measured is few, and after testing according to this initial trial design table, has determined main
Lead the geometric parameter scope of effect, and obtain after carrying out Latin Hypercube Sampling calculating according to the geometric parameter scope determined final
EXPERIMENTAL DESIGN table, test according to this experimental design table, contribute in the case of not affecting experimental result, reduce experiment
Operation, it is achieved draw distortionless experimental result in the experiment that quantity is few.
Refer to Fig. 3 again, be the second embodiment of the EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure of the present invention
Schematic flow sheet；The described method of the present embodiment includes:
S201: determine the value of the number of the geometric parameter factor of man-made microstructure to be measured, each geometric parameter factor
Several and corresponding parameter value.
S202: according to the number of the described geometric parameter factor arranged, the value number of each geometric parameter factor, and root
According to uniform Design principle, make to make good use of grid point method structure uniform designs table；
Concrete, described S202 includes:
Arranging test number (TN) value n, described test number (TN) value n is the least common multiple of the value number of each geometric parameter factor
Number；
According to described test number (TN) value n, extract less than described test number (TN) value n and become mutual with described test number (TN) value n
All numerical value of matter relation.
According to test number (TN) value n and all numerical value of described extraction, generate each of uniform designs table according to good grid point method
Row, it is thus achieved that uniform designs table.
Wherein, it is thus achieved that the computing formula of the good grid point method that uniform designs table is used is:
u1j=hj；
Wherein, i=1,2 ..., n-1, hjIt is designated as jth and becomes relatively prime with n
The value of relation.
S203: dividing the matrix obtaining multiple s row from described uniform designs table, wherein, described s value becomes for parameter values
The number of the geometric parameter factor changed, it is less than total columns of described uniform designs table；.
S204: multiple matrixes of described acquisition are carried out centralization deviation calculating, obtains centralization deviation therein minimum
Matrix；
S205: the matrix minimum according to described centralization deviation and the described parameter value determined, obtain initial trial design
Table.
S206: electromagnetic response delta data when testing according to described initial trial design table, obtains each geometric parameters
Active geometric parameter scope in the number factor；
S207: according to test number (TN) n, uses Latin Hypercube Sampling method to calculate and obtains sample value；
The Latin hypercube that described S207 is used belongs to a kind of affined sampling method, Latin Hypercube Sampling
First [0,1] interval is divided into the subinterval of n non-overlapping copies by rule, then carries out independent in each subinterval respectively
Sampling with equal probability, obtains n the random number i.e. sample value more than 0 less than 1.
The computing formula of random number is:
And relational expressionWherein, i=1,2 ..., n, uiValue be i-th son
The sample value that interval sampling obtains.
Owing to there is described relational expression so that each subinterval is only capable of producing a random number, be i.e. only capable of according to
Test number (TN) n needed for family obtains n random number.
S208: according to active geometric parameter model in calculated sample value and described each geometric parameter factor
Enclosing, structure obtains EXPERIMENTAL DESIGN table.
With an instantiation, the EXPERIMENTAL DESIGN list acquirement method of the man-made microstructure described in the present embodiment is said below
Bright.
Same as a example by the I-shaped man-made microstructure of Fig. 1, the described method idiographic flow of this example is:
S1: arrange: long a, wide b and live width w totally three geometric parameter factors, long parameter factors a takes 5 values, each value
For a=[1.6000,1.9500,2.3000,2.6500,3.0000], wide parameter factors b takes 5 values equally, and each value is b=
[0.0500,0.4000,0.7500,1.1000,1.4500], w takes a value, is worth for w=0.2.
S2: according to number, the value number of each geometric parameter factor of the described geometric parameter factor determined, and according to
Uniform Design principle, makes to make good use of grid point method structure uniform designs table.
As required, default needs carry out frequency n=5 tested, and yield less than n and become the numerical value bag of relatively prime relation with n
Include: 1,2,3,4.
