CN103886393A - Power grid investment optimization method based on simulation investment benefit analysis and learning automatons - Google Patents
Power grid investment optimization method based on simulation investment benefit analysis and learning automatons Download PDFInfo
- Publication number
- CN103886393A CN103886393A CN201410132514.2A CN201410132514A CN103886393A CN 103886393 A CN103886393 A CN 103886393A CN 201410132514 A CN201410132514 A CN 201410132514A CN 103886393 A CN103886393 A CN 103886393A
- Authority
- CN
- China
- Prior art keywords
- investment
- index
- evaluation index
- simulation
- evaluation
- 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 65
- 238000004458 analytical method Methods 0.000 title claims abstract description 31
- 238000004088 simulation Methods 0.000 title claims abstract description 31
- 238000005457 optimization Methods 0.000 title claims abstract description 19
- 230000008901 benefit Effects 0.000 title abstract description 9
- 238000011161 development Methods 0.000 claims abstract description 12
- 238000011156 evaluation Methods 0.000 claims description 80
- 230000006870 function Effects 0.000 claims description 16
- 239000011159 matrix material Substances 0.000 claims description 15
- 230000004044 response Effects 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 10
- 230000009471 action Effects 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000013459 approach Methods 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 238000012067 mathematical method Methods 0.000 claims description 3
- 238000010189 synthetic method Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000000694 effects Effects 0.000 abstract description 2
- 238000012384 transportation and delivery Methods 0.000 description 9
- 230000005611 electricity Effects 0.000 description 8
- 230000008569 process Effects 0.000 description 6
- 230000009466 transformation Effects 0.000 description 5
- 238000011160 research Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000002349 favourable effect Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000003556 assay Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The invention relates to the technical field of electric power system planning, in particular to a power grid investment optimization method based on simulation investment benefit analysis and learning automatons. The method comprises the following steps that first, a grey prediction model is used for conducting data prediction on power grid development data; second, related indexes in an investment benefit index system are calculated according to prediction schemes of various obtained investment data through data prediction for the coming year, and then the simulation investment benefit analysis is carried out to obtain simulation investment comprehensive score results; third, the score results obtained through different schemes are optimized based on a multi-scheme optimization strategy of a learning automaton method, and finally the optimal investment scheme is obtained. The power grid investment optimization method based on the simulation investment benefit analysis and the learning automatons is beneficial to optimizing power grid investment and good in using effect.
Description
Technical field
The present invention relates to Power System Planning technical field, particularly a kind of electric grid investment optimization method based on simulation investment performance analysis and learning automaton.
Background technology
Be accompanied by the stable development of Chinese national economy, the raising year by year of living standards of the people, whole society's need for electricity is growing, the scale of investment of electrical network also continues to increase, the electric grid investment the Eleventh Five-Year Plan period only completing in State Grid Corporation of China's service area just accumulative total has exceeded 1.2 trillion yuan, and during " 12 ", the electric grid investment of State Grid Corporation of China's planning has more reached 1.7 unprecedented trillion yuan.Electrical network is as the important infrastructure that ensures national economy and social sustainable development, and the requirement of its development has determined to need corresponding investment, but company is as operation enterprise, also needs to take into full account the benefit of investment.
The following investment decision of electric grid investment performance analysis and electrical network is two important research contents of electric grid investment.For investigating electric grid investment benefit over the years the analysis of making an appraisal, can set up suitable assessment indicator system, adopt returns of investment comprehensive grade model, obtain and utilize weight index line to calculate the method for trying to achieve comprehensive grading.Electric grid investment benefit analysis methods is widely used at present.And aspect the research of following electric grid investment, for the current situation of electrical network to be studied, how the current needs that can't meet actual electric network investment decision analysis of achievement in research of appropriate investment decision scheme are proposed.For this reason, new method need to be proposed, for electric grid investment funds flow provides favourable reference frame.
Summary of the invention
The object of the present invention is to provide a kind of electric grid investment optimization method based on simulation investment performance analysis and learning automaton, the method is conducive to electric grid investment to be optimized, and result of use is good.
