CN107067092A - A kind of extra-high voltage electric transmission and transformation construction costs combination forecasting method and prediction meanss - Google Patents
A kind of extra-high voltage electric transmission and transformation construction costs combination forecasting method and prediction meanss Download PDFInfo
- Publication number
- CN107067092A CN107067092A CN201611057832.2A CN201611057832A CN107067092A CN 107067092 A CN107067092 A CN 107067092A CN 201611057832 A CN201611057832 A CN 201611057832A CN 107067092 A CN107067092 A CN 107067092A
- Authority
- CN
- China
- Prior art keywords
- submodule
- cost
- value
- engineering
- extra
- 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
- 230000009466 transformation Effects 0.000 title claims abstract description 49
- 230000005540 biological transmission Effects 0.000 title claims abstract description 45
- 238000010276 construction Methods 0.000 title claims abstract description 41
- 238000013277 forecasting method Methods 0.000 title claims abstract description 19
- 238000012549 training Methods 0.000 claims abstract description 26
- 238000000034 method Methods 0.000 claims description 58
- 238000012545 processing Methods 0.000 claims description 7
- 238000012706 support-vector machine Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 6
- 230000002068 genetic effect Effects 0.000 claims description 5
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 238000012417 linear regression Methods 0.000 claims description 4
- 239000002245 particle Substances 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 3
- 230000005611 electricity Effects 0.000 claims description 2
- 238000007689 inspection Methods 0.000 claims description 2
- 210000002569 neuron Anatomy 0.000 claims 1
- 238000007726 management method Methods 0.000 abstract description 5
- 230000006870 function Effects 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 238000009434 installation Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000007418 data mining Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000009897 systematic effect Effects 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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of extra-high voltage electric transmission and transformation construction costs combination forecasting method and prediction meanss.Extra-high voltage electric transmission and transformation engineering historical sample is divided into training sample and checking sample by the present invention；The structure of model is predicted to the submodule of extra-high voltage electric transmission and transformation engineering respectively with a variety of cost forecasting methods；The error between each submodule predicted value and actual value according to obtained by each forecast model, the preliminary weight of each Individual forecast model is calculated with entropy assessment；The preliminary weight of forecast model is modified, the final weight of each submodule combination forecasting of extra-high voltage project is obtained；The submodule cost combined prediction value of engineering to be predicted is calculated according to final weight, each submodule cost combined prediction value, which is simply added, can obtain the cost combined prediction value of extra-high voltage project to be predicted.The present invention is used to carry out cost forecasting to extra-high voltage electric transmission and transformation engineering, particularly for the cost forecasting of engineering yet to be built, so as to instruct the decision-making and engineering management of Practical Project.
Description
Technical field
The invention belongs to technical field of power systems, especially a kind of extra-high voltage electric transmission and transformation construction costs combination forecasting method
And prediction meanss.
Background technology
China refers to alternating voltage grade to the regulation of extra-high voltage in more than 1000kV, DC voltage level ± 800kV with
On.Extra-high voltage project is at present also in development and the initial stage built, and its cost has the features such as investment is high, scale is big.For energetically
Improve the level of security and economic benefit of extra-high voltage electric transmission and transformation engineering construction, it is necessary to which it is pre- to carry out rational cost to engineering yet to be built
Survey.Advanced extra-high voltage project cost forecasting scheme by for the reasonable determination of building cost control target and optimization provide it is important according to
According to, be each unit in charge of construction's cost of investment control and cost optimization solid support is provided.Study extra-high voltage project cost forecasting
Major significance is embodied in the following aspects：First, help to lift the Predicting Technique level of extra-high voltage project cost；Second,
Contribute to the determination and optimization of extra-high voltage cost Control target, improve scientific decision ability；3rd, contribute to General Promotion each
Extra-high voltage project cost management and control lean level, raising economic benefit of investment and the social benefit of interests side.
At present, although studying more to cost forecasting, it is all based on what conventional power transformation engineering was carried out, to extra-high voltage project
Cost forecasting, yet there are no correlative study.Extra-high voltage project has its distinctive cost feature, existing conventional project cost forecasting scheme
And inapplicable extra-high voltage project, under the background that extra-high voltage electric transmission and transformation technology is increasingly mature, extra-high voltage project construction is booming,
Need the cost evaluation level for building a set of prediction scheme guidance yet to be built engineering applicable to its badly.On the one hand, extra-high voltage project is sent out
Exhibition its cost at initial stage has the features such as investment is high, historical sample is few, cost composition is complicated, this power transmission and transformation work ripe with having evolved into
Journey cost feature is different, need to combine extra-high voltage project cost feature and build forecast model；On the other hand conventional project of transmitting and converting electricity should
Individual forecast method does not have universality, even if being applicable certain engineering, is not necessarily suitable Other Engineering yet.