Equation below according to good grid point method:
u1j=hj；
Wherein, i=1,2 ..., n-1, hjIt is designated as jth and becomes relatively prime with n
The numerical value of relation；
Obtain uniform designs table as shown in table 3,
Table 3:
| 1 | 2 | 3 | 4 |
| 2 | 4 | 1 | 3 |
| 3 | 1 | 4 | 2 |
| 4 | 3 | 2 | 1 |
| 5 | 5 | 5 | 5 |
Take s=2 and construct new matrix, and according to L2 centralization deviation computing formula, the new matrix of structure is calculated, can
Obtain:
1. first row and the new matrix of secondary series composition, CD2=0.1125,
2. first row and the new matrix of the 3rd row composition, CD2=0.1125
3. first row and the new matrix of the 4th row composition, CD2=0.1139
4. secondary series and the new matrix of the 3rd row composition, CD2=0.1139
5. secondary series and the new matrix of the 4th row composition, CD2=0.1125
6. the 3rd row and the new matrix of the 4th row composition, CD2=0.1125
According to above-mentioned calculated deviation, can obtain the 1st, 2,5, the matrix that formed of 6 kind of situation the most all have less partially
Difference, this means that these matrixes all have preferable uniformity and spatial representative, can optionally one of which situation arrange
Test.
In this example, the first row selecting CD2 deviation value to be 0.1125 and the new matrix of secondary series composition, as follows
Table 4:
Table 4:
| Sequence number | Parameter 1 | Parameter 2 |
| 1 | 1 | 2 |
| 2 | 2 | 4 |
| 3 | 3 | 1 |
| 4 | 4 | 3 |
| 5 | 5 | 5 |
Corresponding parameter value is updated in described table 4, i.e. can get the initial trial i.e. table 5 below of design table.
Table 5:
| Sequence number | a | b | w |
| 1 | 1.6000 | 0.4000 | 0.2 |
| 2 | 1.9500 | 1.1000 | 0.2 |
| 3 | 2.3000 | 0.0500 | 0.2 |
| 4 | 2.6500 | 0.7500 | 0.2 |
| 5 | 3.0000 | 1.4500 | 0.2 |
S3: during arranging man-made microstructure model to test according to above-mentioned table 5, determines the region that change is violent
For: the geometric parameter scope of geometric parameter factor a is [1.6000,2.0667], and the geometric parameter scope of geometric parameter factor b is
[0.5167,0.9833]。
S4: according to scope obtained above, need to carry out 8 tests, the most again from above-mentioned according to test needs, such as user
Scope gathers n2=8 numerical value and carries out Latin Hypercube Sampling, i.e. obtain random in [0,1] is interval according to equation below
Number.
The computing formula of random number is:
And relational expressionWherein, i=1,2 ..., n, uiBe [0,1] interval in the
The sample value in i subinterval.
I=n2=8 herein.Described n2=8 is also the calculating parameter arranged the most voluntarily by user.
The Latin Hypercube Sampling table that formula according to above-mentioned random number generates is as follows:
Table 6:
Then according to aiValue=1.6000+ (2.0667-1.6) × sampling 1 (i), i=1 ... 8, and biValue=
0.5167+ (0.9833-0.5176) × sampling 2 (i), i=1 ... 8, i.e. can get final EXPERIMENTAL DESIGN table as shown in table 7.
Table 7:
| Sequence number | a | b | w |
| 1 | 1.74584375000000 | 0.895081250000000 | 0.2 |
| 2 | 1.62916875000000 | 0.778656250000000 | 0.2 |
| 3 | 1.80418125000000 | 0.545806250000000 | 0.2 |
| 4 | 1.97919375000000 | 0.662231250000000 | 0.2 |
| 5 | 1.92085625000000 | 0.953293750000000 | 0.2 |
| 6 | 2.03753125000000 | 0.720443750000000 | 0.2 |
| 7 | 1.86251875000000 | 0.836868750000000 | 0.2 |
| 8 | 1.68750625000000 | 0.604018750000000 | 0.2 |
After obtaining table 7, user can design corresponding man-made microstructure model according to the parameter size be given in table 7, enter
The required test of row.
It should be noted that owing to w value is fixed, the most above-mentioned table 3,4 and 6 is also not introduced into geometric parameter factor w.