For achieving the above object, technical scheme of the present invention is: a kind of electric grid investment optimization method based on simulation investment performance analysis and learning automaton, comprises the following steps:
Step 1: adopt grey forecasting model, power network development class data are carried out to data prediction;
Step 2: the power network development class data that obtain based on data prediction, calculate the index of correlation in returns of investment index system, then carry out simulation investment performance analysis, obtain simulation investment comprehensive grading result;
Step 3: the appraisal result that the multi-scheme preference policy based on learning automaton method draws different schemes carries out preferably, finally drawing best capital project.
Further, in described step 2, simulation investment performance analysis comprises the steps:
Step 2.1: achievement data normalization;
Step 2.2: index weights calculates;
Step 2.3: comprehensive grading.
Further, in described step 2.1, achievement data normalization adopts fuzzy membership mathematical method, specific as follows:
For positive correlation index, adopt the army mo(u)ld top half membership function as shown in the formula (1); For negative correlation index, adopt the army mo(u)ld bottom half membership function as shown in the formula (2); Between weak index and target, be part positive correlation, part negative correlation, adopts the osculant membership function as shown in the formula (3);
(1)
(2)
Wherein,
a 1,
a 2be respectively lower limit and the upper limit of membership function;
Wherein,
a,
dbe respectively lower limit and the upper limit of membership function,
b,
cbe respectively appropriateness interval [
b,
c] endpoints thereof.
Further, in described step 2.2, index weights calculates and adopts combining weights computing method, adopts respectively VC Method and entropy power method to calculate weight, composes power by combination, obtains the combining weights of two kinds of methods.
Further, adopting VC Method to calculate weight comprises the following steps:
1) suppose to have
mitem Investment appraisal indicators,
nindividual area power grid evaluation object,
xfor raw data matrix, wherein
x ij be
iof individual evaluation object
jthe numerical value of individual evaluation index:
2) calculate the standard deviation of each evaluation index, to reflect the absolute degree of variation of each evaluation index,
jthe standard deviation of individual evaluation index
s j for:
3) calculate the coefficient of variation of each evaluation index, to reflect the relative variability degree of each evaluation index,
jthe coefficient of variation of individual evaluation index
v j for:
4) coefficient of variation of each evaluation index is normalized, obtains each evaluation index weight
w j for:
Further, adopting entropy power method to calculate weight comprises the following steps:
1) suppose to have
mitem Investment appraisal indicators,
nindividual area power grid evaluation object,
x ij be
iindividual evaluation object is with respect to
jthe predetermined value of individual evaluation index, forms original index data matrix
x=(
x ij )
n × m , wherein
i≤
n,
j≤
m;
2) get the optimal value of each evaluation index
x * j , wherein evaluation index
jif direct index,
x * j =max{
x ij , evaluation index
jif inverse indicators,
x * j =min{
x ij ; Definition
x ij for
x * the degree of approach
d ij =
x ij /
x * j , obtain matrix
d=(
d ij )
n × m ; After matrix normalized, make 0≤
d ij ≤ 1,
, obtain matrix
d=(
d ij )
n × m ; Calculate evaluation index
jconditional entropy
e j for:
Wherein
;
3) calculate evaluation index
jthe entropy of importance
e(
d j ) be:
4) by
e(
d j ) determine evaluation index
jevaluation weights
λ j for:
Wherein,
, and
λ j meet: 0≤
λ j ≤ 1,
.
Further, the method that power is composed in combination is: adopt the synthetic method of multiplication that various weights are combined, and the multiplied by weight of a certain index drawing by different tax power methods, then normalized, finally obtains combining weights;
Supposing has
mindividual index,
qthe weight that kind method is tried to achieve
w, combining weights is:
Wherein,
θ j for the combining weights of index;
j=1,2 ...,
m;
k=1,2 ...,
q.
Further, in described step 2.3, the method for comprehensive grading is: gather each evaluation index value according to weight, to obtain the comprehensive evaluation score value of evaluation object, for examining judge object, specific as follows:
Select linear model to gather each index normalization desired value, finally obtain comprehensive grading;
Wherein,
, 0≤
w j ≤ 1,
j=1,2 ...,
m;
yfor the comprehensive grading value of evaluation object,
w j for evaluation index value
x j weight coefficient.