The content of the invention
There is provided a kind of combination of extra-high voltage electric transmission and transformation construction costs is pre- for the characteristics of present invention combines current extra-high voltage project cost
Survey method, it can obtain preferable cost forecasting result under the historical data support of extra-high voltage electric transmission and transformation engineering small sample,
Reference is provided to extra-high voltage project construction and cost management and control.
In order to achieve the above object, the present invention is adopted the following technical scheme that：A kind of extra-high voltage electric transmission and transformation construction costs combination
Forecasting Methodology, it comprises the following steps：
1) extra-high voltage electric transmission and transformation engineering historical sample data is obtained, initial data is formed, to original before prediction computing
Data are pre-processed, and to improve the quality of data, improve precision, efficiency and the performance of data mining process；
2) after the completion of data prediction, extra-high voltage electric transmission and transformation engineering historical sample is divided into training sample and checking sample,
Training sample is used to calculating and being fitted forecast model, and checking sample is used for the quality for verifying gained forecast model, and for predicting
The weight calculation of model；
3) submodule is divided to each historical sample, obtains the cost value and key influence factor value of each submodule；
4) according to the cost value and key influence factor value of each submodule of training sample, using a variety of cost forecasting method structures
Build a variety of forecast models of each submodule；
5) using step 4) obtained a variety of forecast models of each submodule, to verify the crucial effect of each submodule of sample
The key influence factor value of factor value and each submodule of engineering to be predicted calculates checking sample and engineering pair to be predicted as input
Each submodule cost forecasting value answered；
6) error between checking each submodule cost forecasting value of sample and cost actual value is calculated, according to error entropy weight
Method tries to achieve the preliminary weight of a variety of forecast models of each submodule；
7), will if meeting using verifying whether sample and construction costs predicted value to be predicted inspection forecast model meet requirement
Ask, then preliminary weight is not modified, preliminary weight is final weight；If being unsatisfactory for requiring, preliminary weight is carried out
Amendment；
8) the cost combined prediction value of each submodule of engineering to be predicted is calculated according to final weight result；
9) according to the cost combined prediction value of each submodule of engineering to be predicted, by be simply added engineering to be predicted is made
Valency combined prediction value.
Further, step 1) also include obtaining project data to be predicted and obtain to be predicted according to project data to be predicted
The key influence factor value of each submodule of engineering.
Further, the extra-high voltage electric transmission and transformation engineering include extra-high voltage power transformation engineering, it is ultra-high voltage converter station engineering, extra-high
Press alternating current circuit engineering and UHVDC Transmission Lines engineering.
Further, described a variety of forecast models are based on the small sample feature foundation of extra-high voltage project, and described is more
Kind of Forecasting Methodology include multiple linear regression, artificial neural network, SVMs, the improved SVMs of genetic algorithm and
The improved SVMs of particle cluster algorithm.
Further, described small sample feature refers to based on current extra-high voltage project number is less, construction costs is not advised
Model and the less condition of historical sample data.
Further, described training sample, checking sample and project data to be predicted divide the foundation that submodule is used
Unanimously.
Further, step 6) comprise the following steps that：Obtain checking each submodule cost forecasting value of sample and checking sample
Error between this each submodule cost actual value, builds error assessment system；According to error assessment system, asked using entropy assessment
Obtain the preliminary weight of each a variety of forecast models of submodule.
Further, cost forecasting value of the final weight of a variety of forecast models based on engineering to be predicted is determined, tool
The content of body is as follows：
When the cost forecasting value of engineering to be predicted obtained by certain forecast model is less than 0, the weight of the forecast model is put
For 0, i.e., the forecast model is not considered；
When the cost forecasting value of engineering to be predicted obtained by certain forecast model is in beyond setting range, by the prediction mould
The weight of type is set to 0, i.e., do not consider the forecast model.
Further, step 9) particular content include：
According to the final weight, treat each submodule cost forecasting value of predictive engine and be weighted summation, obtain treating pre-
Survey each submodule cost combined prediction value of engineering；
According to each submodule cost combined prediction value of engineering to be predicted, work to be predicted is tried to achieve by simple addition of algebra
Journey cost combined prediction value.
Present invention also offers the prediction meanss that above-mentioned extra-high voltage electric transmission and transformation construction costs combination forecasting method is used, bag
Data processing unit, Forecasting Methodology unit, model construction unit, weight determining unit and cost integrated unit are included,
Data processing unit, for screening and arranging extra-high voltage electric transmission and transformation engineering historical sample data, forms each submodule
Key influence factor storehouse；Each submodule key influence factor value of historical sample, each submodule cost value are filtered out, by submodule
Sample is divided into training sample and checking sample；
Forecasting Methodology unit, forms the Forecasting Methodology of a variety of forecast models, including multiple linear regression Forecasting Methodology, artificial
Neural net prediction method, SVM prediction method, genetic algorithm SVM prediction method and population support to
Amount machine Forecasting Methodology；
Model construction unit, according to the training sample submodule cost value and training sample submodule crucial effect of input because
Element value, is fitted and builds the forecast model obtained by certain Forecasting Methodology；
Cost forecasting unit, according to the forecast model, inputs each submodule checking sample and engineering to be predicted respectively
Key influence factor, and export the cost forecasting value under the corresponding a variety of forecast models of submodule；
Weight determining unit, it is preliminary to establish according to the error between the cost forecasting value and cost actual value of checking sample
The weight of each each forecast model of submodule, forms preliminary weight；According to the cost forecasting value of engineering to be predicted, preliminary weight is entered
Row amendment obtains each forecast model final weight of each submodule；
Cost integrated unit, according to making for each forecast model final weight of each submodule and each forecast model of engineering to be predicted
Valency predicted value, each submodule cost combined prediction value of engineering to be predicted is formed by weighted sum；Finally by engineering to be predicted
Simple be added is carried out to submodule cost combined prediction value can obtain the cost forecasting of the extra-high voltage electric transmission and transformation engineering to be predicted
Value.
The present invention can obtain preferable cost forecasting under the historical data support of extra-high voltage electric transmission and transformation engineering small sample
As a result, reference is provided to extra-high voltage project construction and cost management and control；The present invention is used to make extra-high voltage electric transmission and transformation engineering
Valency is predicted, particularly for the cost forecasting of engineering yet to be built, so as to direct the decision-making and engineering management of Practical Project.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the accompanying drawing used required in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can be with
Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is a kind of schematic flow sheet for extra-high voltage electric transmission and transformation construction costs Forecasting Methodology that the embodiment of the present invention 1 is provided；
Fig. 2 is a kind of schematic flow sheet for extra-high voltage electric transmission and transformation construction costs Forecasting Methodology that the embodiment of the present invention 2 is provided；
Fig. 3 is the structural representation of prediction meanss of the present invention；
Fig. 4 is UHVDC converter station engineering hierarchical decomposition figure of the present invention；
Fig. 5 divides figure for the submodule of DC line engineering of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.It is based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made
Embodiment, belongs to the scope of protection of the invention.
Embodiment 1
Shown in reference picture 1, a kind of extra-high voltage electric transmission and transformation construction costs Forecasting Methodology specifically includes following steps：
101st, sample data collection should be carried out first, form initial data.But initial data has scrambling, imperfect
Property, the structure and execution efficiency of model algorithm can be had a strong impact on, calculating error result is even resulted in, therefore, prediction computing it
Before initial data need to be pre-processed, to improve the quality of data, improve precision, efficiency and the performance of data mining process.
102nd, after the completion of data prediction, training sample and the checking class of sample two are splitted data into, training sample is based on
Calculate and fitting forecast model, checking sample is used for verifying the quality of gained model, and for the weight calculation of forecast model.
103rd, submodule is divided to each historical sample.
Divide from engineering characteristic, extra-high voltage project can be divided into exchange power transformation engineering, DC converter station engineering, alternating current circuit work
Journey, DC line engineering.Different engineerings can continue to be divided into sub- engineering level even submodule level according to expense again.With the change of current
Exemplified by engineering of standing, converter station project can be divided into architectural engineering expense, installing engineering expense and original equipment cost three by expense category
The expense of individual sub- engineering level.Sub- engineering can continue, by function different demarcation submodule level, to be illustrated in figure 4 extra-high voltage direct-current
Converter station project hierarchical decomposition figure, the division of Other Engineering is similar with this, i.e., first divided by expense, then divided by function.Fig. 4
In, converter station project original equipment cost can divide four submodules by functions of modules, be respectively：Valve Room installation cost, change of current transformation
Device system equipment takes, power distribution equipment installation cost, other main production engineering installation costs.In addition, also can recognize that according to submodule each
The influence factor of submodule cost, and then determine the key influence factor larger to submodule cost influence.Fig. 5 is DC line
The submodule of engineering is divided, and submodule is used as because line project data organization is special, therefore using the result divided by expense.
104th, each a variety of forecast models of submodule are built according to training sample.
Input training sample key influence factor value and cost value, respectively using five kinds of Forecasting Methodology MLR, ANN, SVM,
GASVM and PSOSVM carries out Function Fitting and model construction, obtains five kinds of method forecast models of each submodule.
105th, a variety of forecast models obtained according to 104, with the key for each submodule for verifying sample and engineering to be predicted
Influence factor value tries to achieve the cost forecasting value of checking sample and the corresponding each submodule of engineering to be predicted as input.
106th, the error between checking each submodule cost forecasting value of sample and actual value is calculated, according to error entropy weight
Method can try to achieve the preliminary weight of forecast model of each submodule.
Error is to reflect predict the outcome quality, the important indicator of precision of forecasting model.Checking obtained by five kinds of forecast models
Sample results are different, thus can the error analysis based on checking sample, using a variety of error criterions, then according to entropy assessment structure
Build combination forecasting proportional system.The present invention obtains mean square error (Mean from the predicted value and actual value of checking sample
Squared Error, MSE), mean absolute error (Mean Absolute Error, MAE) and mean absolute percentage error
(Mean Absolute Percent Error, MAPE) is predicted Model Weight determination.The calculating of each error criterion is respectively such as
Under, provided with n checking sample, yiFor actual value, Δi, (i=1,2 ..., n) for checking sample predicted value and actual value it is exhausted
To error.
The emphasis of a variety of forecast models is to determine the weight of each forecast model, to finally obtain cost combined prediction value.
Entropy assessment is a kind of method of metrics evaluation and program decisions, is had a wide range of applications.Entropy assessment basic thought is each to be evaluated
The information of valency unit is quantified with being integrated；Each factor is assigned using entropy assessment and weighed, evaluation procedure can be simplified.
107th, judge whether gained submodule cost forecasting value meets requirement.
Require, preliminary weight is not modified, preliminary weight is final weight if meeting.
If being unsatisfactory for requiring, preliminary weight is modified.
108th, it is right according to checking sample submodule cost forecasting value and the size of engineering submodule cost forecasting value to be predicted
The method weight of each forecast model is modified, and should specifically meet following two criterions.
Criterion 1, non-negative criterion.Practical Project cost is always greater than 0, if checking sample predictions result obtained by certain forecast model
Or Engineering prediction result to be predicted is when bearing, then it is believed that the model is not suitable for this group of data, such as to allow it to participate in combined prediction,
Precision of prediction can be reduced.Therefore its forecast model weight is assigned to 0 value.
Criterion 2,5±1Criterion.Arrange historical sample data to find, the sample close to submodule cost mode is more, up to going through
More than half of history sample；Also, the ratio between the maximum of all submodule historical sample costs and mode are no more than five times, minimum
Value is not less than 1/5. compared with mode accordingly it is recognized herein that a variety of actual values fluctuation of extra-high voltage project submodule cost typically exists
The 5 of average±1In the range of times.Then think obtained by the Forecasting Methodology in advance more than this scope when having the ratio between partial predictor and mode
Measured value is inaccurate.Specifically, if certain Forecasting Methodology obtained by checking sample predictions result or construction costs to be predicted predict the outcome with
During other Forecasting Methodology accordingly results difference more (being more than 5 times or less than 1/5 times), then it is believed that the model is not suitable for this group
Data, such as allow it to participate in combined prediction, can reduce precision of prediction.Therefore this kind of forecast model weight is assigned to 0 value, i.e., do not considered
The model.
110th, each submodule cost combined prediction value is calculated according to final weight result.Example, it is as follows
To submodule zi, wherein i is submodule block number, and the cost forecasting value of p kind forecast models is respectively cij, weight difference
For ωj, wherein j=1,2 ..., p, then submodule ziCost combined prediction value
ci=ωjcij
111st, according to the cost combined prediction value of each submodule, the cost combination of engineering to be predicted can be obtained by being simply added
Predicted value, such as following formula
Embodiment 2
Shown in reference picture 2, a kind of extra-high voltage electric transmission and transformation construction costs Forecasting Methodology specifically includes following steps：
201st, extra-high voltage electric transmission and transformation engineering historical data is combed and pre-processed.
With reference to above-described embodiment step 101, it will not be repeated here.
202nd, according to historical data, select a part of sample as checking sample, checking sample is used for verifying gained model
Quality, and for combination forecasting weight calculation.
203rd, according to historical data, select a part of sample as training sample, training sample is pre- for calculating and being fitted
Survey model.
204th, submodule is divided to all training samples, sorts out the submodule cost value and crucial effect of each training sample
Factor value.The division of submodule is referring to 103.
205th, the forecast model of different submodules under a variety of Forecasting Methodologies is built.
To each submodule, using a variety of Forecasting Methodologies, including but not limited to MLR, ANN, SVM, GASVM and PSOSVM this
Five kinds of methods.It is fitted with training sample submodule cost value and key influence factor value and trains each Forecasting Methodology, obtains each son
A variety of forecast models of module.
206th, submodule is divided to all checking samples, sorts out the submodule cost value and crucial effect of each checking sample
Factor value.The division of submodule is referring to 103.
207th, a variety of forecast models of each submodule described in 205 are called, to verify each submodule key influence factor of sample
Value calculates checking each submodule cost forecasting value of sample as input.
208th, according to forecast model, output checking each submodule cost forecasting value of sample.
209th, according to the error between the cost forecasting value and actual value of a variety of forecast models of each submodule, mean square error is calculated
Poor (MSE), mean absolute error (MAE) and mean absolute percentage error (MAPE).
210th, according to each error criterion, the initial weight of each forecast model is calculated with entropy assessment, specifically
Predicated error matrix M is built to each submodule(i×j), (i=1 ..., 5；J=1,2,3).Element mijRepresent i-th
Checking sample predictions value and the jth class error of actual value that kind Forecasting Methodology is obtained.Wherein j=1,2,3 represents error class respectively
Wei not MSE, MAE, MAPE.If checking sample has multiple values, mijFor the average value of multiple predicted values and actual value error.
Data normalization processing, eliminates error dimension.Normalized error
Calculate the accounting of the different forecast models of jth kind error
Calculate the entropy of each forecast model
Calculate the weight of jth error criterion, that is, coefficient of variation
Calculate the preliminary weight of each forecast model.Define the preliminary weight of each forecast model
Illustratively, such as the prediction data that table 1 takes for converter station project submodule change of current change systematic building, MSE, MAE are calculated,
Tri- kinds of errors of MAPE, preliminary weight calculation step can obtain the preliminary weight such as table 2 of each forecast model more than.
Table 1
Table 2
211st, each submodule key influence factor value of engineering to be predicted is obtained.
212nd, a variety of forecast models of each submodule described in 205 are called, with each submodule crucial effect of engineering to be predicted because
Element value calculates each submodule cost forecasting value of engineering to be predicted as input.
213rd, output is according to a variety of forecast models prediction gained each submodule cost forecasting value of engineering to be predicted.
214th, the non-negative criterion and 5 according to 108±1Criterion is modified to forecast model weight.
Illustratively, table 3 is the forecast model weight of converter station project submodule AC-DC power distribution unit installation cost, due to treating
The cost forecasting value of predictive engine MLR models violates non-negative criterion, therefore its weight is set to 0, obtains final each forecast model
Weight.
Table 3
| MLR | ANN | SVM | GA-SVM | PSO-SVM | |
| Construction costs predicted value to be predicted | -7.5E+09 | 9.45E+08 | 1.2E+09 | 1.2E+09 | 1192400000 |
| Preliminary weight | 0.99036 | 0.00041 | 0.00479 | 0.00401 | 0.00043 |
| Final weight | 0 | 0.0429 | 0.4969 | 0.4160 | 0.0443 |
Illustratively, table 4 is the checking sample predictions value and preliminary weight of DC line engineering annex engineering submodule, due to
Construction costs predicted value to be predicted violates 5 obtained by ANN forecast models±1Criterion, therefore its weight is set to 0, obtains final each pre-
Survey Model Weight.
Table 4
215th, it is pre- according to each forecast model weight of each submodule and each forecast model of each submodule of engineering to be predicted tried to achieve
Measured value, the cost combined prediction value of submodule can be tried to achieve by weighted sum.As described in 110.
216th, by each submodule cost combined prediction value, the cost of engineering to be predicted can be tried to achieve by simple algebraic addition
Combined prediction value.
Embodiment 3
Shown in reference picture 3, a kind of extra-high voltage electric transmission and transformation construction costs combined prediction device, for implementing above-described embodiment 1-
Extra-high voltage electric transmission and transformation construction costs combination forecasting method in 2, is specifically included：
Data processing unit 301, for screening and arranging extra-high voltage electric transmission and transformation engineering historical sample data.Form each submodule
The key influence factor storehouse of block.Filter out each submodule key influence factor value of historical sample, each submodule cost value.By submodule
Block sample is divided into training sample and checking sample.
Forecasting Methodology unit 302, forms the method class of a variety of forecast models.Mainly include but is not limited to multiple linear regression
Forecasting Methodology, neural network prediction method, SVM prediction method, genetic algorithm SVM prediction method,
Particle swarm support vector machine Forecasting Methodology.
Model construction unit 303, it is crucial for the training sample submodule cost according to input and training sample submodule
Influence factor value, is fitted and builds the cost forecasting model obtained by certain Forecasting Methodology.
Cost forecasting unit 304, for according to the cost forecasting model, inputting each submodule checking sample respectively and treating
The key influence factor of predictive engine, and export the cost forecasting value under the corresponding a variety of forecast models of submodule.
Weight determining unit 305, it is preliminary true according to the error between the cost forecasting value and cost actual value of checking sample
The weight of each each forecast model of submodule is found, preliminary weight is formed；According to the quality of the cost forecasting value of engineering to be predicted, to first
Step weight, which is modified, obtains each forecast model final weight of each submodule.
Cost integrated unit 306, according to each forecast model weight of each submodule and the cost forecasting value of each forecast model, leads to
Cross weighted sum and form submodule cost forecasting value.Institute can be obtained finally by simple be added is carried out to submodule cost forecasting value
The cost forecasting value of the extra-high voltage electric transmission and transformation engineering to be predicted.
Those of ordinary skill in the art are it is to be appreciated that the list of each example described with reference to the embodiments described herein
Member, can be realized with the combination of electronic hardware or computer software and electronic hardware.
It is apparent to those skilled in the art that, for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, may be referred to the corresponding process in preceding method embodiment, will not be repeated here.
, can be by it in several embodiments provided herein, it should be understood that disclosed apparatus and method
Its mode is realized.For example, apparatus embodiments described above are only schematical, for example, the division of the unit, only
Only a kind of division of logic function, can there is other dividing mode when actually realizing.In addition, in each embodiment of the invention
Each functional unit can be integrated in a processing unit or unit is individually physically present, can also be two
Or two or more unit is integrated in a unit.
If the function is realized using in the form of SFU software functional unit and is used as independent production marketing or in use, can be with
It is stored in a computer read/write memory medium.Understood based on such, technical scheme is substantially in other words
The part contributed to prior art or the part of the technical scheme can be embodied in the form of software product, the meter
Calculation machine software product is stored in a storage medium, including some instructions are to cause a computer equipment (can be individual
People's computer, server, or network equipment etc.) perform all or part of step of each of the invention embodiment methods described.
And foregoing storage medium includes：USB flash disk, mobile hard disk, read-only storage (English abbreviation：ROM, English full name:Read-Only
Memory), random access memory (English abbreviation：RAM, English full name：Random Access Memory), magnetic disc or light
Disk etc. is various can be with the medium of store program codes.
The foregoing is only a specific embodiment of the invention, but protection scope of the present invention is not limited thereto, any
Those familiar with the art the invention discloses technical scope in, change or replacement can be readily occurred in, should all be contained
Cover within protection scope of the present invention.Therefore, protection scope of the present invention described should be defined by scope of the claims.
Claims (10)
1. a kind of extra-high voltage electric transmission and transformation construction costs combination forecasting method, it comprises the following steps：
1) extra-high voltage electric transmission and transformation engineering historical sample data is obtained, initial data is formed, to initial data before prediction computing
Pre-processed；
2) after the completion of data prediction, extra-high voltage electric transmission and transformation engineering historical sample is divided into training sample and checking sample, training
Sample is used to calculating and being fitted forecast model, and checking sample is used for the quality for verifying gained forecast model, and for forecast model
Weight calculation；
3) submodule is divided to each historical sample, obtains the cost value and key influence factor value of each submodule；
4) according to the cost value and key influence factor value of each submodule of training sample, built using a variety of cost forecasting methods each
A variety of forecast models of submodule；
5) using step 4) obtained a variety of forecast models of each submodule, to verify the key influence factor of each submodule of sample
Value and the key influence factor value of each submodule of engineering to be predicted calculate checking sample and engineering to be predicted are corresponding as input
Each submodule cost forecasting value；
6) error between checking each submodule cost forecasting value of sample and cost actual value is calculated, is asked according to error with entropy assessment
Obtain the preliminary weight of a variety of forecast models of each submodule；
7) using verifying whether sample and construction costs predicted value to be predicted inspection forecast model meet requirement, required if meeting,
Preliminary weight is not modified then, preliminary weight is final weight；If being unsatisfactory for requiring, preliminary weight is repaiied
Just；
8) the cost combined prediction value of each submodule of engineering to be predicted is calculated according to final weight result；
9) according to the cost combined prediction value of each submodule of engineering to be predicted, by be simply added engineering to be predicted cost group
Close predicted value.
2. extra-high voltage electric transmission and transformation construction costs combination forecasting method according to claim 1, it is characterised in that step 1) also
Including obtain project data to be predicted and according to project data to be predicted obtain the crucial effect of each submodule of engineering to be predicted because
Element value.
3. extra-high voltage electric transmission and transformation construction costs combination forecasting method according to claim 1, it is characterised in that described extra-high
Project of transmitting and converting electricity is pressed to include extra-high voltage power transformation engineering, ultra-high voltage converter station engineering, extra-high voltage AC circuit engineering and extra-high straightening
Flow Line engineering.
4. extra-high voltage electric transmission and transformation construction costs combination forecasting method according to claim 1, it is characterised in that described is more
The small sample feature foundation that forecast model is based on extra-high voltage project is planted, described a variety of Forecasting Methodologies include multiple linear and returned
Return, artificial neural network, SVMs, the improved SVMs of genetic algorithm and the improved supporting vector of particle cluster algorithm
Machine.
5. extra-high voltage electric transmission and transformation construction costs combination forecasting method according to claim 4, it is characterised in that described is small
Sample feature refers to based on the bar that current extra-high voltage project number is less, construction costs is lack of standardization and historical sample data is less
Part.
6. extra-high voltage electric transmission and transformation construction costs combination forecasting method according to claim 1, it is characterised in that described instruction
It is consistent with the foundation that project data to be predicted divides submodule use to practice sample, checking sample.
7. extra-high voltage electric transmission and transformation construction costs combination forecasting method according to claim 1, it is characterised in that
Step 6) comprise the following steps that：Obtain checking each submodule cost forecasting value of sample and checking each submodule cost of sample
Error between actual value, builds error assessment system；According to error assessment system, each submodule is tried to achieve using entropy assessment a variety of
The preliminary weight of forecast model.
8. extra-high voltage electric transmission and transformation construction costs combination forecasting method according to claim 1, it is characterised in that
Cost forecasting value of the final weight of a variety of forecast models based on engineering to be predicted determines that specific content is as follows：
When the cost forecasting value of engineering to be predicted obtained by certain forecast model is less than 0, the weight of the forecast model is set to 0,
The forecast model is not considered；
When the cost forecasting value of engineering to be predicted obtained by certain forecast model is in beyond setting range, by the forecast model
Weight is set to 0, i.e., do not consider the forecast model.
9. extra-high voltage electric transmission and transformation construction costs combination forecasting method according to claim 1, it is characterised in that step 9)
Particular content includes：
According to the final weight, treat each submodule cost forecasting value of predictive engine and be weighted summation, obtain work to be predicted
Each submodule cost combined prediction value of journey；
According to each submodule cost combined prediction value of engineering to be predicted, engineering to be predicted is tried to achieve by simple addition of algebra and made
Valency combined prediction value.
10. the prediction meanss that any one of the claim 1-9 extra-high voltage electric transmission and transformation construction costs combination forecasting methods are used,
Including data processing unit, Forecasting Methodology unit, model construction unit, weight determining unit and cost integrated unit, its feature
It is：
Data processing unit, for screening and arranging extra-high voltage electric transmission and transformation engineering historical sample data, forms the pass of each submodule
Key influence factor storehouse；Each submodule key influence factor value of historical sample, each submodule cost value are filtered out, by submodule sample
It is divided into training sample and checking sample；
Forecasting Methodology unit, forms the Forecasting Methodology of a variety of forecast models, including multiple linear regression Forecasting Methodology, artificial neuron
Neural network forecast method, SVM prediction method, genetic algorithm SVM prediction method and particle swarm support vector machine
Forecasting Methodology；
Model construction unit, according to the training sample submodule cost value and training sample submodule key influence factor of input
Value, is fitted and builds the forecast model obtained by certain Forecasting Methodology；
Cost forecasting unit, according to the forecast model, inputs the key of each submodule checking sample and engineering to be predicted respectively
Influence factor, and export the cost forecasting value under the corresponding a variety of forecast models of submodule；
Weight determining unit, according to the error between the cost forecasting value and cost actual value of checking sample, tentatively establishes each son
The weight of each forecast model of module, forms preliminary weight；According to the cost forecasting value of engineering to be predicted, preliminary weight is repaiied
Just obtaining each forecast model final weight of each submodule；
Cost integrated unit, it is pre- according to the cost of each forecast model final weight of each submodule and each forecast model of engineering to be predicted
Measured value, each submodule cost combined prediction value of engineering to be predicted is formed by weighted sum；Finally by engineering antithetical phrase to be predicted
Module cost combined prediction value, which carries out simple be added, can obtain the cost forecasting value of the extra-high voltage electric transmission and transformation engineering to be predicted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611057832.2A CN107067092A (en) | 2016-11-25 | 2016-11-25 | A kind of extra-high voltage electric transmission and transformation construction costs combination forecasting method and prediction meanss |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611057832.2A CN107067092A (en) | 2016-11-25 | 2016-11-25 | A kind of extra-high voltage electric transmission and transformation construction costs combination forecasting method and prediction meanss |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107067092A true CN107067092A (en) | 2017-08-18 |
Family
ID=59618416
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611057832.2A Pending CN107067092A (en) | 2016-11-25 | 2016-11-25 | A kind of extra-high voltage electric transmission and transformation construction costs combination forecasting method and prediction meanss |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107067092A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108537412A (en) * | 2018-03-19 | 2018-09-14 | 国网冀北电力有限公司经济技术研究院 | Power distribution network Project Cost processing method and processing device |
| CN108921334A (en) * | 2018-06-15 | 2018-11-30 | 国家电网公司 | A kind of project amount control target of extra-high voltage electric transmission and transformation engineering determines method and device |
| CN112508512A (en) * | 2020-11-26 | 2021-03-16 | 国网河北省电力有限公司经济技术研究院 | Power grid engineering cost data management method and device and terminal equipment |
| CN112840385A (en) * | 2018-10-12 | 2021-05-25 | 韩国水力原子力株式会社 | Power plant early warning device and method adopting multiple prediction models |
| CN113011633A (en) * | 2021-02-05 | 2021-06-22 | 国网湖南省电力有限公司 | Overhead transmission line engineering cost prediction method, device and medium |
| CN113807613A (en) * | 2021-10-13 | 2021-12-17 | 国网安徽省电力有限公司 | Power grid engineering full-period budget optimization configuration tool |
| CN114066074A (en) * | 2021-11-19 | 2022-02-18 | 广东电网有限责任公司 | Method, device, equipment and medium for predicting maintenance frequency of old transformer assets |
| CN114169598A (en) * | 2021-11-29 | 2022-03-11 | 国网福建省电力有限公司 | Power transmission and transformation project progress early warning method considering complex environment |
-
2016
- 2016-11-25 CN CN201611057832.2A patent/CN107067092A/en active Pending
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108537412A (en) * | 2018-03-19 | 2018-09-14 | 国网冀北电力有限公司经济技术研究院 | Power distribution network Project Cost processing method and processing device |
| CN108921334A (en) * | 2018-06-15 | 2018-11-30 | 国家电网公司 | A kind of project amount control target of extra-high voltage electric transmission and transformation engineering determines method and device |
| CN112840385A (en) * | 2018-10-12 | 2021-05-25 | 韩国水力原子力株式会社 | Power plant early warning device and method adopting multiple prediction models |
| US11960271B2 (en) | 2018-10-12 | 2024-04-16 | Korea Hydro & Nuclear Power Co., Ltd | Power plant early warning device and method employing multiple prediction model |
| CN112508512A (en) * | 2020-11-26 | 2021-03-16 | 国网河北省电力有限公司经济技术研究院 | Power grid engineering cost data management method and device and terminal equipment |
| CN113011633A (en) * | 2021-02-05 | 2021-06-22 | 国网湖南省电力有限公司 | Overhead transmission line engineering cost prediction method, device and medium |
| CN113807613A (en) * | 2021-10-13 | 2021-12-17 | 国网安徽省电力有限公司 | Power grid engineering full-period budget optimization configuration tool |
| CN114066074A (en) * | 2021-11-19 | 2022-02-18 | 广东电网有限责任公司 | Method, device, equipment and medium for predicting maintenance frequency of old transformer assets |
| CN114169598A (en) * | 2021-11-29 | 2022-03-11 | 国网福建省电力有限公司 | Power transmission and transformation project progress early warning method considering complex environment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107067092A (en) | A kind of extra-high voltage electric transmission and transformation construction costs combination forecasting method and prediction meanss | |
| Zhang et al. | Hierarchically-coordinated voltage/VAR control of distribution networks using PV inverters | |
| Zhang et al. | Multi-agent safe policy learning for power management of networked microgrids | |
| Babacan et al. | Siting and sizing of distributed energy storage to mitigate voltage impact by solar PV in distribution systems | |
| Xu et al. | Multi-objective dynamic VAR planning against short-term voltage instability using a decomposition-based evolutionary algorithm | |
| Xu et al. | Optimal placement of static compensators for multi‐objective voltage stability enhancement of power systems | |
| Galvani et al. | Multiobjective predictability-based optimal placement and parameters setting of UPFC in wind power included power systems | |
| Reddy et al. | 2Index and GA based optimal location and sizing of distribution system capacitors | |
| Sedighizadeh et al. | Hybrid approach to FACTS devices allocation using multi-objective function with NSPSO and NSGA-II algorithms in Fuzzy framework | |
| Galvani et al. | Data clustering based probabilistic optimal power flow in power systems | |
| CN107368961A (en) | A kind of regional power grid carbon emission management method under the access background suitable for new energy | |
| CN109034559A (en) | Charging station distribution planning method, apparatus and computer equipment | |
| CN115800385A (en) | Electric energy quality regulation and control method based on adjustable and controllable capacity of photovoltaic inverter and charging pile | |
| Nascimento et al. | Multi-objective adaptive evolutionary algorithm to enhance voltage stability in power systems | |
| Pinto et al. | Reliability‐constrained robust expansion planning of active distribution networks | |
| CN112508254B (en) | Method for determining investment prediction data of transformer substation engineering project | |
| Gantayet et al. | Stochastic investigation for solid‐state transformer integration in distributed energy resources integrated active distribution network | |
| Wang et al. | Bi‐level optimisation dispatch method for photovoltaic hosting capacity enhancement of distribution buses | |
| CN106874607B (en) | Power grid self-organization critical state quantitative evaluation method based on multi-level variable weight theory | |
| CN108923413A (en) | Quality of voltage prediction technique based on support vector machines | |
| CN113919162B (en) | Voltage sag risk early warning method based on simulation and multi-source measured data fusion | |
| Lotfi et al. | Simultaneous placement of capacitor and DG in distribution networks using particle swarm optimization algorithm | |
| Dong et al. | Sensitivity‐based reliability coordination for power systems considering wind power reserve based on hybrid correlation control method for wind power forecast error | |
| CN114548828A (en) | Method, device and equipment for locating and fixing volume of distributed photovoltaic power supply | |
| Lu et al. | Prediction technology of power transmission and transformation project cost based on the decomposition-integration |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170818 |