By above-described embodiment, according to orthogonal design mode, carrying out this test needs to carry out 5*5*1 that is 25
Secondary, and use the present invention only to need according to user's needs, arrange to test for 8 times, considerably reduce test number (TN), and, this
Invention is in the EXPERIMENTAL DESIGN table acquisition process of man-made microstructure, based on uniform Design principle and Latin Hypercube Sampling method
Obtain EXPERIMENTAL DESIGN table, also do not interfere with experimental result.
Refer to Fig. 4, be that the structure composition of the EXPERIMENTAL DESIGN table acquisition device of the man-made microstructure of the embodiment of the present invention shows
Being intended to, this device includes: parameter determination module the 1, first constructing module 2, parameter choose module 3 and the second constructing module 4.Its
In:
Described parameter determination module 1, for determining the number of the geometric parameter factor of man-made microstructure to be measured, each geometry
The value number of parameter factors and the parameter value of correspondence thereof；
Described first constructing module 2, for according to the number of the described geometric parameter factor, the taking of each geometric parameter factor
Value number and the parameter value of correspondence thereof, and according to uniform Design principle structure initial trial design table；
Described parameter chooses module 3, the electromagnetic response change in time testing according to described initial trial design table
Data, obtain active geometric parameter scope in each geometric parameter factor；
Described second constructing module 4, for according to each geometric parameter of Latin Hypercube Sampling method and described acquisition because of
Active geometric parameter scope in son, structure obtains EXPERIMENTAL DESIGN table.
Concrete, user can arrange the geometric parameter factor of man-made microstructure to be measured by described parameter determination module 1
Number, the value number of each geometric parameter factor and the parameter value of correspondence thereof.
Concrete, can make to make good use of grid point method structure uniformly set according to uniform Design principle at described first constructing module 2
Meter table, then obtains multiple new matrix according to described uniform designs table, and uses the centralization L2 deviation deviation to each new matrix
Calculating, the new matrix less in deviation chooses one of them, and as initial after the parameter value corresponding to the interpolation of its list item
EXPERIMENTAL DESIGN table.
Referring to Fig. 4, described first constructing module 2 comprises the steps that structural unit 21 and calculates acquiring unit 22.
Described structural unit 21, for according to the number of the described geometric parameter factor arranged, each geometric parameter factor
Value number, and according to uniform Design principle, make to make good use of grid point method structure uniform designs table；
Described calculating acquiring unit 22, for dividing the matrix obtaining multiple s row from described uniform designs table, wherein,
Described s value, less than total columns of described uniform designs table, carries out centralization deviation calculating to multiple matrixes of described acquisition, obtains
The matrix that centralization deviation therein is minimum, the matrix minimum according to described centralization deviation and the described parameter value determined,
Table is designed to initial trial.
Described structural unit 21 determine according to user needed for test number (TN) value n, determine and described test number (TN) value n
Become all numerical value of relatively prime relation, then construct according to the equation below of the good grid point tactical deployment of troops and obtain uniform designs table:
u1j=hj
Wherein, i=1,2 ..., n-1, hjIt is designated as jth and becomes relatively prime with n
The value of relation.
Described calculating acquiring unit 22 then arbitrarily selects s row composition from the uniform designs table that described structural unit 21 obtains
New matrix；Calculate each described s according to L2 centralization deviation computational methods and arrange the deviation of new matrix, obtain deviation therein less
Matrix.Wherein, s is also that user is arranged as required to, and described s is total less than the uniform designs table of described acquiring unit 22 acquisition
Columns.
Test user can arrange the test to Meta Materials, by data analysis hands according to above-mentioned initial trial design table
Section finds the region that the change of electromagnetic response in Meta Materials to be measured is violent, and change acutely shows that quantity of information that this region comprised is relatively
Greatly.Described parameter is chosen module 3 and the value of electromagnetic response delta data is made more than the geometric parameter corresponding to the threshold value preset
For active parameter value, and according to these active parameter values, determine active geometric parameter model
Enclose.
