CN109961229A - A kind of power source planning comprehensive estimation method - Google Patents
A kind of power source planning comprehensive estimation method Download PDFInfo
- Publication number
- CN109961229A CN109961229A CN201910243008.3A CN201910243008A CN109961229A CN 109961229 A CN109961229 A CN 109961229A CN 201910243008 A CN201910243008 A CN 201910243008A CN 109961229 A CN109961229 A CN 109961229A
- Authority
- CN
- China
- Prior art keywords
- index
- power source
- degree
- grade
- follows
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 239000011159 matrix material Substances 0.000 claims abstract description 59
- 238000011156 evaluation Methods 0.000 claims abstract description 26
- 238000012545 processing Methods 0.000 claims abstract description 13
- 230000008569 process Effects 0.000 claims abstract description 9
- 238000011426 transformation method Methods 0.000 claims abstract description 7
- 238000013459 approach Methods 0.000 claims abstract description 3
- 230000000052 comparative effect Effects 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 12
- 230000008030 elimination Effects 0.000 claims description 12
- 238000003379 elimination reaction Methods 0.000 claims description 12
- 230000004927 fusion Effects 0.000 claims description 9
- 230000015572 biosynthetic process Effects 0.000 claims description 6
- 238000012217 deletion Methods 0.000 claims description 6
- 230000037430 deletion Effects 0.000 claims description 6
- 238000009826 distribution Methods 0.000 claims description 6
- 238000003786 synthesis reaction Methods 0.000 claims description 6
- 230000007613 environmental effect Effects 0.000 claims description 5
- 230000002776 aggregation Effects 0.000 claims description 3
- 238000004220 aggregation Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000011161 development Methods 0.000 claims description 3
- 238000013210 evaluation model Methods 0.000 claims description 3
- 238000000926 separation method Methods 0.000 claims description 3
- 241001269238 Data Species 0.000 claims description 2
- 238000007493 shaping process Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 238000005304 joining Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of power source planning comprehensive estimation methods, belong to power source planning index system evaluation areas.The present invention obtains the index for the purpose of influencing power source planning scheme entirety, classifies according to respective Criterion Attribute during power source planning；Nondimensionalization, forward directionization processing are carried out to each achievement data using range transformation method；Grey relational grade in every class index between index is calculated using grey correlation theory, while formulating the degree of association to influence criterion；Raw performance system is handled to obtain final index system；Each index weights are calculated using analytic hierarchy process (AHP)；Formulate power source planning index system Rule set；Construct evaluations matrix；Then it is brought into using fuzzy algorithmic approach and asks to obtain final appraisal results.The method of the present invention eliminates the juxtaposition between index, and weakens the degree of association influence to overall merit generation more by force between index.
Description
Technical field
The present invention relates to power source planning index system evaluation areas, are a kind of power source planning comprehensive estimation methods.
Background technique
To alleviate environmental pollution and energy crisis, large-scale renewable energy power generation is extremely urgent, however wind-powered electricity generation, photovoltaic
Etc. renewable energy have it is intermittent and uncertain, with the increase of renewable energy grid connection capacity, the stabilization of electric system
Property, reliability etc. are by severe challenge, and to the operations of modern power systems, more stringent requirements are proposed.
Power source planning plays an important role in Power System Planning, is whether assessment electric power system power source layout is reliable and stable
Important method.And whether the power source planning scheme established is rationally effective, then needs a set of applicability and feasibility extremely strong
Assessment indicator system assesses the various aspects of power source planning scheme.
For power source planning comprehensive assessment index system, juxtaposition and association, i.e., two fingers are usually present between index
When the case where mark is reflected is essentially identical or judges an index, between the influence and each index that will receive other indexs
The degree of association is stronger, and influence is bigger, so needing to delete the index of juxtaposition, relatively by force but reports situations also to the degree of association
There are the eliminations of the index degree of being associated of other great influences.It is not right in the method for the degree of association between handling index in relation to document
The degree of association influences to carry out grade classification, and the deletion index and the elimination degree of association of blindness will cause the not perfect and information of index system
Missing.
Summary of the invention
The present invention is in place of solving the above the deficiencies in the prior art, to provide a kind of power source planning for the technical issues of solving
Comprehensive estimation method, it is intended to consider that power source planning is related to multiple attributes such as reliable, economic, environmentally friendly, to power source planning comprehensive assessment
Index system carries out layered shaping, then calculates grey relational grade using grey correlation theory, and constructs power source planning evaluation
The degree of association influences criterion, to carry out layered shaping, deletion and pass being associated extremely strong index to index system to index
The degree of association is eliminated between joining stronger index, has not only eliminated the juxtaposition between index, but also is eliminated because the degree of association is relatively strong right between index
It is influenced caused by overall merit.
The method for constructing power source planning comprehensive assessment index system, using grey correlation theory and evidence theory and is applied
Reasonable index system is improved in the method foundation of close spy's orthogonalization, and carries out overall merit on this basis.
