CN109460917A - A kind of bus load prediction technique based on distribution factor and support vector machines - Google Patents
A kind of bus load prediction technique based on distribution factor and support vector machines Download PDFInfo
- Publication number
- CN109460917A CN109460917A CN201811325946.XA CN201811325946A CN109460917A CN 109460917 A CN109460917 A CN 109460917A CN 201811325946 A CN201811325946 A CN 201811325946A CN 109460917 A CN109460917 A CN 109460917A
- Authority
- CN
- China
- Prior art keywords
- load
- support vector
- bus
- region
- vector machines
- 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 50
- 238000012706 support-vector machine Methods 0.000 title claims abstract description 37
- 238000009826 distribution Methods 0.000 title claims abstract description 31
- 230000008676 import Effects 0.000 claims abstract description 3
- 240000002853 Nelumbo nucifera Species 0.000 claims description 5
- 235000006508 Nelumbo nucifera Nutrition 0.000 claims description 5
- 235000006510 Nelumbo pentapetala Nutrition 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 abstract description 4
- 238000010248 power generation Methods 0.000 abstract description 4
- 238000004519 manufacturing process Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 241001269238 Data Species 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000013277 forecasting method Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 230000003442 weekly effect Effects 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 231100000350 mutagenesis Toxicity 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
- G06Q10/06375—Prediction of business process outcome or impact based on a proposed change
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
-
- 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
- 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
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The bus load prediction technique based on distribution factor and support vector machines that the invention discloses a kind of, the specific steps are as follows: step 1, load data import；Step 2, load data pretreatment；Step 3 classifies bus load, determines each object in the hierarchical structure of system distribution model；Step 4, Support Vector Machines Optimized model parameter；With algorithm of support vector machine forecasting system total load；Step 5 calculates the breadth coefficient of each region load；Step 6 calculates the breadth coefficient of all types of loads；Step 7, according to Step 4: step 5 and step 6 data, solve the predicted value of bus load, can be completed.The present invention can preferably handle the emergency situations of various loads in bus load prediction, realize good prediction effect, and it is scientific horizontal with fining to improve load prediction comprehensively；Method of the invention can sufficiently meet the needs of operation of power networks lean management, the plan of reasonable arrangement production scheduling and implementation energy-saving power generation dispatching.
Description
Technical field
The present invention relates to power system load field, specifically a kind of bus based on distribution factor and support vector machines is negative
Lotus prediction technique.
Background technique
Electric power becomes the main energy sources in people's work and life, and electric load is the electrical equipment of electric power users a certain
The summation for the electrical power that moment takes to electric system, in order to guarantee the normal power supply in area, people can bear electric system
Lotus is predicted.Traditional Load Prediction In Power Systems include: short-term load forecasting (as unit of day), medium term load forecasting
(as unit of week and the moon) and long term load forecasting (as unit of season and year).
In general load prediction, the object of various Study on Forecasting Method is usually some region or some power grid system
System, demand history data be it is comprehensive, the single characteristic of various types of loads can not be screened, in actual prediction
There may be biggish errors in the process.
It is negative that the summation that the main transformer of substation supplies the end loads in a relatively small region may be defined as bus
Lotus.The prediction of bus load pointed by the present invention, it has the following characteristics that bus load is predicted compared with system loading is predicted
Load it is smaller, the part throttle characteristics of route outlet may be single, it is also possible to the synthesis of several load types.
For single load type, part throttle characteristics is quite obvious, but possible amplitude of variation is larger simultaneously, is not easy to grasp each period
Load level, such as steel plant's impact load；System median generatrix substantial amounts, and the changing rule of every bus has spy
Color can not be accomplished to analyze one by one；Due to being influenced by power supply area user behavior, bus load stability is poor, is also easy to produce prominent
Become；Part bus is not obvious vulnerable to small power supply power generation, the influence of load transfer and scheduled overhaul, load rule；Historical accumulation
Data it is not accurate enough, often have abnormal data appearance；By changes of operating modes, small power plant mounting, overhaul of the equipments, load transfer
Equal related factors are big.These features increase the difficulty of bus load prediction.
Main method employed in current loads prediction is: the prediction technique based on bus load Self-variation rule.
The core of this method be prediction model foundation and historical data it is comprehensive, Neither of the two can be dispensed.Wherein historical data
It is basis, prediction algorithm is core, and influence of the two to precision of prediction is all very big.But bus is negative unlike system loading
The changing rule of lotus itself is unstable, is also also easy to produce mutation, and causing to use for reference system loading prediction technique merely can not obtain very
High precision of prediction.
Summary of the invention
The bus load prediction technique based on distribution factor and support vector machines that the purpose of the present invention is to provide a kind of, with
Solve the problems mentioned above in the background art.
To achieve the above object, the invention provides the following technical scheme:
A kind of bus load prediction technique based on distribution factor and support vector machines, the specific steps are as follows:
Step 1, load data import；
Step 2, load data pretreatment；
Step 3 understands system overall situation, bus load is classified, in the hierarchical structure for determining system distribution model
Each object；
Step 4, according to history total load data, Support Vector Machines Optimized model parameter；It is pre- with algorithm of support vector machine
Examining system total load；
Step 5 determines the predicted value of future time instance region load according to the historical load data of each region load, according to
The predictor calculation of each region load goes out the breadth coefficient of each region load；
Step 6 determines the predicted value of the following all types of loads, according to each class according to the historical data of all types of loads
The predictor calculation of type load goes out the breadth coefficient of all types of loads；
Step 7, according to the breadth coefficient of each region load in the predicted value, step 5 of system total load in step 4
With the breadth coefficient of loads all types of in step 6, the predicted value of bus load is solved, can be completed.
As a further solution of the present invention: load of the algorithm of support vector machine for nonlinear change mode, prediction
Model are as follows:X is the whole network system total load historical forecast data；B is the dimension of selected input variable；f
It (x) is predicted load；It is the Nonlinear Mapping from the input space to high-dimensional feature space.
As a further solution of the present invention: load of the algorithm of support vector machine for nonlinear change mode, prediction
Model are as follows:K(xi,xj)=exp (- | | xi-xj||2/δ2), f (x) is predicted load, K
(xi,xj) it is Nonlinear Mapping from the input space to high-dimensional feature space, δ is kernel functional parameter, the constant being positive, αiTo draw
Ge Lang multiplier, b are the dimensions of selected input variable.
As a further solution of the present invention: determining the prediction of following each region load in step 5 using least square method
Value.
As a further solution of the present invention: the breadth coefficient of region load is KA,k,n+1(t), PA,k,i(t)
=k+q (n+1), C indicate the quantity that region is divided in forecasting system, and c is counting variable, PA,k,iIt (t) is i-th day region k in t
The region load at quarter, k and t are definite values in each calculate.
As a further solution of the present invention: determining the prediction of the following all types of loads in step 6 using least square method
Value.
As a further solution of the present invention: the breadth coefficient of type load is KB,l,n+1(t),PB,l,iIt (t) is i-th day load value in l type load t moment, L indicates to draw in forecasting system
Divide the quantity of load type, l is counting variable.
Compared with prior art, the beneficial effects of the present invention are:
Compared with traditional Power System Bus Load Forecasting method, the present invention can preferably handle bus load prediction
In various loads emergency situations, realize good prediction effect, improve that load prediction is scientific and fining is horizontal comprehensively；This
The method of invention can sufficiently meet the needs of operation of power networks lean management, the plan of reasonable arrangement production scheduling and implementation energy conservation
Power generation dispatching has good social benefit and prospect of the application.
Detailed description of the invention
Fig. 1 is the knot of tree-shaped constant load model in the bus load prediction technique based on distribution factor and support vector machines
Structure schematic diagram.
Fig. 2 is the inconsistent model of the bus load prediction technique intermediate load region based on distribution factor and support vector machines
Structural schematic diagram.
Fig. 3 is the model that load type is inconsistent in the bus load prediction technique based on distribution factor and support vector machines
Structural schematic diagram.
Fig. 4 is that the structure of mixing load model in the bus load prediction technique based on distribution factor and support vector machines is shown
It is intended to.
Fig. 5 is power grid first bus in somewhere in the bus load prediction technique based on distribution factor and support vector machines
Prediction result.
Fig. 6 is somewhere power grid Article 2 bus in the bus load prediction technique based on distribution factor and support vector machines
Prediction result.
Fig. 7 is somewhere power grid Article 3 bus in the bus load prediction technique based on distribution factor and support vector machines
Prediction result.
Fig. 8 is somewhere power grid Article 4 bus in the bus load prediction technique based on distribution factor and support vector machines
Prediction result.
Fig. 9 is somewhere power grid Article 5 bus in the bus load prediction technique based on distribution factor and support vector machines
Prediction result.
Figure 10 is that somewhere power grid Article 6 is female in the bus load prediction technique based on distribution factor and support vector machines
The prediction result of line.
Figure 11 is the work flow diagram of the bus load prediction technique based on distribution factor and support vector machines.
Specific embodiment
The technical solution of the patent is explained in further detail With reference to embodiment.
A kind of bus load prediction technique based on distribution factor and support vector machines mainly includes two main contents:
The whole network system loading prediction technique and bus load factor location mode.
In the whole network system loading forecast period, load prediction is carried out using algorithm of support vector machine.The study energy of this method
Power is stronger, it is not easy to fall into locally optimal solution, prediction effect is good.For the load of nonlinear change mode, prediction model
For
In formula, x is input variable, that is, the variable closely related with prediction, usually historical data, in the present invention
For the whole network system total load historical forecast data；D is the dimension of selected input variable；F (x) is predicted load；From defeated
Enter space to high-dimensional feature space Nonlinear Mapping.B and w is model parameter, equally will affect the precision of prediction.
Structure risk function is introduced below, converts the regression problem in (1) to the double optimization problem of formula (2):
The constraint condition of the problem are as follows:
In formula, N is the dimension of vector；eiRepresentative errors, e ∈ RN×1For error vector；γ is punishment parameter, and control is to accidentally
The punishment degree of difference, and it is greater than 0.yiFor the condition value of equality constraint.
In order to solve the optimization problem of above-mentioned formula (2), lagrange's method of multipliers is constructed, Lagrange factor, construction are introduced
Lagrangian:
In formula, αiFor Lagrange multiplier, and αi∈RN×1。
From Lagrangian optimal conditions:
It can further derive:
It is available from cancellation ω and e in formula (6):
In formula, E=[1,1 ..., 1]T, α=[α1,α2,...,αN]T, Y=[y1,y2,...,yN]T, R ∈ RN×N
It can be acquired by following formula:
In formula, K is the kernel function for the condition that meets, and the dot-product operation of high-order feature space is replaced with the kernel function in first space,
To make that simplification has entirely been calculated.
In this way, the expression formula of entire Nonlinear Prediction Models can simplify are as follows:
In formula, b, αiIt can be found out from formula (6).K(xi,xj) non-linear from the input space to high-dimensional feature space reflect
It penetrates.
Using the most common radial basis function as kernel function in the present invention:
K(xi,xj)=exp (- | | xi-xj||2/δ2) (10)
In formula, δ is kernel functional parameter, the constant being positive.
Supporting vector machine model needs to determine the cover width δ of punishment parameter γ and gaussian kernel function in modeling process.Its
In, γ is balance factor, can be according to the characteristic of sample data come the complexity of decision model and to the punishment degree of fitness bias；
δ reflects the associated degree between supporting vector, and δ value is small, and the connection between supporting vector is relatively loose, and Learning machine is relatively multiple
Miscellaneous, Generalization Ability cannot be guaranteed；δ value is big, and the influence between supporting vector is too strong, and regression model is difficult to reach enough accuracy.
In practical applications, γ and δ value is mainly arranged by rule of thumb, will affect precision of prediction, it is therefore necessary to they are optimized,
Reduce prediction error.
The present invention is as follows to the optimization method of two parameters during prediction: choosing nearest one week load data and makees
There are within one day 96 future positions using 15min as a prediction time for test sample, then share within one week 96 × 7 future positions,
The numerical value of continuous adjusting parameter, calculates the mean error of future position, final using the smallest parameter point of mean error as the present invention
Parameter value.
Bus load distributed model is to set up hierarchical relationship between system loading and each bus load, is distributed by adjusting
Coefficient is assigned as the predicted value of system loading the predicted value of every bus, the bus load of four kinds of more commonly used the type
Prediction model is the inconsistent model of the inconsistent model of tree-shaped constant load model, load area, load type, mixing respectively
Load model is described below respectively.
(1) tree-shaped constant load model
This model uses simplest one layer of branched structure, and the predicted value of bus load is each of system loading predicted value
A branch, this model directly set up breadth coefficient between the predicted value of system loading and the predicted value of bus load, and
Within each period, which is definite value.This model structure is simple, and the factor of consideration is less, applies in general to tie
The prediction occasion that structure is simple and precision of prediction is of less demanding uses.This kind of structure is shown in Fig. 1.
(2) the inconsistent model of load area
Under normal conditions, the bus load variation tendency of areal is identical, so bus load is pressed into geographic classification, and
Introduce region load concept can continue between each bus in the region same in this way and corresponding region load with tree-shaped
The method of Constant Model is analyzed, and between each region load and system loading, it usually sets up one and changes over time
Parameter.The structure chart of the model is shown in Fig. 2.
(3) the inconsistent model of load type
The inconsistent model of load type and the inconsistent model of load area are essentially identical, and difference is the method for bus classification
Difference, one is bus load is classified according to load type, another kind is that bus load is classified according to area.
The model enters the concept of type load, using the breadth coefficient changed over time, class between system loading and type load
Constant breadth coefficient is used between type load and corresponding bus load.Model structure is shown in Fig. 3
(4) load model is mixed
Comprehensively consider the influence of load type and regional factor to bus load, while increasing type load and region
The concept of load.First using the breadth coefficient changed over time in system loading distribution and each type load, next is adopted
With the breadth coefficient changed over time by type power load distributing into each region load, in a period of time, region load
Constant is generally used with the breadth coefficient of bus load.The structure chart of mixing load model is shown in Fig. 4 to increase of the invention lead to
With property, using mixing load model.
It is thus necessary to determine that parameter include type power load distributing coefficient and region power load distributing coefficient.With region load
It is illustrated for breadth coefficient, type power load distributing coefficient is similar therewith.
Assuming that first n days bus loads hourly are given data, by taking the inconsistent model of region load as an example, using most
Small two, which multiply model, analyzes historical data known to n days, and then predicts (n+1)th day each region burden apportionment coefficient.
If PA,k,iIt (t) is region load of i-th day region k in t moment.We utilize the primary song in least square model
Line y=k+qi calculates first n days historical datas, by the correlation formula of least square method, can obtain:
Wherein, k and t is definite value in each calculate.
We just obtain a curve y=k+qi in this way, this curve can be good at the variation of n days loads before characterizing
Trend, we predict n+1 days loads using this trend curve.N+1 days region loads can be obtained (to pay attention to not being final
Accurate predicted value) be
PA,k,i(t)=k+q (n+1) (13)
Bus load not only has date periodicity, also has weekly pattern, i-th day bus load is predicted, here
Before only choosing in n days the historical data of same type day as basis for forecasting, but it is such based on the amount of prediction data can be substantially
It reduces, increases influence of the bus load randomness to prediction result, so comprehensively considering the influence of weekly pattern and randomness, select
Two days historical datas are as the foundation predicted before and after same type day and same type day in n days before taking, so as to comprehensively consider
Influence of the various factors to bus load.
By changing the value of t, the repeated application above process, so that it may to determine n+1 days region load datas.Utilize this
A little data, so that it may to calculate the region sharing of load COEFFICIENT K of the n+1 days region k t momentsA,k,n+1(t):
C indicates the quantity that region is divided in forecasting system in formula, and c is counting variable.
The calculation method of type power load distributing coefficient is similar, the sharing of load coefficient of n+1 days l type load t moments
KB,l,n+1(t):
Wherein PB,l,iIt (t) is i-th day load value in l type load t moment.L indicates to divide load class in forecasting system
The quantity of type, l are counting variable.
After predicting the whole network system loading, bus load distribution factor is calculated further according to historical load data, just
The calculated value of each bus load can be calculated.
It chooses 6 buses of somewhere power grid to be predicted as case, solid line represents actual value in figure, and dotted line represents prediction
Value, Fig. 5-10 expression use the prediction result of the method for the present invention, predict standard from can be seen that 6 listed type loads in Fig. 5-10
True rate (average value) reaches 95% or more, it was demonstrated that the feasibility of the algorithm.
The working principle of the invention is: the present invention is allocated system loading using distribution factor method.Its forecast reason
It is: is predicted first with load of the prediction algorithm to system entirety, to obtains the load value of following a certain moment the whole network；
Then calculated load molecular group, system loading is distributed to every bus by the factor according to a certain percentage.In the whole network load prediction rank
Section, is predicted using algorithm of support vector machine, to reach good system loading precision of prediction；In the sharing of load stage, examine
Consider and sufficiently excavate the characteristic of bus load and the uncertainty of bus load, proposes the calculating based on the power load distributing factor
Method.
The present invention can preferably handle the emergency situations of various loads in bus load prediction, realize good prediction effect
It is scientific horizontal with fining to improve load prediction comprehensively for fruit.This method can sufficiently meet operation of power networks lean management
Demand, the plan of reasonable arrangement production scheduling and implementation energy-saving power generation dispatching.
It is obvious to a person skilled in the art that invention is not limited to the details of the above exemplary embodiments, Er Qie
In the case where without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter
From the point of view of which point, the present embodiments are to be considered as illustrative and not restrictive, and the scope of the present invention is by appended power
Benefit requires rather than above description limits, it is intended that all by what is fallen within the meaning and scope of the equivalent elements of the claims
Variation is included within the present invention.Any reference signs in the claims should not be construed as limiting the involved claims.
In addition, it should be understood that although this specification is described in terms of embodiments, but not each embodiment is only wrapped
Containing an independent technical solution, this description of the specification is merely for the sake of clarity, and those skilled in the art should
It considers the specification as a whole, the technical solutions in the various embodiments may also be suitably combined, forms those skilled in the art
The other embodiments being understood that.
Claims (7)
1. a kind of bus load prediction technique based on distribution factor and support vector machines, which is characterized in that specific step is as follows:
Step 1, load data import；
Step 2, load data pretreatment；
Step 3 understands system overall situation, bus load is classified, is determined each in the hierarchical structure of system distribution model
Object；
Step 4, according to history total load data, Support Vector Machines Optimized model parameter；System is predicted with algorithm of support vector machine
System total load；
Step 5 determines the predicted value of future time instance region load, according to each according to the historical load data of each region load
The predictor calculation of region load goes out the breadth coefficient of each region load；
Step 6 determines the predicted value of the following all types of loads according to the historical data of all types of loads, negative according to each type
The predictor calculation of lotus goes out the breadth coefficient of all types of loads；
Step 7, according to the breadth coefficient and step of each region load in the predicted value, step 5 of system total load in step 4
The breadth coefficient of all types of loads in rapid six, solves the predicted value of bus load, can be completed.
2. the bus load prediction technique according to claim 1 based on distribution factor and support vector machines, feature exist
In, load of the algorithm of support vector machine for nonlinear change mode, prediction model are as follows:x
It is the whole network system total load historical forecast data；B is the dimension of selected input variable；F (x) is predicted load；Be from
Nonlinear Mapping of the input space to high-dimensional feature space.
3. the bus load prediction technique according to claim 1 based on distribution factor and support vector machines, feature exist
In, load of the algorithm of support vector machine for nonlinear change mode, prediction model are as follows:K(xi,xj)=exp (- | | xi-xj||2/δ2), f (x) is predicted load, K (xi,xj) be
Nonlinear Mapping from the input space to high-dimensional feature space, δ are kernel functional parameter, the constant being positive, αiMultiply for Lagrange
Son, b are the dimensions of selected input variable.
4. the bus load prediction technique according to claim 1 or 2 based on distribution factor and support vector machines, feature
It is, determines the predicted value of following each region load in the step 5 using least square method.
5. the bus load prediction technique according to claim 4 based on distribution factor and support vector machines, feature exist
In the breadth coefficient of the region load is KA,k,n+1(t), PA,k,i(t)=k+q (n+1), C indicate the quantity that region is divided in forecasting system,
C is counting variable, PA,k,iIt (t) is i-th day region k in the region load of t moment, k and t are definite values in each calculate.
6. the bus load prediction technique according to claim 1 or 3 based on distribution factor and support vector machines, feature
It is, determines the predicted value of the following all types of loads in the step 6 using least square method.
7. the bus load prediction technique according to claim 6 based on distribution factor and support vector machines, feature exist
In the breadth coefficient of the type load is KB,l,n+1(t),PB,l,i(t) for i-th day in l class
The load value of type load t moment, L indicate the quantity that load type is divided in forecasting system, and l is counting variable.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811325946.XA CN109460917A (en) | 2018-11-08 | 2018-11-08 | A kind of bus load prediction technique based on distribution factor and support vector machines |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811325946.XA CN109460917A (en) | 2018-11-08 | 2018-11-08 | A kind of bus load prediction technique based on distribution factor and support vector machines |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109460917A true CN109460917A (en) | 2019-03-12 |
Family
ID=65609725
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811325946.XA Pending CN109460917A (en) | 2018-11-08 | 2018-11-08 | A kind of bus load prediction technique based on distribution factor and support vector machines |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109460917A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111708987A (en) * | 2020-06-16 | 2020-09-25 | 重庆大学 | Method for predicting load of multiple parallel transformers of transformer substation |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102930347A (en) * | 2012-10-15 | 2013-02-13 | 河海大学 | Method for forecasting short term load under demand response |
| CN103617564A (en) * | 2013-12-04 | 2014-03-05 | 国家电网公司 | Whole-network load prediction method based on local load predicted value comprehensive evaluation |
| CN104598999A (en) * | 2015-02-04 | 2015-05-06 | 河海大学 | Power distribution network time-of-use power price optimizing method taking reliability into consideration |
| CN105205547A (en) * | 2015-08-25 | 2015-12-30 | 广西电网有限责任公司 | Bus load prediction algorithm based on similarity matching of multiple uncertain factors |
| CN105529747A (en) * | 2016-01-06 | 2016-04-27 | 河海大学 | Modeling method for wind power admission power of power grid partitions through rationalized uniform redistribution |
| CN107527116A (en) * | 2017-08-28 | 2017-12-29 | 山东科技大学 | short-term load forecasting method based on support vector regression |
-
2018
- 2018-11-08 CN CN201811325946.XA patent/CN109460917A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102930347A (en) * | 2012-10-15 | 2013-02-13 | 河海大学 | Method for forecasting short term load under demand response |
| CN103617564A (en) * | 2013-12-04 | 2014-03-05 | 国家电网公司 | Whole-network load prediction method based on local load predicted value comprehensive evaluation |
| CN104598999A (en) * | 2015-02-04 | 2015-05-06 | 河海大学 | Power distribution network time-of-use power price optimizing method taking reliability into consideration |
| CN105205547A (en) * | 2015-08-25 | 2015-12-30 | 广西电网有限责任公司 | Bus load prediction algorithm based on similarity matching of multiple uncertain factors |
| CN105529747A (en) * | 2016-01-06 | 2016-04-27 | 河海大学 | Modeling method for wind power admission power of power grid partitions through rationalized uniform redistribution |
| CN107527116A (en) * | 2017-08-28 | 2017-12-29 | 山东科技大学 | short-term load forecasting method based on support vector regression |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111708987A (en) * | 2020-06-16 | 2020-09-25 | 重庆大学 | Method for predicting load of multiple parallel transformers of transformer substation |
| CN111708987B (en) * | 2020-06-16 | 2023-04-07 | 重庆大学 | Method for predicting load of multiple parallel transformers of transformer substation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhong et al. | Optimal energy management for multi-energy multi-microgrid networks considering carbon emission limitations | |
| He et al. | Urban long term electricity demand forecast method based on system dynamics of the new economic normal: the case of Tianjin | |
| Reneses et al. | Distribution pricing: theoretical principles and practical approaches | |
| Yuan et al. | A hybrid prediction-based microgrid energy management strategy considering demand-side response and data interruption | |
| Yi et al. | Coordinated operation strategy for a virtual power plant with multiple DER aggregators | |
| Solanke et al. | Control and management of a multilevel electric vehicles infrastructure integrated with distributed resources: A comprehensive review | |
| CN101888087A (en) | Method for realizing distributed super-short-term area load forecasting in distribution network terminal | |
| CN104638636B (en) | A kind of electric power daily load characteristic index Forecasting Methodology | |
| Arcos-Vargas et al. | Residential peak electricity management. A storage and control systems application taking advantages of smart meters | |
| CN106355285B (en) | Power load prediction method based on parameter correction | |
| Berredo et al. | Decision making in fuzzy environment and multicriteria power engineering problems | |
| CN105243609A (en) | Orderly power utilization decision method of user on the basis of fuzzy analytic hierarchy process | |
| Yang et al. | Optimal bidding strategy of renewable-based virtual power plant in the day-ahead market | |
| Yu et al. | Optimal scheduling strategy of cascade hydropower plants under the joint market of day-ahead energy and frequency regulation | |
| Wen et al. | Optimal intra-day operations of behind-the-meter battery storage for primary frequency regulation provision: A hybrid lookahead method | |
| Kim et al. | Economical energy storage systems scheduling based on load forecasting using deep learning | |
| Li et al. | Operator-as-a-consumer: A novel energy storage sharing approach under demand charge | |
| Yin et al. | Collaborative decision-making model for capacity allocation of photovoltaics energy storage system under Energy Internet in China | |
| Wei et al. | Day-ahead optimal dispatch of a virtual power plant in the joint energy-reserve-carbon market | |
| Aguilar et al. | Intent profile strategy for virtual power plant participation in simultaneous energy markets with dynamic storage management | |
| CN109460917A (en) | A kind of bus load prediction technique based on distribution factor and support vector machines | |
| Heggie et al. | Power rationing in a long-term power shortage | |
| Zeng et al. | Optimal Planning of IDR-Integrated Electricity-Heat Systems Considering Techno-Economic and Carbon Emission Issues | |
| Dash et al. | A bilayer clustered-priority-driven energy management model for inclining block rate tariff environment | |
| Maldonado et al. | Optimal planning of collective photovoltaic arrays in energy communities through a multi-cut benders’ decomposition strategy |
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 |