Described second structural unit 4 is then chosen module 3 according to Latin Hypercube Sampling method and described parameter and has been determined
The geometric parameter scope structure of mastery reaction obtains EXPERIMENTAL DESIGN table.
Seeing Fig. 4, described second structural unit 4 specifically comprises the steps that computing unit 41 and acquiring unit 42.
Described computing unit 41, for according to described test number (TN) value n, uses Latin Hypercube Sampling method to calculate and obtains
Sample value；
Described acquiring unit 42, for according to playing leading work in calculated sample value and described each geometric parameter factor
Geometric parameter scope, structure obtain EXPERIMENTAL DESIGN table.
The present invention, in the EXPERIMENTAL DESIGN table acquisition process of man-made microstructure, builds uniform Design according to uniform Design principle
Table, obtains the initial trial design table that parameter to be measured is few, and after testing according to this initial trial design table, has determined main
Lead the geometric parameter scope of effect, and obtain after carrying out Latin Hypercube Sampling calculating according to the geometric parameter scope determined final
EXPERIMENTAL DESIGN table, test according to this experimental design table, contribute in the case of not affecting experimental result, reduce experiment
Operation, it is achieved draw distortionless experimental result in the experiment that quantity is few.
One of ordinary skill in the art will appreciate that all or part of flow process realizing in above-described embodiment method, be permissible
Instructing relevant hardware by computer program to complete, described program can be stored in a computer read/write memory medium
In, this program is upon execution, it may include such as the flow process of the embodiment of above-mentioned each method.Wherein, described storage medium can be magnetic
Dish, CD, read-only store-memory body (Read-Only Memory, ROM) or random store-memory body (Random Access
Memory, RAM) etc..
The above disclosed present pre-ferred embodiments that is only, can not limit the right model of the present invention with this certainly
Enclose, the equivalent variations therefore made according to the claims in the present invention, still belong to the scope that the present invention is contained.
Claims (8)
1. the EXPERIMENTAL DESIGN list acquirement method of a man-made microstructure, it is characterised in that including:
Determine the number of the geometric parameter factor of man-made microstructure to be measured, the value number of each geometric parameter factor and correspondence thereof
Parameter value；
Number, the value number of each geometric parameter factor and the parameter value of correspondence thereof according to the described geometric parameter factor, and
According to uniform Design principle structure initial trial design table；
Electromagnetic response delta data when testing according to described initial trial design table, obtains in each geometric parameter factor and rises
The geometric parameter scope of mastery reaction；
Active geometric parameter model in each geometric parameter factor according to Latin Hypercube Sampling method and described acquisition
Enclosing, structure obtains EXPERIMENTAL DESIGN table；
Described each geometric parameter factor value number and the parameter value of correspondence, according to the size of required electromagnetic material
It is determined；
Described according to active geometric parameters in each geometric parameter factor of Latin Hypercube Sampling method and described acquisition
Number scope, structure obtains EXPERIMENTAL DESIGN table and includes:
According to test number (TN) value n, use Latin Hypercube Sampling method to calculate and obtain sample value；
According to active geometric parameter scope in calculated sample value and described each geometric parameter factor, it is configured to
To EXPERIMENTAL DESIGN table；
Described according to test number (TN) value n, use Latin Hypercube Sampling method calculate obtain sample value particularly as follows:
[0,1] interval is divided into the subinterval of n non-overlapping copies, in each subinterval, carries out independent equiprobability respectively take out
Sample, obtains n the random number i.e. sample value more than 0 less than 1；
The computing formula of described random number is:
And relational expressionWherein, i=1,2 ..., n, uiValue be i-th subinterval and take out
The sample value that sample obtains, u is the random number more than 0 less than 1.
2. the method for claim 1, it is characterised in that the described number according to the described geometric parameter factor determined,
The value number of each geometric parameter factor and the parameter value of correspondence thereof, and according to the structure initial trial design of uniform Design principle
Table includes:
According to the number of the described geometric parameter factor arranged, the value number of each geometric parameter factor, and according to uniformly setting
Meter principle, makes to make good use of grid point method structure uniform designs table；
Dividing the matrix obtaining multiple s row from described uniform designs table, wherein, described s value is less than described uniform designs table
Total columns；
Multiple matrixes of described acquisition are carried out centralization deviation calculating, obtains the matrix that centralization deviation therein is minimum；
The matrix minimum according to described centralization deviation and the described parameter value determined, obtain initial trial design table.