To achieve the above object, use following technical scheme: a kind of power source planning comprehensive estimation method, feature exist
In, comprising the following steps:
Step 1, it the foundation of Raw performance system: during power source planning, obtains whole to influence power source planning scheme
For the purpose of index, classify according to respective Criterion Attribute；
Step 2, nondimensionalization, forward directionization place are carried out to achievement data each in Raw performance system using range transformation method
Reason；
Step 3, it for step 2 treated Raw performance system, is calculated in every class index using grey correlation theory
Grey relational grade between index, while formulating the degree of association influence criterion deleted according to grey relational grade index；
Step 4, criterion and calculated grey relational grade are influenced according to the degree of association in step 3, to Raw performance system
It is handled to obtain final index system；
Step 5, each index weights are calculated using analytic hierarchy process (AHP), obtains the weight sets of index；
Step 6, the power source planning index system Rule set that all kinds of achievement datas are carried out with grade classification is formulated；Using triangle
Type subordinating degree function determines degree of membership of each index at each grade, and constructs matrix as evaluations matrix using this；
Step 7, by step 4-6, weight sets, the evaluations matrix of final index system are obtained, then uses fuzzy algorithmic approach
It brings into and asks to obtain final appraisal results.
A further technical solution lies in the Criterion Attribute of the step 1 includes safety, the feature of environmental protection, reliability and warp
Ji property.
A further technical solution lies in index is carried out nondimensionalization, forward directionization processing mode by the range transformation method
It is as follows:
If sharing m scheme in overall merit, n index, each index is respectively y1, y2..., yn, use yijIt indicates i-th
J-th of original index value of scheme, i=1,2 ..., n；J=1,2 ... m；xijIndicate i-th of side by nondimensionalization processing
J-th of index value of case；
For profit evaluation model index:
For cost type index:
A further technical solution lies in calculate the mistake of the grey relational grade in every class index between index in the step 3
Journey is as follows:
(1) construction reference index and Comparative indices matrix；Using first index as reference index, remaining, which is used as, compares
Index；
Comparative indices matrix are as follows:
In formula: xijIndicate j-th of original index value of i-th of scheme；
Reference index vector are as follows:
x1n=[x11 x12 … x1n]
M is the number of Comparative indices under some 1 grade of index, and n is the feature quantity of comparator matrix；
(2) calculate the absolute difference of each Comparative indices sequence and reference sequences corresponding element one by one, i.e., | x1n-xin|(i
=1 ..., i)
Wherein xin=[xi1 xi2 … xin] (i=1 ..., m)
Obtain the absolute difference matrix between each Comparative indices and reference index:
Then it determinesWith
(3) grey incidence coefficient is calculated
Calculate separately the incidence coefficient of each relatively sequence and reference sequences:
ρ is resolution ratio in formula, and in (0,1) interior value, if ρ is smaller, difference is bigger between incidence coefficient, and separating capacity is got over
By force；
(4) grey relational grade is calculated
The mean value of the incidence coefficient of some index and reference sequences corresponding element is calculated separately, to each relatively sequence with reflection
The correlation degree of each relatively sequence and reference sequences, is denoted as grey relational grade:
After finding out the degree of association of first index and other indexs, needs reference index line number adding 1, then calculate this again
The degree of correlation of R-matrix and its comparator matrix, and so on, it need to calculate m-1 times altogether, m is this unit index number；
The correlation matrix between each index is obtained after calculating:
In formula: rijIndicate the grey relational grade between index i and index j.
A further technical solution lies in the degree of association influences criterion are as follows: according to the size of grey relational grade to shadow
Index for the purpose of sound power source planning scheme entirety carries out influence grade classification, carries out two each other to index according to grade is influenced
The correlation of a deletion and two each other between strong correlation index of selecting of extremely strong correlation metric is eliminated.
A further technical solution lies in detailed process is as follows for the step 4:
Grey relational grade influences Rule set in the degree of association in the rate range of degree of association influence Rule set between judging two indexes
The middle grade threshold for setting adjustable rate range, successively judges the size of this grey relational grade Yu each grade threshold, if greatly
In, then with evidence theory carry out fiducial probability compared with insincere probability, if insincere probability greatly if adjust this threshold value, if can
Believe that probability then deletes greatly an index, retain an index or carries out the elimination of correlation；If being less than, it is not processed；Successively
This grey relational grade is compared with each grade threshold, processing result is integrated into final index system；
Wherein, fiducial probability is calculated using evidence theory and insincere probability is as follows:
(1) the distance d (m between evidence body is calculatedi,mj)；
In formula: n indicates that corroboration body sum, i.e., the basic trust degree that expert provides distribute sum,
M indicates focus element sum；
miIndicate the basic trust degree distribution of i-th of corroboration body；
mjIndicate the basic trust degree distribution of j-th of corroboration body；
BkIndicate k-th of focus element；
(2) support of evidence body is calculated；
Similarity between evidence body are as follows:
sim(mi,mj)=1-d (mi,mj)
Evidence body miSupport are as follows:
(3) evidence body confidence level Dcrd (m is calculatedi)；
(4) data fusion is carried out to evidence body according to evidence, with evidence body m1、m2For, calculation formula is such as
Under:
Elimination using Schimidt orthogonalization degree of being associated is as follows:
Enable Z1=X1
Z2=X2-k21X1Wherein k21For undetermined constant, meet cov (Z2,Z1)=0；
I.e.
cov(X2-k21X1,X1)=cov (X2,X1)-k21cov(X1,X1)=0
?