Further, in described step 3, the multi-scheme preference policy based on learning automaton comprises the following steps:
Step 3.1: random environment by a ternary set
α, β, crepresent, wherein,
αrepresent limited set of actions,
βrepresent response sets,
crepresent probability penalty set; Set
cin each element
c i corresponding set
αin element
α i ; Environment output variable
βvalue mode have multiplely, work as output
βbe 0 or 1 o'clock, be output as " 1 " and represent negativity response, " 0 " represents certainty response;
Step 3.2: learning automaton according to the information interaction of random environment produce one group of feasible solution
α 1,
α 2...,
α r ,
nwhen inferior renewal, environment is according to the feasible solution to input
α(
n) response, output
β(
n) { 0,1} upgrades feasible solution to ∈
α(
n) acceptance probability, and as the input of automat, for automat, next step selects feasible action that information is provided; Wherein, acceptance probability vector
p(
n)={
p 1(
n),
p 2(
n) ...,
p r (
n) in each element representation be:
p i (
n) = Pr{
α(
n) =
α i }
Wherein,
i=1,2 ...,
r; Under original state,
n=1 o'clock, the acceptance probability of each feasible solution was identical, is
p i (1)=1/
r, i=1,2 ...,
r;
Step 3.3:
nwhen inferior renewal,
n>1, if
α(
n)=
α i , based on
β(
n) output, acceptance probability phasor
p(
n)={
p 1(
n),
p 2(
n),
, p r (
n) update rule be:
When
β(
n)=0 o'clock:
When
β(
n)=1 o'clock:
In formula,
; 0<
a<1,0<
b<1,
afor rewarding parameter,
bfor punishment parameter; If
β(
n)=0, shows that output is sure, accepts feasible solution
, should increase corresponding acceptance probability
p i , reduce the probability that other actions are accepted
p j (
i≠
j); Otherwise, if
β(
n)=1, will reduce
p i , increase
p j (
j≠
i);
Step 3.4: use average punishment index
m(
n) assessment automat constringency performance:
Wherein, under original state (
n=1 o'clock), have
;
Step 3.5: by constantly feasible solution being selected, acceptance probability vector finally converges to a range of stability, and meet
m(
n) <
m(1); In acceptance probability vector, the highest feasible solution of acceptance probability is the optimum solution of multi-scheme optimal selection problem.
Compared to prior art, the invention has the beneficial effects as follows can be based on to electric grid investment benefit assay, optimize following electric grid investment, evaluation effect is objective, reasonable, for electric grid investment funds flow provides favourable reference frame, meet the demand of power network development and engineering actual investment, there is very strong practicality and wide application prospect.
Brief description of the drawings
Fig. 1 is realization flow figure of the present invention.
Fig. 2 is the learning process schematic diagram of learning automat of the present invention.
Fig. 3 is each scheme acceptance probability change procedure figure of embodiment of the present invention learning automat study process.
Fig. 4 is the average punishment index changing trend diagram in embodiment of the present invention learning automat study process.
Embodiment
The electric grid investment optimization method that the present invention is based on simulation investment performance analysis and learning automaton, as shown in Figure 1, comprises the following steps:
Step 1: adopt based on GM(1 1) grey forecasting model of model, carries out data prediction to power network development class data.
In described step 1, need the data of prediction to comprise: the delivery of company of next year province, power administration of city, electricity sales amount, electrical network original value of fixed assets, electric grid investment, 500kV power transformation capacity, 220kV power transformation capacity, 110kV power transformation capacity, 35kV power transformation capacity, 10kV power transformation capacity, average sales rate of electricity, 500 KV length (kilometer), 220 KV length (kilometer), 110 KV length (kilometer), 35 KV length (kilometer), 10(20) KV length (kilometer).
Step 2: the power network development class data that obtain based on data prediction, calculate the index of correlation in returns of investment index system, then carry out simulation investment performance analysis, obtain simulation investment comprehensive grading result.
In described step 2, simulation investment performance analysis comprises the steps:
Step 2.1: achievement data normalization;
Step 2.2: index weights calculates;
Step 2.3: comprehensive grading.
In described step 2.1, achievement data normalization adopts fuzzy membership mathematical method, specific as follows:
For positive correlation index, adopt the army mo(u)ld top half membership function as shown in the formula (1); For negative correlation index, adopt the army mo(u)ld bottom half membership function as shown in the formula (2); Between weak index and target, be part positive correlation, part negative correlation, adopts the osculant membership function as shown in the formula (3);
(1)
Wherein,
a 1,
a 2be respectively lower limit and the upper limit of membership function;
Wherein,
a,
dbe respectively lower limit and the upper limit of membership function,
b,
cbe respectively appropriateness interval [
b,
c] endpoints thereof.
In described step 2.2, index weights calculates and adopts combining weights computing method, adopts respectively VC Method and entropy power method to calculate weight, composes power by combination, obtains the combining weights of two kinds of methods.
Adopting VC Method to calculate weight comprises the following steps:
1) suppose to have
mitem Investment appraisal indicators,
nindividual area power grid evaluation object,
xfor raw data matrix, wherein
x ij be
iof individual evaluation object
jthe numerical value of individual evaluation index:
2) calculate the standard deviation of each evaluation index, to reflect the absolute degree of variation of each evaluation index,
jthe standard deviation of individual evaluation index
s j for:
3) calculate the coefficient of variation of each evaluation index, to reflect the relative variability degree of each evaluation index,
jthe coefficient of variation of individual evaluation index
v j for:
4) coefficient of variation of each evaluation index is normalized, obtains each evaluation index weight
w j for:
。
Adopting entropy power method to calculate weight comprises the following steps:
1) suppose to have
mitem Investment appraisal indicators,
nindividual area power grid evaluation object,
x ij be
iindividual evaluation object is with respect to
jthe predetermined value of individual evaluation index, forms original index data matrix
x=(
x ij )
n × m , wherein
i≤
n,
j≤
m;
2) get the optimal value of each evaluation index
x * j , wherein evaluation index
jif direct index,
x * j =max{
x ij , evaluation index
jif inverse indicators,
x * j =min{
x ij ; Definition
x ij for
x * the degree of approach
d ij =
x ij /
x * j , obtain matrix
d=(
d ij )
n × m ; After matrix normalized, make 0≤
d ij ≤ 1,
, obtain matrix
d=(
d ij )
n × m ; Calculate evaluation index
jconditional entropy
e j for:
Wherein
;
3) calculate evaluation index
jthe entropy of importance
e(
d j ) be:
4) by
e(
d j ) determine evaluation index
jevaluation weights
λ j for:
Wherein,
, and
λ j meet: 0≤
λ j ≤ 1,
.
The method that power is composed in combination is: adopt the synthetic method of multiplication that various weights are combined, and the multiplied by weight of a certain index drawing by different tax power methods, then normalized, finally obtains combining weights;
Supposing has
mindividual index,
qthe weight that kind method is tried to achieve
w, combining weights is:
Wherein,
θ j for the combining weights of index;
j=1,2 ...,
m;
k=1,2 ...,
q.
In described step 2.3, the method for comprehensive grading is: gather each evaluation index value according to weight, to obtain the comprehensive evaluation score value of evaluation object, for examining judge object, specific as follows:
Select linear model to gather each index normalization desired value, finally obtain comprehensive grading;
Wherein,
, 0≤
w j ≤ 1,
j=1,2 ...,
m;
yfor the comprehensive grading value of evaluation object,
w j for evaluation index value
x j weight coefficient.
Step 3: the appraisal result that the multi-scheme preference policy based on learning automaton method draws different schemes carries out preferably, finally drawing best capital project.
In described step 3, the multi-scheme preference policy based on learning automaton, as shown in Figure 2, comprises the following steps:
Step 3.1: random environment by a ternary set
α, β, crepresent, wherein,
αrepresent limited set of actions,
βrepresent response sets,
crepresent probability penalty set; Set
cin each element
c i corresponding set
αin element
α i ; Environment output variable
βvalue mode have multiplely, work as output
βbe 0 or 1 o'clock, be output as " 1 " and represent negativity response, " 0 " represents certainty response;
Step 3.2: learning automaton according to the information interaction of random environment produce one group of feasible solution
α 1,
α 2...,
α r ,
nwhen inferior renewal, environment is according to the feasible solution to input
α(
n) response, output
β(
n) { 0,1} upgrades feasible solution to ∈
α(
n) acceptance probability, and as the input of automat, for automat, next step selects feasible action that information is provided; Wherein, acceptance probability vector
p(
n)={
p 1(
n),
p 2(
n) ...,
p r (
n) in each element representation be:
p i (
n) = Pr{
α(
n) =
α i }
Wherein,
i=1,2 ...,
r; Under original state,
n=1 o'clock, the acceptance probability of each feasible solution was identical, is
p i (1)=1/
r, i=1,2 ...,
r;
Step 3.3:
nwhen inferior renewal,
n>1, if
α(
n)=
α i , based on
β(
n) output, acceptance probability phasor
p(
n)={
p 1(
n),
p 2(
n),
, p r (
n) update rule be:
When
β(
n)=0 o'clock:
When
β(
n)=1 o'clock:
In formula,
; 0<
a<1,0<
b<1,
afor rewarding parameter,
bfor punishment parameter; If
β(
n)=0, shows that output is sure, accepts feasible solution
, should increase corresponding acceptance probability
p i , reduce the probability that other actions are accepted
p j (
i≠
j); Otherwise, if
β(
n)=1, will reduce
p i , increase
p j (
j≠
i);
Step 3.4: use average punishment index
m(
n) assessment automat constringency performance:
Wherein, under original state (
n=1 o'clock), have
;
Step 3.5: by constantly feasible solution being selected, acceptance probability vector finally converges to a range of stability, and meet
m(
n) <
m(1); In acceptance probability vector, the highest feasible solution of acceptance probability is the optimum solution of multi-scheme optimal selection problem.
Below in conjunction with drawings and the specific embodiments, the invention will be further described.
Based on 9, certain province historical investment data in area, to each department next year electric grid investment carry out suggestion for investment.Historical data comprises 2009～investment data of each department in 2012, need to analyze the Optimal Investment decision-making of each department electrical network in 2013.
Plan to build the investment orientation of vertical three kinds of schemes.The first scheme is according to power administration of the each city delivery of 2009 ~ 2012 years, electricity sales amount, electric grid investment situation, predicts respectively the corresponding data of power administration of each city next year.Alternative plan is according to economizing the net delivery of 2009 ~ 2012 years, electricity sales amount, electric grid investment situation, prediction economizes the corresponding data of net next year, to economizing the predicted value of net, account for the ratio of whole network data according to 2012 power administration of Nian Ge city deliveries, electricity sales amount, electric grid investment, distribute in proportion accordingly respectively, the delivery, electricity sales amount, the electric grid investment that obtain power administration of each city predict the outcome.Third party's case is the actual electric network development according to each department, predicted data in scheme 2 is carried out to the result of coefficient adjustment.
1) according to three kinds of capital projects, adopt grey forecasting model, the Data Growth situation under prediction electrical network different schemes, obtains the index of correlation result of calculation under each scheme.
2) index adopting in simulation investment performance analysis comprises that unit electric grid investment increases electricity sales amount, line loss per unit (%), newly-increased unit capacity increases delivery (220kV), newly-increased unit capacity increases delivery (110kV), unit holds delivery (220kV), and unit holds delivery (110kV).Through normalization, weight calculation and comprehensive grading, can obtain simulation investment performance analysis evaluation result.Table 1 provides respectively under three kinds of different capital projects, the weighted value that utilizes three kinds of methods of weighting to calculate.
The index weights that the lower three kinds of methods of weighting of three kinds of schemes of table 1 are calculated
The comprehensive grading result that adopts three kinds of different tax power methods to obtain, under different schemes, the good and bad scheme relation that different regions electrical network is corresponding is different, in table 2.
The comprehensive grading result of three kinds of schemes of table 2
3) adopt learning automaton method, three of somewhere kinds of capital projects are trained, through what set
nafter=8000 study, the acceptance probability of three kinds of schemes is respectively: scheme 1:0.5608, scheme 2:0.1876, scheme 3:0.2516.The acceptance probability of scheme 1 is far longer than other two schemes.Fig. 3 is each scheme acceptance probability change procedure of learning automaton learning process.
Fig. 4 is the average punishment index variation tendency in learning process.As shown in Figure 3, the average punishment index in learning process is also continuous downtrending, and has average punishment index
m(
n)=0.0170,
m(1)=0.0208, meets M (n) <M (1) and sets up.Therefore, scheme 1 becomes the Optimal Investment scheme of final adaptation this area electrical network with larger acceptance probability.
Visible, the electric grid investment optimization method based on simulation investment performance analysis and learning automaton can effectively obtain prime investment scheme from multiple alternatives.
Be more than preferred embodiment of the present invention, all changes of doing according to technical solution of the present invention, when the function producing does not exceed the scope of technical solution of the present invention, all belong to protection scope of the present invention.
Claims (9)
1. the electric grid investment optimization method based on simulation investment performance analysis and learning automaton, is characterized in that, comprises the following steps:
Step 1: adopt grey forecasting model, power network development class data are carried out to data prediction;
Step 2: the power network development class data that obtain based on data prediction, calculate the index of correlation in returns of investment index system, then carry out simulation investment performance analysis, obtain simulation investment comprehensive grading result;
Step 3: the appraisal result that the multi-scheme preference policy based on learning automaton method draws different schemes carries out preferably, finally drawing best capital project.
2. the electric grid investment optimization method based on simulation investment performance analysis and learning automaton according to claim 1, is characterized in that, in described step 2, simulation investment performance analysis comprises the steps:
Step 2.1: achievement data normalization;
Step 2.2: index weights calculates;
Step 2.3: comprehensive grading.
3. the electric grid investment optimization method based on simulation investment performance analysis and learning automaton according to claim 2, is characterized in that, in described step 2.1, achievement data normalization adopts fuzzy membership mathematical method, specific as follows:
For positive correlation index, adopt the army mo(u)ld top half membership function as shown in the formula (1); For negative correlation index, adopt the army mo(u)ld bottom half membership function as shown in the formula (2); Between weak index and target, be part positive correlation, part negative correlation, adopts the osculant membership function as shown in the formula (3);
Wherein,
a 1,
a 2be respectively lower limit and the upper limit of membership function;
Wherein,
a,
dbe respectively lower limit and the upper limit of membership function,
b,
cbe respectively appropriateness interval [
b,
c] endpoints thereof.
4. the electric grid investment optimization method based on simulation investment performance analysis and learning automaton according to claim 2, it is characterized in that, in described step 2.2, index weights calculates and adopts combining weights computing method, adopt respectively VC Method and entropy power method to calculate weight, compose power by combination, obtain the combining weights of two kinds of methods.
5. the electric grid investment optimization method based on simulation investment performance analysis and learning automaton according to claim 4, is characterized in that, adopts VC Method to calculate weight and comprises the following steps:
1) suppose to have
mitem Investment appraisal indicators,
nindividual area power grid evaluation object,
xfor raw data matrix, wherein
x ij be
iof individual evaluation object
jthe numerical value of individual evaluation index:
2) calculate the standard deviation of each evaluation index, to reflect the absolute degree of variation of each evaluation index,
jthe standard deviation of individual evaluation index
s j for:
3) calculate the coefficient of variation of each evaluation index, to reflect the relative variability degree of each evaluation index,
jthe coefficient of variation of individual evaluation index
v j for:
4) coefficient of variation of each evaluation index is normalized, obtains each evaluation index weight
w j for:
6. the electric grid investment optimization method based on simulation investment performance analysis and learning automaton according to claim 4, is characterized in that, adopts entropy power method to calculate weight and comprises the following steps:
1) suppose to have
mitem Investment appraisal indicators,
nindividual area power grid evaluation object,
x ij be
iindividual evaluation object is with respect to
jthe predetermined value of individual evaluation index, forms original index data matrix
x=(
x ij )
n × m , wherein
i≤
n,
j≤
m;
2) get the optimal value of each evaluation index
x * j , wherein evaluation index
jif direct index,
x * j =max{
x ij , evaluation index
jif inverse indicators,
x * j =min{
x ij ; Definition
x ij for
x * the degree of approach
d ij =
x ij /
x * j , obtain matrix
d=(
d ij )
n × m ; After matrix normalized, make 0≤
d ij ≤ 1,
, obtain matrix
d=(
d ij )
n × m ; Calculate evaluation index
jconditional entropy
e j for:
Wherein
;
3) calculate evaluation index
jthe entropy of importance
e(
d j ) be:
4) by
e(
d j ) determine evaluation index
jevaluation weights
λ j for:
Wherein,
, and
λ j meet: 0≤
λ j ≤ 1,
.
7. the electric grid investment optimization method based on simulation investment performance analysis and learning automaton according to claim 4, it is characterized in that, the method that power is composed in combination is: adopt the synthetic method of multiplication that various weights are combined, the multiplied by weight of a certain index drawing by different tax power methods, then normalized, finally obtains combining weights;
Supposing has
mindividual index,
qthe weight that kind method is tried to achieve
w, combining weights is:
Wherein,
θ j for the combining weights of index;
j=1,2 ...,
m;
k=1,2 ...,
q.
8. the electric grid investment optimization method based on simulation investment performance analysis and learning automaton according to claim 2, it is characterized in that, in described step 2.3, the method of comprehensive grading is: gather each evaluation index value according to weight, to obtain the comprehensive evaluation score value of evaluation object, be used for examining judge object, specific as follows:
Select linear model to gather each index normalization desired value, finally obtain comprehensive grading;
Wherein,
, 0≤
w j ≤ 1,
j=1,2 ...,
m;
yfor the comprehensive grading value of evaluation object,
w j for evaluation index value
x j weight coefficient.
9. the electric grid investment optimization method based on simulation investment performance analysis and learning automaton according to claim 1, is characterized in that, in described step 3, the multi-scheme preference policy based on learning automaton comprises the following steps:
Step 3.1: random environment by a ternary set
α, β, crepresent, wherein,
αrepresent limited set of actions,
βrepresent response sets,
crepresent probability penalty set; Set
cin each element
c i corresponding set
αin element
α i ; Environment output variable
βvalue mode have multiplely, work as output
βbe 0 or 1 o'clock, be output as " 1 " and represent negativity response, " 0 " represents certainty response;
Step 3.2: learning automaton according to the information interaction of random environment produce one group of feasible solution
α 1,
α 2...,
α r ,
nwhen inferior renewal, environment is according to the feasible solution to input
α(
n) response, output
β(
n) { 0,1} upgrades feasible solution to ∈
α(
n) acceptance probability, and as the input of automat, for automat, next step selects feasible action that information is provided; Wherein, acceptance probability vector
p(
n)={
p 1(
n),
p 2(
n) ...,
p r (
n) in each element representation be:
p i (
n) = Pr{
α(
n) =
α i }
Wherein,
i=1,2 ...,
r; Under original state,
n=1 o'clock, the acceptance probability of each feasible solution was identical, is
p i (1)=1/
r, i=1,2 ...,
r;
Step 3.3:
nwhen inferior renewal,
n>1, if
α(
n)=
α i , based on
β(
n) output, acceptance probability phasor
p(
n)={
p 1(
n),
p 2(
n),
, p r (
n) update rule be:
When
β(
n)=0 o'clock:
When
β(
n)=1 o'clock:
In formula,
; 0<
a<1,0<
b<1,
afor rewarding parameter,
bfor punishment parameter; If
β(
n)=0, shows that output is sure, accepts feasible solution
, should increase corresponding acceptance probability
p i , reduce the probability that other actions are accepted
p j (
i≠
j); Otherwise, if
β(
n)=1, will reduce
p i , increase
p j (
j≠
i);
Step 3.4: use average punishment index
m(
n) assessment automat constringency performance:
Wherein, under original state (
n=1 o'clock), have
;
Step 3.5: by constantly feasible solution being selected, acceptance probability vector finally converges to a range of stability, and meet
m(
n) <
m(1); In acceptance probability vector, the highest feasible solution of acceptance probability is the optimum solution of multi-scheme optimal selection problem.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410132514.2A CN103886393A (en) | 2014-04-03 | 2014-04-03 | Power grid investment optimization method based on simulation investment benefit analysis and learning automatons |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410132514.2A CN103886393A (en) | 2014-04-03 | 2014-04-03 | Power grid investment optimization method based on simulation investment benefit analysis and learning automatons |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103886393A true CN103886393A (en) | 2014-06-25 |
Family
ID=50955275
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410132514.2A Pending CN103886393A (en) | 2014-04-03 | 2014-04-03 | Power grid investment optimization method based on simulation investment benefit analysis and learning automatons |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103886393A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104063763A (en) * | 2014-06-30 | 2014-09-24 | 江苏华大天益电力科技有限公司 | Share-based method for monitoring indirect economic benefit of informatization |
| CN104376435A (en) * | 2014-12-03 | 2015-02-25 | 国家电网公司 | Electric power and energy balance scheme evaluating method |
| CN111144605A (en) * | 2018-11-06 | 2020-05-12 | 财团法人工业技术研究院 | Investment strategy rule generating method and investment strategy rule generating device using same |
-
2014
- 2014-04-03 CN CN201410132514.2A patent/CN103886393A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104063763A (en) * | 2014-06-30 | 2014-09-24 | 江苏华大天益电力科技有限公司 | Share-based method for monitoring indirect economic benefit of informatization |
| CN104376435A (en) * | 2014-12-03 | 2015-02-25 | 国家电网公司 | Electric power and energy balance scheme evaluating method |
| CN111144605A (en) * | 2018-11-06 | 2020-05-12 | 财团法人工业技术研究院 | Investment strategy rule generating method and investment strategy rule generating device using same |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Majidi et al. | Application of information gap decision theory in practical energy problems: A comprehensive review | |
| CN104037943B (en) | A kind of voltage monitoring method and system that improve grid voltage quality | |
| CN103679544A (en) | Comprehensive assessment method for running of intelligent power distribution network | |
| CN104331844B (en) | A kind of power network construction project investment decision method | |
| CN107800140A (en) | A kind of large user for considering load characteristic, which powers, accesses decision-making technique | |
| CN103577679A (en) | Real-time computing method for theoretical line loss of low-voltage distribution room | |
| CN106022614A (en) | Data mining method of neural network based on nearest neighbor clustering | |
| CN108710992A (en) | A kind of demand response participates in the appraisal procedure of power grid peak load shifting effect | |
| CN108448620A (en) | High permeability distributed generation resource assemblage classification method based on integrated performance index | |
| Galvani et al. | A multi-objective probabilistic approach for smart voltage control in wind-energy integrated networks considering correlated parameters | |
| CN107358332A (en) | A kind of dispatching of power netwoks runs lean evaluation method | |
| CN107748944A (en) | Decontrol user side demand response method under environment in a kind of sale of electricity side | |
| CN105046354A (en) | Multi-agent power distribution network planning scene simulation generation method and system | |
| CN110198042A (en) | A kind of dynamic optimization method and storage medium of power grid energy storage | |
| Lashgari et al. | Transportation energy demand forecasting in Taiwan based on metaheuristic algorithms | |
| CN103886393A (en) | Power grid investment optimization method based on simulation investment benefit analysis and learning automatons | |
| CN105654194B (en) | A kind of power network development index measurement method method | |
| CN106384307A (en) | Differentiated evaluation method for county-area power distribution network plan | |
| Yammani et al. | Optimal dispatch of vehicle-to-grid (V2G) battery storage using p-ELECTRE method and its impact on optimal scheduling of DGs in distribution system | |
| CN108898322A (en) | A kind of electric grid investment strategy benefit integrated evaluating method based on FCM algorithm | |
| CN113947330A (en) | Multi-target power grid project optimization decision method and system considering investment time sequence | |
| Sanjani | The prediction of increase or decrease of electricity cost using fuzzy expert systems | |
| Sahraei-Ardakani et al. | Multi attribute optimal DG planning under uncertainty using AHP method | |
| Liu et al. | Research on power system load forecasting model based on data mining technology | |
| Tan | Research on sustainable carrying capacity of urban tourism environment based on multi objective optimization algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140625 |
|
| RJ01 | Rejection of invention patent application after publication |