3. method as claimed in claim 2, it is characterised in that the described number according to the described geometric parameter factor arranged,
The value number of each geometric parameter factor, and according to uniform Design principle, make to make good use of grid point method structure uniform designs table bag
Include:
Arranging test number (TN) value n, described test number (TN) value n is the least common multiple of the value number of each geometric parameter factor；
According to described test number (TN) value n, extract less than described test number (TN) value n and become relatively prime pass with described test number (TN) value n
All numerical value of system；
According to test number (TN) value n and all numerical value of described extraction, generate each row of uniform designs table according to good grid point method, obtain
Obtain uniform designs table.
4. method as claimed in claim 3, it is characterised in that the described geometric parameter factor determining man-made microstructure to be measured
The parameter value of number, the value number of each geometric parameter factor and correspondence thereof includes:
Determine the number of the geometric parameter factor of man-made microstructure to be measured, the value number of each geometric parameter factor；
Determine the span of each geometric parameter factor to determine each geometric parameter factor according to described value number
Corresponding parameter value.
5. the EXPERIMENTAL DESIGN table acquisition device of a man-made microstructure, it is characterised in that including:
Parameter determination module, for determining the number of the geometric parameter factor of man-made microstructure to be measured, each geometric parameter factor
Value number and the parameter value of correspondence；
First constructing module, for according to the number of the described geometric parameter factor, the value number of each geometric parameter factor and
The parameter value of its correspondence, and according to uniform Design principle structure initial trial design table；
Parameter chooses module, the electromagnetic response delta data in time testing according to described initial trial design table, obtains
Active geometric parameter scope in each geometric parameter factor；
Second constructing module, rises leading in each geometric parameter factor according to Latin Hypercube Sampling method and described acquisition
The geometric parameter scope of effect, structure obtains EXPERIMENTAL DESIGN table；
Described each geometric parameter factor value number and the parameter value of correspondence, according to the size of required electromagnetic material
It is determined；
Described second constructing module includes:
Computing unit, for according to test number (TN) value n, uses Latin Hypercube Sampling method to calculate and obtains sample value；
Acquiring unit, for according to active geometric parameters in calculated sample value and described each geometric parameter factor
Number scope, structure obtains EXPERIMENTAL DESIGN table；
Described computing unit, according to test number (TN) value n, uses Latin Hypercube Sampling method to calculate the process tool obtaining sample value
Body is:
[0,1] interval is divided into the subinterval of n non-overlapping copies, in each subinterval, carries out independent equiprobability respectively take out
Sample, obtains n the random number i.e. sample value more than 0 less than 1；
The computing formula of described random number is:
And relational expressionWherein, i=1,2 ..., n, uiValue be i-th subinterval and take out
The sample value that sample obtains, u is the random number more than 0 less than 1.
6. device as claimed in claim 5, it is characterised in that described first constructing module includes:
Structural unit, for the number of the described geometric parameter factor according to setting, the value number of each geometric parameter factor,
And according to uniform Design principle, make to make good use of grid point method structure uniform designs table；
Calculating acquiring unit, for dividing the matrix obtaining multiple s row from described uniform designs table, wherein, described s value is less than
Multiple matrixes of described acquisition are carried out centralization deviation calculating, obtain center therein by total columns of described uniform designs table
Change the matrix that deviation is minimum, the matrix minimum according to described centralization deviation and the described parameter value determined, obtain initial trial
Design table.
7. device as claimed in claim 6, it is characterised in that
Described structural unit be according to preset test number (TN) value n and less than described test number (TN) value n and with described test number (TN)
Value n becomes all numerical value of relatively prime relation, and generates each row of uniform designs table according to good grid point method, it is thus achieved that uniform designs table；
Wherein, described default test number (TN) value n is the minimum public affairs of the sample point number that each geometric parameter arranged is corresponding
Multiple.