Wherein:
cov(X2,X1)=E (X1X2)-E(X1)E(X2)
E(X1)、E(X2)、E(X1X2) indicate expectation, the desired value of index is indicated using the mean value of n group sample here, i.e.,
Wherein X1(i)、X2(i) index X is respectively indicated1、X2Numerical value in i-th group of sample；
Z is enabled again3=X3-k32X2-k31X1
Wherein k32、k31For undetermined constant, meet cov (X3,X2)=cov (X3,X1)=0,
?
Z can similarly be obtained4,Z5,…Zn, then the index for having carried out correlation elimination is not needed to locate with what is filtered out
The aggregation of reason obtains final System of Comprehensive Evaluation, then uniformly uses xijIndicate that i scheme j refers to target value.
A further technical solution lies in detailed process is as follows for the step 5:
(1) development of judgment matrix
The relative importance judgement that each index is carried out according to following table, constructs fuzzy judgment matrix AJ, wherein element meets
aii=0.5 ... i=1,2 ... n, aij+aji=1, j=1,2 ... n, then its judgment matrix be
Importance scale criterion table
| Scale numbering | Meaning |
| 1 | Index i and index j no less important |
| 3 | Index i is slightly more important than index j |
| 5 | Index i is obviously more important than index j |
| 7 | Index i is more extremely important than index j |
| 9 | Index i is more of crucial importance than index j |
| 2,4,6,8 | The median of above-mentioned adjacent scale |
| The inverse of above each number | Inverse ratio is compared with aji=1/aij |
(2) weight calculation
Weight calculation formula is
In formula, n is the total number of same level index
Then the weight sets W of index is obtained；
(3) consistency check
Relative uniformity examine formula be
In formula: CI is to calculate coincident indicator；RI is random index, and the order of value and judgment matrix is related such as
Following table；λmaxFor the Maximum characteristic root of judgment matrix；
Aver-age Random Consistency Index RI
| n | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
| RI | 0.58 | 0.90 | 1.12 | 1.24 | 1.32 | 1.41 | 1.45 |
If CR < 0.1, then it is assumed that judgment matrix is feasible, and consistency check passes through, and general CR value is the smaller the better；If CR >=
0.1, then not over consistency check, the first step is returned, compares again, constructs suitable judgment matrix.
A further technical solution lies in the triangular form subordinating degree function includes the following two kinds type:
(1) the smaller the better type triangle π membership function model
Membership function model are as follows:
(2) type that is the bigger the better triangle π membership function model
Membership function model are as follows:
A further technical solution lies in calculate power source planning index Rule set:
The multi-group data of each index in the programme planning phase is determined first；Then, each index in this several groups of data is calculated
Average value；The Number synthesis averaged together for being greater than average value, and using it as the maximum data boundary of Rule set；Equally
Principle is it is found that the Number synthesis averaged together for being less than average value, and using it as the smallest data boundary of Rule set；Most
Three equidistant branches are inserted between the two data boundaries afterwards, it is determined that outstanding, good, preferable, general, poor five etc.
The value range of grade, and then obtain Rule set.
A further technical solution lies in Calculation Estimation matrixes: specific steps in the step 7 are as follows:
After weight sets W, the evaluations matrix H that each index has been determined, final appraisal results S uses fuzzy composition algorithm such as
Under:
S=W ο H
ο is fuzzy operator in formula；Evaluation knot is determined according to maximum membership grade principle after finding out final comprehensive evaluation result
The affiliated grade of fruit, according to final appraisal results and Rule set grade separation numerical procedure final score.
Compared with prior art, the method for the present invention has the advantages that
1. establishing the degree of association in power source planning overall merit influences criterion, influence grade classification is carried out on the degree of association；
2. first calculate grey relational grade, filter out the index with the extremely strong degree of association carry out evidence fusion adjustment threshold value and
Index is deleted, and is filtered out the index with the stronger degree of association and is carried out the degree of association between evidence fusion adjustment threshold value and index and disappears
Remove, not only eliminated the juxtaposition between index, but eliminate because between index the degree of association more by force influence caused by overall merit.
Detailed description of the invention
The flow diagram of Fig. 1 the method for the present invention.
The structural block diagram of the power source planning evaluation index layered system of Fig. 2 the method for the present invention.
The final structure block diagram of the power source planning evaluation index layered system of Fig. 3 the method for the present invention.
Specific embodiment
The present invention will be further described with reference to the accompanying drawing:
Embodiment 1
As shown in Figure 1, disclosure sets forth a kind of power source planning comprehensive estimation methods, now to a cost minimization electricity
Source programme carries out index system foundation and overall merit, specific embodiment are as follows:
1. comprehensive assessment index layered shaping
Power source planning fully considers the safety for considering power source planning, the feature of environmental protection, reliability, economy etc. to planning
The influence of scheme entirety carries out layered shaping to index, and establishing includes decision objective layer, decision attribute layer, decision object layer three
The integral framework of a level, obtained index system (only list safety, the feature of environmental protection, reliability, warp as shown in Figure 2 in figure
Four Criterion Attributes of Ji property).
2. indices non-dimension is handled
Stochastic Production Simulation is carried out according to the requirement of power source planning status and programme and obtains each achievement data, then
It is using range transformation method that index progress nondimensionalization processing is as follows:
It is using range transformation method that index progress nondimensionalization, forward directionization processing is as follows:
If sharing m scheme in overall merit, n index, each index is respectively y1, y2..., yn, use yij(i=1,
2 ..., n；J=1,2 ... m) indicate i-th of original index value of i-th of scheme, xijIndicate i-th by nondimensionalization processing
J-th of index value of a scheme.
For profit evaluation model index:
For cost type index:
Each index of the standardization of table 1 front and back calculates data
3. calculating grey relational grade matrix:
(1) construction reference index and Comparative indices matrix.First index is first reference index by the present embodiment, remaining
For Comparative indices.
Comparative indices matrix are as follows:
In formula: XijIndicate j-th of original index value of i-th of scheme.
Reference index vector are as follows:
x1n=[x11 x12 … x1n] (4)
M is the number of Comparative indices under some 1 grade of index, and n is the feature quantity of comparator matrix.
(2) calculate the absolute difference of each Comparative indices sequence and reference sequences corresponding element one by one, i.e., | x1n-xin|(i
=1 ..., m)
Wherein xin=[xi1 xi2 … xin] (i=1 ..., M)
Obtain the absolute difference matrix between each Comparative indices and reference index:
Then it determinesWith
(3) calculate correlation coefficient
Calculate separately the incidence coefficient of each relatively sequence and reference sequences:
ρ is resolution ratio in formula, and in (0,1) interior value, if ρ is smaller, difference is bigger between incidence coefficient, and separating capacity is got over
By force.Usual ρ takes 0.5.
(4) calculating correlation
The mean value of the incidence coefficient of some index and reference sequences corresponding element is calculated separately, to each relatively sequence with reflection
The correlation degree of each relatively sequence and reference sequences, is denoted as:
After finding out the degree of association of first index and other indexs, needs reference index line number adding 1, then calculate this again
The degree of correlation of R-matrix and its comparator matrix, and so on, it need to calculate m-1 times altogether, m is this unit index number.
The correlation matrix between each index is obtained after calculating:
In formula: rijIndicate the grey relational grade between index i and index j.
The grey relational grade of each index of 2 safety of table
The grey relational grade of each index of 3 reliability of table
The grey relational grade of each index of 4 economy of table
The grey relational grade of each index of 5 feature of environmental protection of table
4. the degree of association for establishing power source planning overall merit influences criterion:
The pass of power source planning overall merit is established according to degree of association influence degree between each index of power source planning index system
Connection degree influences grade (substantially without influence, there is weaker influence, there is stronger influence, there is extremely strong influence):
The 6 power source planning index system degree of association of table influences grade classification
Wherein: r1It indicates weaker influence and has the degree of association threshold value between two grades of stronger influence, r2Indicate stronger shadow
Ring and have the degree of association threshold value between two grades of extremely strong influence.
Tentatively selected r in the example1It is 0.7, r2For 0.85.
5. the screening and the degree of association that carry out between index are eliminated:
Whether the degree of association is greater than threshold value r between first determining whether two indexes2, if more than then being carried out with evidence theory credible general
Rate compared with insincere probability, if insincere probability greatly if adjust threshold value r2, if fiducial probability greatly if delete an index, retain
One index, is integrated into the index system changed for the first time；If being less than, the index system directly integrating to change for the first time.
Then whether the degree of association is greater than threshold value r between judging two indexes1, if more than then being carried out with evidence theory credible general
Rate compared with insincere probability, if insincere probability greatly if adjust threshold value r1, if fiducial probability greatly if use Schimidt orthogonalization
The degree of association is eliminated between method carries out index, is integrated into final index system；If both less than r1, then it is integrated directly into final index
System.
Fiducial probability is calculated using evidence theory and insincere probability is as follows:
(1) the distance d (m between evidence body is calculatedi,mj)。
In formula: n indicates that corroboration body sum, i.e., the basic trust degree that expert provides distribute sum, n=in the present embodiment
4；
M indicates that focus element sum, m=2 in the present embodiment, focus element are respectively confidence level and can not reliabilities；
miIndicate the basic trust degree distribution of i-th of corroboration body；
mjIndicate the basic trust degree distribution of j-th of corroboration body；
BkIndicate k-th of focus element.
(2) support of evidence body is calculated.
Similarity between evidence body are as follows:
sim(mi,mj)=1-d (mi,mj) (10)
Evidence body MiSupport are as follows:
(3) evidence body confidence level Dcrd (m is calculatedi)。
(4) data fusion is carried out to evidence body according to evidence, with evidence body m1、m2For, calculation formula is such as
Under:
Elimination using Schimidt orthogonalization degree of being associated is as follows:
Enable Z1=X1
Z2=X2-k21X1Wherein k21For undetermined constant, meet cov (Z2,Z1)=0.
I.e.
cov(X2-k21X1,X1)=cov (X2,X1)-k21cov(X1,X1)=0 (17)
?
Wherein:
cov(X2,X1)=E (X1X2)-E(X1)E(X2) (19)
E(X1)、E(X2)、E(X1X2) indicate expectation, the desired value of index is indicated using the mean value of n group sample here, i.e.,
Wherein X1(i)、X2(i) index X is respectively indicated1、X2Numerical value in i-th group of sample.
Z is enabled again3=X3-k32X2-k31X1
Wherein k32、k31For undetermined constant, meet cov (X3,X2)=cov (X3,X1)=0,
?
Z can similarly be obtained4,Z5,…Zn, then the index for having carried out correlation elimination is not needed to locate with what is filtered out
The aggregation of reason obtains final System of Comprehensive Evaluation, then uniformly uses xijIndicate that i scheme j refers to target value.
It is greater than threshold value r in table 22It is X12-X13, the degree of association 0.88, relevance is stronger between showing index, needs to use
Evidence theory decides whether to screen index, in the present embodiment by 4 experts provide two indices needs screen it is (credible
Probability, insincere probability) apportioning cost are as follows: (m1, m2, m3, m4)=((0.8,0.2), (0.5,0.5), (0.4,0.6), (0.3,
0.7)), the expert's fiducial probability screened is needed to merge index, result is (m1-2, m1-3, m1-4)=((0.691,
0.309), (0.562,0.438), (0.397,0.603)).
Wherein, m1-2、m1-3、m1-4Respectively indicate the fusion value of first 2, the fiducial probability of preceding 3 and preceding 4 experts.4
The fiducial probability fusion value of expert are as follows: fiducial probability 0.397 is less than insincere probability 0.603, illustrates that the event is insincere, i.e., and two
A index does not need to screen.Therefore to threshold value r2Being adjusted is 0.88, and two indexes satisfaction does not screen requirement.
It is greater than threshold value r in table 32It is X21-X22, the degree of association 0.95, X25-X26, the degree of association 0.89, to X25-
X26It is (m that fiducial probability, which carries out fusion results,1-2, m1-3, m1-4)=((0.716,0.284), (0.515,0.485), (0.660,
0.340)).Illustrate that the event is credible, i.e., two indices need screening and final threshold value r2It is 0.88, retains more excellent index,
Therefore delete index X22, X26；It is greater than threshold value r in table 42It is X31-X32, the degree of association 0.92, deletion index X32；It is big in table 5
In threshold value r2It is X42-X45, the degree of association 0.91, deletion index X45；Then the index system after screening is obtained as schemed
Shown in 3.
It is greater than threshold value r in table 21It is X11-X13, the degree of association 0.70, X14-X15, the degree of association 0.79, to X11-
X13It is (m that fiducial probability, which carries out fusion results,1-2, m1-3, m1-4)=((0.691,0.309), (0.485,0.515), (0.620,
0.380)).Illustrate that the event is credible, i.e., two indices need degree of being associated elimination and final threshold value r1It is 0.70, X14-
X15It is also required to degree of being associated elimination；X in table 321-X23、X23-X24Degree of being associated is needed to eliminate；X in table 431-X34Need into
The row degree of association is eliminated；X in table 543-X44Degree of being associated is needed to eliminate.
Then it is as shown in table 7 Schimidt orthogonalization processing final result to be carried out to the index for needing degree of being associated to eliminate:
7 degree of association of table eliminates front and back data comparison
6. parameter weight sets W and evaluations matrix H:
Analytic hierarchy process (AHP) parameter weight sets:
(1) development of judgment matrix
Relative importance according to each index of table 2 judges, constructs fuzzy judgment matrix AJ, wherein element meets aii=
0.5 ... i=1,2 ... n, aij+aji=1, j=1,2 ... n, then its judgment matrix be
8 importance scale criterion table of table
| Scale numbering | Meaning |
| 1 | Index i and index j no less important |
| 3 | Index i is slightly more important than index j |
| 5 | Index i is obviously more important than index j |
| 7 | Index i is more extremely important than index j |
| 9 | Index i is more of crucial importance than index j |
| 2,4,6,8 | The median of above-mentioned adjacent scale |
| The inverse of above each number | Inverse ratio is compared with aji=1/aij |
(2) weight calculation
Weight calculation formula is
In formula, n is the total number of same level index
Then the weight sets W of index is obtained.
(3) consistency check
Relative uniformity examine formula be
9 Aver-age Random Consistency Index RI of table
| n | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
| RI | 0.58 | 0.90 | 1.12 | 1.24 | 1.32 | 1.41 | 1.45 |
In formula: CI is to calculate coincident indicator；RI is random index, and the order of value and judgment matrix is related such as
Table 3；λmaxFor the Maximum characteristic root of judgment matrix.
If CR < 0.1, then it is assumed that judgment matrix is feasible, and consistency check passes through, and general CR value is the smaller the better.If CR >=
0.1, then not over consistency check, the first step is retracted, the expert that please give a mark compares again, constructs suitable judgment matrix.
The weight sets W established in the present embodiment is as shown in table 10:
The weight sets W of 10 index of table
Triangular form subordinating degree function:
(1) the smaller the better type triangle π membership function model
Membership function model are as follows:
(2) type that is the bigger the better triangle π membership function model
Membership function model are as follows:
Calculate power source planning index Rule set:
The multi-group data of each index in the programme planning phase is determined first；Then, each index in this several groups of data is calculated
Average value；The Number synthesis averaged together for being greater than average value, and using it as the maximum data boundary of Rule set；Equally
Principle is it is found that the Number synthesis averaged together for being less than average value, and using it as the smallest data boundary of Rule set；Most
Three equidistant branches are inserted between the two data boundaries afterwards, it is determined that outstanding, good, preferable, general, poor five etc.
The value range of grade, and then it is as shown in table 11 to obtain Rule set:
11 power source planning index system Rule set of table
Calculation Estimation matrix:
Evaluations matrix is sought according to Rule set and subordinating degree function:
Wherein: m indicates index number；N indicates number of degrees, hijIndicate degree of membership of the index i at grade j.
It is H without the evaluations matrix that the degree of association is eliminated1:
It is H by the evaluations matrix that the degree of association is eliminated2:
7. calculating final appraisal results:
After weight sets W, the evaluations matrix H that each index has been determined, final appraisal results S uses fuzzy composition algorithm such as
Under:
S=W ο H (35)
ο is fuzzy operator in formula.Evaluation knot is determined according to maximum membership grade principle after finding out final comprehensive evaluation result
The affiliated grade of fruit, according to final appraisal results and Rule set grade separation numerical procedure final score.
Show that final result is as shown in table 12 by fuzzy operation:
12 power supply overall merit final result of table
By comparing as can be seen that two schemes belong to preferable grade, final score is when eliminating without the degree of association
73.86 points, when being eliminated by the degree of association final score be 68.86 points, this meet eliminate the degree of association after because partial data numerical value drop
Score variation after low.Also show the reasonability of present invention building index system method.
Embodiment described above only describe the preferred embodiments of the invention, not to model of the invention
It encloses and is defined, without departing from the spirit of the design of the present invention, those of ordinary skill in the art are to technical side of the invention
The various changes and improvements that case is made should all be fallen into the protection scope that claims of the present invention determines.
Claims (10)
1. a kind of power source planning comprehensive estimation method, which comprises the following steps:
Step 1, it the foundation of Raw performance system: during power source planning, obtains to influence power source planning scheme generally mesh
Index, classify according to respective Criterion Attribute；
Step 2, nondimensionalization, forward directionization processing are carried out to achievement data each in Raw performance system using range transformation method；
Step 3, for step 2 treated Raw performance system, index in every class index is calculated using grey correlation theory
Between grey relational grade, while formulating the degree of association deleted according to grey relational grade on index and influence criterion；
Step 4, criterion and calculated grey relational grade are influenced according to the degree of association in step 3, Raw performance system is carried out
Processing obtains final index system；
Step 5, each index weights are calculated using analytic hierarchy process (AHP), obtains the weight sets of index；
Step 6, the power source planning index system Rule set that all kinds of achievement datas are carried out with grade classification is formulated；It is subordinate to using triangular form
Category degree function determines degree of membership of each index at each grade, and constructs matrix as evaluations matrix using this；
Step 7, by step 4-6, weight sets, the evaluations matrix of final index system is obtained, is then brought into using fuzzy algorithmic approach
It asks to obtain final appraisal results.
2. a kind of power source planning comprehensive estimation method according to claim 1, which is characterized in that the index of the step 1
Attribute includes safety, the feature of environmental protection, reliability and economy.
3. a kind of power source planning comprehensive estimation method according to claim 1, which is characterized in that the range transformation method will
It is as follows that index carries out nondimensionalization, forward directionization processing mode:
If sharing m scheme in overall merit, n index, each index is respectively y1, y2..., yn, use yijIndicate i-th of scheme
J-th of original index value, i=1,2 ..., n；J=1,2 ... m；xijIndicate i-th of scheme by nondimensionalization processing
J-th of index value；
For profit evaluation model index:
For cost type index:
4. a kind of power source planning comprehensive estimation method according to claim 1, which is characterized in that calculated in the step 3
The process of grey relational grade in every class index between index is as follows:
(1) construction reference index and Comparative indices matrix；Using first index as reference index, remaining, which is used as, compares finger
Mark；
Comparative indices matrix are as follows:
In formula: xijIndicate j-th of original index value of i-th of scheme；
Reference index vector are as follows:
x1n=[x11 x12 … x1n]
M is the number of Comparative indices under some 1 grade of index, and n is the feature quantity of comparator matrix；
(2) calculate the absolute difference of each Comparative indices sequence and reference sequences corresponding element one by one, i.e., | x1n-xin| (i=
1 ..., i)
Wherein xin=[xi1 xi2 … xin] (i=1 ..., m)
Obtain the absolute difference matrix between each Comparative indices and reference index:
Then it determinesWith
(3) grey incidence coefficient is calculated
Calculate separately the incidence coefficient of each relatively sequence and reference sequences:
ρ is resolution ratio in formula, and in (0,1) interior value, if ρ is smaller, difference is bigger between incidence coefficient, and separating capacity is stronger；
(4) grey relational grade is calculated
The mean value of the incidence coefficient of some index and reference sequences corresponding element is calculated separately, to each relatively sequence to reflect each ratio
Compared with the correlation degree of sequence and reference sequences, it is denoted as grey relational grade:
After finding out the degree of association of first index and other indexs, needs reference index line number adding 1, then calculate the reference again
The degree of correlation of matrix and its comparator matrix, and so on, it need to calculate m-1 times altogether, m is this unit index number；
The correlation matrix between each index is obtained after calculating:
In formula: rijIndicate the grey relational grade between index i and index j.
5. a kind of power source planning comprehensive estimation method according to claim 1, which is characterized in that the degree of association influences quasi-
Then are as follows: the index for the purpose of influencing power source planning scheme entirety is carried out according to the size of grey relational grade to influence grade stroke
Point, according to influence grade to index carry out two each other extremely strong correlation metric select the strong correlation each other of a deletion and two
Property index between correlation eliminate.
6. a kind of power source planning comprehensive estimation method according to claim 1, which is characterized in that the step 4 it is specific
Process is as follows:
Grey relational grade influences to set in Rule set in the rate range of degree of association influence Rule set in the degree of association between judging two indexes
Surely the grade threshold of adjustable rate range, judges the size of this grey relational grade Yu each grade threshold, successively if more than then
With evidence theory carry out fiducial probability with insincere probability compared with, if insincere probability greatly if adjust this threshold value, if it is credible generally
Rate then deletes greatly an index, retains an index or carries out the elimination of correlation；If being less than, it is not processed；Successively by this
Grey relational grade is compared with each grade threshold, and processing result is integrated into final index system；
Wherein, fiducial probability is calculated using evidence theory and insincere probability is as follows:
(1) the distance d (m between evidence body is calculatedi,mj)；
In formula: n indicates that corroboration body sum, i.e., the basic trust degree that expert provides distribute sum,
M indicates focus element sum；
miIndicate the basic trust degree distribution of i-th of corroboration body；
mjIndicate the basic trust degree distribution of j-th of corroboration body；
BkIndicate k-th of focus element；
(2) support of evidence body is calculated；
Similarity between evidence body are as follows:
sim(mi,mj)=1-d (mi,mj)
Evidence body miSupport are as follows:
(3) evidence body confidence level Dcrd (m is calculatedi)；
(4) data fusion is carried out to evidence body according to evidence, with evidence body m1、m2For, calculation formula is as follows:
Elimination using Schimidt orthogonalization degree of being associated is as follows:
Enable Z1=X1
Z2=X2-k21X1Wherein k21For undetermined constant, meet cov (Z2,Z1)=0；
I.e.
cov(X2-k21X1,X1)=cov (X2,X1)-k21cov(X1,X1)=0
?
Wherein:
cov(X2,X1)=E (X1X2)-E(X1)E(X2)
E(X1)、E(X2)、E(X1X2) indicate expectation, the desired value of index is indicated using the mean value of n group sample here, i.e.,
Wherein X1(i)、X2(i) index X is respectively indicated1、X2Numerical value in i-th group of sample；
Z is enabled again3=X3-k32X2-k31X1
Wherein k32、k31For undetermined constant, meet cov (X3,X2)=cov (X3,X1)=0,
?
Z can similarly be obtained4,Z5,…Zn, then the index for having carried out correlation elimination is not needed to handle with what is filtered out
Aggregation obtains final System of Comprehensive Evaluation, then uniformly uses xijIndicate that i scheme j refers to target value.
7. a kind of power source planning comprehensive estimation method according to claim 1, which is characterized in that the step 5 it is specific
Process is as follows:
(1) development of judgment matrix
The relative importance judgement that each index is carried out according to following table, constructs fuzzy judgment matrix AJ, wherein element meets aii=
0.5 ... i=1,2 ... n, aij+aji=1, j=1,2 ... n, then its judgment matrix be
Importance scale criterion table
(2) weight calculation
Weight calculation formula is
In formula, n is the total number of same level index；
Then the weight sets W of index is obtained；
(3) consistency check
Relative uniformity examine formula be
In formula: CI is to calculate coincident indicator；RI is random index, and the order of value and judgment matrix is related as follows
Table；λmaxFor the Maximum characteristic root of judgment matrix；
Aver-age Random Consistency Index RI
If CR < 0.1, then it is assumed that judgment matrix is feasible, and consistency check passes through, and general CR value is the smaller the better；If CR >=0.1,
Not over consistency check, the first step is returned, compares again, constructs suitable judgment matrix.
8. a kind of power source planning comprehensive estimation method according to claim 1, which is characterized in that the triangular form degree of membership
Function includes the following two kinds type:
(1) the smaller the better type triangle π membership function model
Membership function model are as follows:
(2) type that is the bigger the better triangle π membership function model
Membership function model are as follows:
9. a kind of power source planning comprehensive estimation method according to claim 1, which is characterized in that calculate power source planning index
Rule set:
The multi-group data of each index in the programme planning phase is determined first；Then, being averaged for each index in this several groups of data is calculated
Value；The Number synthesis averaged together for being greater than average value, and using it as the maximum data boundary of Rule set；Same principle
It is found that the Number synthesis averaged together for being less than average value, and using it as the smallest data boundary of Rule set；Finally exist
Three equidistant branches are inserted between the two data boundaries, it is determined that outstanding, good, preferable, general, poor five grades
Value range, and then obtain Rule set.
10. a kind of power source planning comprehensive estimation method according to claim 1, which is characterized in that Calculation Estimation matrix: institute
State specific steps in step 7 are as follows:
After weight sets W, the evaluations matrix H that each index has been determined, final appraisal results S is as follows using fuzzy composition algorithm:
In formulaFor fuzzy operator；Evaluation result institute is determined according to maximum membership grade principle after finding out final comprehensive evaluation result
Belong to grade, according to final appraisal results and Rule set grade separation numerical procedure final score.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910243008.3A CN109961229A (en) | 2019-03-28 | 2019-03-28 | A kind of power source planning comprehensive estimation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910243008.3A CN109961229A (en) | 2019-03-28 | 2019-03-28 | A kind of power source planning comprehensive estimation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109961229A true CN109961229A (en) | 2019-07-02 |
Family
ID=67025134
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910243008.3A Pending CN109961229A (en) | 2019-03-28 | 2019-03-28 | A kind of power source planning comprehensive estimation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109961229A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889207A (en) * | 2019-11-08 | 2020-03-17 | 江南机电设计研究所 | System combination model credibility intelligent evaluation method based on deep learning |
| CN111861271A (en) * | 2020-07-31 | 2020-10-30 | 中国石油化工股份有限公司 | Method for evaluating heat preservation performance of pipeline |
| CN112633631A (en) * | 2020-11-26 | 2021-04-09 | 上海交通大学 | Method for evaluating complementarity of multi-power-supply system |
| CN112766628A (en) * | 2020-08-26 | 2021-05-07 | 中铁电气工业有限公司 | Method for comprehensively evaluating health state of prefabricated substation |
| CN114358568A (en) * | 2021-12-29 | 2022-04-15 | 国网天津市电力公司 | Electric energy substitution benefit evaluation method based on subentry measurement |
| CN116862292A (en) * | 2023-06-26 | 2023-10-10 | 同济大学 | Water-energy-carbon association analysis method, system, equipment and medium |
-
2019
- 2019-03-28 CN CN201910243008.3A patent/CN109961229A/en active Pending
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889207A (en) * | 2019-11-08 | 2020-03-17 | 江南机电设计研究所 | System combination model credibility intelligent evaluation method based on deep learning |
| CN110889207B (en) * | 2019-11-08 | 2023-08-22 | 江南机电设计研究所 | Deep learning-based intelligent assessment method for credibility of system combination model |
| CN111861271A (en) * | 2020-07-31 | 2020-10-30 | 中国石油化工股份有限公司 | Method for evaluating heat preservation performance of pipeline |
| CN111861271B (en) * | 2020-07-31 | 2024-04-12 | 中国石油化工股份有限公司 | Pipeline heat insulation performance evaluation method |
| CN112766628A (en) * | 2020-08-26 | 2021-05-07 | 中铁电气工业有限公司 | Method for comprehensively evaluating health state of prefabricated substation |
| CN112633631A (en) * | 2020-11-26 | 2021-04-09 | 上海交通大学 | Method for evaluating complementarity of multi-power-supply system |
| CN112633631B (en) * | 2020-11-26 | 2023-03-24 | 上海交通大学 | Method for evaluating complementarity of multi-power-supply system |
| CN114358568A (en) * | 2021-12-29 | 2022-04-15 | 国网天津市电力公司 | Electric energy substitution benefit evaluation method based on subentry measurement |
| CN116862292A (en) * | 2023-06-26 | 2023-10-10 | 同济大学 | Water-energy-carbon association analysis method, system, equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109961229A (en) | A kind of power source planning comprehensive estimation method | |
| Nandurge et al. | Analyzing road accident data using machine learning paradigms | |
| CN111476435B (en) | Charging pile load prediction method based on density peak value | |
| CN106131862B (en) | Optimization covering method based on multi-objective Evolutionary Algorithm in a kind of wireless sensor network | |
| CN102054002A (en) | Method and device for generating decision tree in data mining system | |
| Chang et al. | Applying a modified VIKOR method to classify land subdivisions according to watershed vulnerability | |
| CN108550077A (en) | A kind of individual credit risk appraisal procedure and assessment system towards extensive non-equilibrium collage-credit data | |
| CN106845642A (en) | A kind of self adaptation multi-target evolution method of belt restraining cloud workflow schedule | |
| CN104809658A (en) | Method for rapidly analyzing low-voltage distributing network area line loss | |
| CN105335902A (en) | Reliability determining method and device for electric power communication net | |
| Swindiarto et al. | Integration of fuzzy C-means clustering and TOPSIS (FCM-TOPSIS) with silhouette analysis for multi criteria parameter data | |
| CN107391515A (en) | Power system index analysis method based on Association Rule Analysis | |
| KR102268676B1 (en) | 3D Point Cloud Generative Adversarial Network Based on Tree Structured Graph Convolutions | |
| CN104217015A (en) | Hierarchical clustering method based on mutual shared nearest neighbors | |
| CN110363230A (en) | Stacking integrated sewage handling failure diagnostic method based on weighting base classifier | |
| CN108573274A (en) | A kind of selective clustering ensemble method based on data stability | |
| CN108171432B (en) | Ecological risk evaluation method based on multi-dimensional cloud model-fuzzy support vector machine | |
| CN109034232A (en) | The automation output system and control method of urban planning condition verification achievement Report | |
| CN106447025B (en) | Testability index distribution based on discrete particle cluster chooses integrated processes with test | |
| CN106021386A (en) | Theta-join method for massive distributed data | |
| CN104794234A (en) | Data processing method and device for benchmarking | |
| CN116933139A (en) | Village classification method integrating membership value, grade and sequence | |
| CN114781703A (en) | Hierarchical multi-objective optimization method, terminal equipment and storage medium | |
| CN114330526B (en) | Traffic flow missing data restoration method based on TGO-SSA-FCM algorithm | |
| CN110458722A (en) | Flood interval prediction method based on multiple target random vector function connection network |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190702 |
|
| RJ01 | Rejection of invention patent application after publication |