8. device as claimed in claim 7, it is characterised in that described parameter determination module is additionally operable to determine each geometric parameters
The span of the number factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110396039.6A CN103136402B (en) | 2011-12-02 | 2011-12-02 | The EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110396039.6A CN103136402B (en) | 2011-12-02 | 2011-12-02 | The EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103136402A CN103136402A (en) | 2013-06-05 |
| CN103136402B true CN103136402B (en) | 2016-08-31 |
Family
ID=48496225
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110396039.6A Active CN103136402B (en) | 2011-12-02 | 2011-12-02 | The EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103136402B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106709218A (en) * | 2017-03-01 | 2017-05-24 | 中铁二院工程集团有限责任公司 | Elastic strip geometric parameter optimization method based on two times of experiment design |
-
2011
- 2011-12-02 CN CN201110396039.6A patent/CN103136402B/en active Active
Non-Patent Citations (5)
| Title |
|---|
| "Matlab工具箱在导弹试验设计中的应用";刘云南等;《弹箭与制导学报》;20070531;第27卷(第5期);第91-94页 * |
| "一种基于拉丁超立方体抽样的多目标进化算法";郑金华等;《模式识别与人工智能》;20090430;第22卷(第2期);第223-233页 * |
| "均匀设计表的MATLAB实现";曹慧荣 等;《统计与决策》;20080331(第6期);第144-146页 * |
| "基于均匀设计的产品优化变型设计";李正祥 等;《机械科学与技术》;20110630;第30卷(第6期);第892-896页 * |
| "基于进化算法改进拉丁超立方抽样的概率潮流计算";李俊芳等;《中国电机工程学报》;20110905;第31卷(第25期);第90-96页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103136402A (en) | 2013-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tian et al. | Spatial externalities in China regional economic growth | |
| Li et al. | The adequacy of different landscape metrics for various landscape patterns | |
| Peng et al. | Evaluating the effectiveness of landscape metrics in quantifying spatial patterns | |
| Alam et al. | Cosmic web dependence of galaxy clustering and quenching in SDSS | |
| CN104951580B (en) | The time domain spectral element emi analysis method of unconditional stability and condition stability mixing | |
| Raimbault et al. | Space matters: Extending sensitivity analysis to initial spatial conditions in geosimulation models | |
| Xiao et al. | Emergence of symmetry in complex networks | |
| CN102722603B (en) | Reliability measuring method for mechanical and electrical products | |
| CN102968813A (en) | Surface sampling method of triangular patch mesh model | |
| Zhang et al. | Fuzzy analysis of community detection in complex networks | |
| Dubarry et al. | Battery durability and reliability under electric utility grid operations: analysis of on-site reference tests | |
| Ruishan et al. | Land use dynamics and landscape patterns in Shanghai, Jiangsu and Zhejiang | |
| CN103136402B (en) | The EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure and device | |
| Laurie et al. | A multifractal model for spatial variation in species richness | |
| Fu et al. | Pinning control of clustered complex networks with different size | |
| CN103136401B (en) | The EXPERIMENTAL DESIGN list acquirement method of a kind of man-made microstructure and device | |
| CN104572588A (en) | Matrix inversion processing method and device | |
| CN106257506A (en) | Three layers of associating of big data quantity prediction dynamically select optimal models method | |
| CN108985493A (en) | A kind of ground class variation prediction method based on self-adapting changeable filter | |
| CN103136396B (en) | A kind of metamaterial structure unit parameter simulation selection method and device | |
| CN103793339B (en) | Data Cache performance heuristic approach based on internal storage access storehouse distance | |
| CN110489816B (en) | Analysis method for quantifying fatigue damage grade of crane boom | |
| CN103135004B (en) | A kind of method, a kind of method and device thereof measuring electromagnetic property of test card structure | |
| CN103136403B (en) | A kind of test design list acquirement method of man-made microstructure and device | |
| Garza et al. | Evaluation of two novel wake models in offshore wind farms |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |