CN106529706A - Support-vector-machine-regression-based method for predicting wind speed of wind power plant - Google Patents
Support-vector-machine-regression-based method for predicting wind speed of wind power plant Download PDFInfo
- Publication number
- CN106529706A CN106529706A CN201610936945.3A CN201610936945A CN106529706A CN 106529706 A CN106529706 A CN 106529706A CN 201610936945 A CN201610936945 A CN 201610936945A CN 106529706 A CN106529706 A CN 106529706A
- Authority
- CN
- China
- Prior art keywords
- data
- wind
- support vector
- training
- vector regression
- 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 38
- 238000012549 training Methods 0.000 claims abstract description 25
- 238000012360 testing method Methods 0.000 claims abstract description 14
- 230000006870 function Effects 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 238000002790 cross-validation Methods 0.000 claims description 2
- 230000001373 regressive effect Effects 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims description 2
- 230000002123 temporal effect Effects 0.000 claims description 2
- 241001269238 Data Species 0.000 claims 1
- 238000012706 support-vector machine Methods 0.000 abstract description 24
- 230000005684 electric field Effects 0.000 description 4
- 230000005611 electricity Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000013277 forecasting method Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000009987 spinning 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
- G06Q10/04—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
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)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Wind Motors (AREA)
Abstract
Disclosed in the invention is a support-vector-machine-regression-based method for predicting a wind speed of a wind power plant. The invention relates to a support-vector-machine-regression-based method for predicting a wind speed of a wind power plant, thereby solving problems of low prediction precision and slow convergence speed of the existing wind speed prediction method. The method comprises the following steps: step one, selecting sample data collected by a wind power plant; step two, determining a sample training set and a testing set; step three, carrying out pretreatment on the sample data; step four, selecting a support vector machine (SVM) regression kernel function and determining a to-be-optimized parameter of an SVM model; and step five, training the SVM model by using an optimal parameter and predicting a wind speed value at a future time. The method is applied to the wind power prediction field.
Description
Technical field
The present invention relates to the method for forecasting based on Support vector regression.
Background technology
Development wind-powered electricity generation, to improving energy resource structure, preserving the ecological environment, ensure that clean energy is safe and realizes holding for economy
The aspects such as supervention exhibition have extremely important meaning, and this has become global common recognition.But, current Wind turbines output work
The characteristics of rate is that, with intermittence, non-linear, pace of change is fast, the features such as fluctuation range is larger.Wind-electricity integration is to the quality of power supply
And the very tremendous influence that power system has.Want to realize that the scale of wind-powered electricity generation is utilized, optimize dispatching of power netwoks, strengthen wind
The electric market competitiveness, wind energy turbine set must carry out wind power prediction forecast, should possess daily forecast and real-time prediction ability.
Therefore, Accurate Prediction is carried out to wind power output power, especially ultra-short term, short-term forecast, wind-powered electricity generation can be improved simultaneously
Impact of the net to power system, formulates more rational generation schedule, reduces spinning reserve and operating cost and wind to electric field
Electric field participates in competition in power generation and all has important function.Wind power prediction is used as the key played a significant role in wind-electricity integration
Technology, is the problem of urgent need to resolve, has broad application prospects.
The content of the invention
The invention aims to it is low slow with convergence rate to solve existing short wind speed forecasting method precision of prediction
Problem, and propose a kind of method for forecasting based on Support vector regression.
A kind of method for forecasting based on Support vector regression comprises the steps:
Step one, the sample data for choosing wind energy turbine set collection;
Step 2, determine sample training collection and test set;
Step 3, pretreatment is carried out to sample data;
Step 4, selection Support vector regression (SVM) kernel function, determine that SVM models treat optimizing parameter;
Step 5, using optimal parameter train SVM models, predict future time instance air speed value.
Beneficial effects of the present invention are:
The present invention relates to a kind of method for forecasting based on Support vector regression, the present invention is by choosing wind
The sample data of electric field collection;Determine sample training collection and test set;Pretreatment is carried out to sample data;Select support vector machine
(SVM) kernel function is returned, determines that SVM models treat optimizing parameter;SVM models are trained using optimal parameter, future time instance is predicted
Air speed value.The aspects such as the convergence rate of precision of prediction and Forecasting Methodology in wind speed load all have been improved, and the method has weight
The realistic meaning wanted and application prospect.
Description of the drawings
Fig. 1 is the method for forecasting block diagram based on Support vector regression;
Fig. 2 is a kind of SVM model parameters optimization method flow chart.
Specific embodiment
Specific embodiment one:A kind of method for forecasting based on Support vector regression of present embodiment
Comprise the steps:
Step one, the sample data for choosing wind energy turbine set collection;
Step 2, sample training collection and test set are determined according to sample data;
Step 3, pretreatment is carried out to sample data;
Step 4, selection Support vector regression (SVM) kernel function, determine that SVM models treat optimizing parameter, are most preferably joined
Number training SVM models;
Step 5, using optimal parameter train SVM models, predict future time instance air speed value.
Specific embodiment two:Present embodiment from unlike specific embodiment one:Wind is chosen in the step one
The sample data of electric field collection;Detailed process is:
The sample data of wind energy turbine set collection is to set up anemometer tower in the key position point of wind energy turbine set, mainly studies wind-power electricity generation
Following 10 minutes, the 30 minutes and 1 hour real-time estimate of unit output, temporal resolution are 10min;So right first
Wind turbines Performance Characteristics is analyzed.
From the ambient temperature that the wind power generating set sampling interval is 10min, wind speed and output historical average evidence.
Other steps and parameter are identical with specific embodiment one.
Specific embodiment three:Present embodiment from unlike specific embodiment one or two:Root in the step 2
Sample training collection and test set are determined according to sample data;Detailed process is:
No. 1 Wind turbines real-time running data of 7 days is chosen in experiment, represents that (3 row are respectively environment with the matrix of N × 3
Temperature, wind speed, output), build regressive prediction model;Consider above-mentioned analysis, by emulation experiment, with front 5 days 720 samples
Training set of the notebook data as SVM models, the power definition of wind power generating set is:
PS=1/2 ρ v3fCp (1)
Wherein PSFor wind power generating set performance number, unit is W, and ρ is atmospheric density, and unit is kg/m3, v numbers are to carry out flow velocity
Degree, unit is m/s, and f is area, unit m2, CpFor power coefficient, refer to that wind energy conversion system absorbs energy from natural wind energy big
Little degree.
Other steps and parameter are identical with specific embodiment one or two.
Specific embodiment four:Unlike one of present embodiment and specific embodiment one to three:The step 3
In pretreatment is carried out to sample data;Detailed process is:
Wind power generating set not all moment is all in operational excellence state, it is therefore desirable to reject some improper shapes
Condition data, including unit breaks down shutdown, power is in negative value, and data collecting system is unstable to cause the presence of the feelings such as abnormity point
Condition.In addition, also training set and test set data to be normalized, makes initial data by regular to [0,1] model
In enclosing, unified computing.
(1) missing data is processed, when lacking for individual data, its value is generally in moment load before and after its correspondence
Intermediate value, can be described using linear interpolation method;
(2) data normalization, computing formula are as follows:
Y=(X-Xmin)/(Xmax-Xmin) (2)
Y represents the value after normalization, Xmax, XminThe maximum and minima of respectively original output parameter X.
Other steps and parameter are identical with one of specific embodiment one to three.
Specific embodiment five:Unlike one of present embodiment and specific embodiment one to four:The SVM models
Training set be changed into 600 × 3 matrix data, test set is changed into 120 × 3 matrix data.
Other steps and parameter are identical with one of specific embodiment one to four.
Specific embodiment six:Unlike one of present embodiment and specific embodiment one to five:The step 4
Middle selection Support vector regression (SVM) kernel function, determines that SVM models treat optimizing parameter, obtains optimal parameter training SVM moulds
Type;Detailed process is:
Described selection Support vector regression (SVM) kernel function, determines that SVM models treat optimizing parameter, insensitive coefficient ε
Excursion is less, affects also less to model, therefore can set fixed value as 10-2, to penalty factor and RBF kernel functional parameter σ
Carry out optimizing selection;Extensive search is carried out using cross validation grid data service, finally selecting makes training pattern error minimum
Parameter combination.
Other steps and parameter are identical with one of specific embodiment one to five.
Specific embodiment seven:Unlike one of present embodiment and specific embodiment one to six:The step 5
Middle utilization optimal parameter trains SVM models, predicts the air speed value of future time instance;Detailed process is:
Described input sample test data, predicts the air speed value of future time instance, and the input quantity of training pattern is upper a period of time
The wind speed at quarter, ambient temperature and output, the output of model is subsequent time unit output.So as to use direct method
Carry out wind power prediction;Regression machine model is obtained by training set data learning training, using the measurable unit of the model not
Carry out the output at moment;
According to optimal solution structure forecast function:
Input sample test data predicts the air speed value of subsequent time using anticipation function f (x).
Other steps and parameter are identical with one of specific embodiment one to six.
Claims (7)
1. a kind of method for forecasting based on Support vector regression, it is characterised in that:It is a kind of to be based on supporting vector
The method for forecasting that machine is returned comprises the steps:
Step one, the sample data for choosing wind energy turbine set collection;
Step 2, sample training collection and test set are determined according to sample data;
Step 3, pretreatment is carried out to sample data;
Step 4, selection Support vector regression kernel function, determine that SVM models treat optimizing parameter, obtain optimal parameter training SVM
Model;
Step 5, using optimal parameter train SVM models, predict future time instance air speed value.
2. a kind of method for forecasting based on Support vector regression according to claim 1, it is characterised in that:
The sample data of wind energy turbine set collection is chosen in the step one;Detailed process is:
The sample data of wind energy turbine set collection is to set up anemometer tower in wind energy turbine set, main research wind power generating set output future
The real-time estimate of 10 minutes, 30 minutes and 1 hour, temporal resolution are 10min;
From the ambient temperature that the wind power generating set sampling interval is 10min, wind speed and output historical average evidence.
3. a kind of method for forecasting based on Support vector regression according to claim 2, it is characterised in that:
Sample training collection and test set are determined according to sample data in the step 2;Detailed process is:
The Wind turbines real-time running data of 7 days is chosen, is represented with the matrix of N × 3, is built regressive prediction model;With first 5 days
Training set of 720 sample datas as SVM models, the power definition of wind power generating set is:
PS=1/2 ρ v3fCp (1)
Wherein PSFor wind power generating set performance number, unit is W, and ρ is atmospheric density, and unit is kg/m3, v numbers are speed of incoming flow,
Unit is m/s, and f is area, unit m2, CpFor power coefficient, refer to the size journey that wind energy conversion system absorbs energy from natural wind energy
Degree.
4. a kind of method for forecasting based on Support vector regression according to claim 3, it is characterised in that:
Pretreatment is carried out to sample data in the step 3;Detailed process is:
(1) missing data is processed, during for shortage of data, is described using linear interpolation method;
(2) data normalization, computing formula are as follows:
Y=(X-Xmin)/(Xmax-Xmin) (2)
Y represents the value after normalization, Xmax, XminThe maximum and minima of respectively original output parameter X.
5. a kind of method for forecasting based on Support vector regression according to claim 4, it is characterised in that:
The training set of the SVM models is changed into 600 × 3 matrix data, and test set is changed into 120 × 3 matrix data.
6. a kind of method for forecasting based on Support vector regression according to claim 5, it is characterised in that:
Support vector regression kernel function is selected in the step 4, determines that SVM models treat optimizing parameter, obtain optimal parameter training
SVM models;Detailed process is:
Described selection Support vector regression kernel function, determines that SVM models treat optimizing parameter, and insensitive coefficient ε values are 10-2,
Optimizing selection is carried out to penalty factor and RBF kernel functional parameters σ;Scanned for using cross validation grid data service, it is final to select
The minimum parameter combination of training pattern of sening as an envoy to error.
7. a kind of method for forecasting based on Support vector regression according to claim 6, it is characterised in that:
SVM models are trained using optimal parameter in the step 5, the air speed value of future time instance is predicted;Detailed process is:
Described input sample test data, predicts the air speed value of future time instance, and the input quantity of training pattern was a upper moment
Wind speed, ambient temperature and output, the output of model is subsequent time unit output.So as to be carried out using direct method
Wind power prediction;Regression machine model is obtained by training set data learning training, during unit future measurable using the model
The output at quarter;
According to optimal solution structure forecast function:
Input sample test data predicts the air speed value of subsequent time using anticipation function f (x).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610936945.3A CN106529706A (en) | 2016-10-25 | 2016-10-25 | Support-vector-machine-regression-based method for predicting wind speed of wind power plant |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610936945.3A CN106529706A (en) | 2016-10-25 | 2016-10-25 | Support-vector-machine-regression-based method for predicting wind speed of wind power plant |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106529706A true CN106529706A (en) | 2017-03-22 |
Family
ID=58292137
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610936945.3A Pending CN106529706A (en) | 2016-10-25 | 2016-10-25 | Support-vector-machine-regression-based method for predicting wind speed of wind power plant |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106529706A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106960260A (en) * | 2017-03-27 | 2017-07-18 | 深圳汇创联合自动化控制有限公司 | A kind of wind power forecasting system for being easy to power scheduling |
CN106979126A (en) * | 2017-04-12 | 2017-07-25 | 浙江大学 | Wind power generating set high wind speed section effective wind speed method of estimation based on SVR |
CN107045574A (en) * | 2017-04-12 | 2017-08-15 | 浙江大学 | The low wind speed section effective wind speed method of estimation of wind power generating set based on SVR |
CN110685857A (en) * | 2019-10-16 | 2020-01-14 | 湘潭大学 | Mountain wind turbine generator behavior prediction model based on ensemble learning |
CN110942182A (en) * | 2019-11-14 | 2020-03-31 | 国网福建省电力有限公司建设分公司 | Method for establishing typhoon prediction model based on support vector regression |
CN110985290A (en) * | 2019-12-04 | 2020-04-10 | 浙江大学 | Optimal torque control method based on support vector regression |
CN110992101A (en) * | 2019-12-05 | 2020-04-10 | 中国铁道科学研究院集团有限公司电子计算技术研究所 | Station advertisement media resource value and income prediction regression method and prediction model |
-
2016
- 2016-10-25 CN CN201610936945.3A patent/CN106529706A/en active Pending
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106960260A (en) * | 2017-03-27 | 2017-07-18 | 深圳汇创联合自动化控制有限公司 | A kind of wind power forecasting system for being easy to power scheduling |
CN106979126A (en) * | 2017-04-12 | 2017-07-25 | 浙江大学 | Wind power generating set high wind speed section effective wind speed method of estimation based on SVR |
CN107045574A (en) * | 2017-04-12 | 2017-08-15 | 浙江大学 | The low wind speed section effective wind speed method of estimation of wind power generating set based on SVR |
CN106979126B (en) * | 2017-04-12 | 2019-01-29 | 浙江大学 | Wind power generating set high wind speed section effective wind speed estimation method based on SVR |
CN107045574B (en) * | 2017-04-12 | 2020-02-28 | 浙江大学 | SVR-based effective wind speed estimation method for low wind speed section of wind generating set |
CN110685857A (en) * | 2019-10-16 | 2020-01-14 | 湘潭大学 | Mountain wind turbine generator behavior prediction model based on ensemble learning |
CN110942182A (en) * | 2019-11-14 | 2020-03-31 | 国网福建省电力有限公司建设分公司 | Method for establishing typhoon prediction model based on support vector regression |
CN110985290A (en) * | 2019-12-04 | 2020-04-10 | 浙江大学 | Optimal torque control method based on support vector regression |
CN110985290B (en) * | 2019-12-04 | 2022-02-11 | 浙江大学 | Optimal torque control method based on support vector regression |
CN110992101A (en) * | 2019-12-05 | 2020-04-10 | 中国铁道科学研究院集团有限公司电子计算技术研究所 | Station advertisement media resource value and income prediction regression method and prediction model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106529706A (en) | Support-vector-machine-regression-based method for predicting wind speed of wind power plant | |
Yang et al. | Day-ahead wind power forecasting based on the clustering of equivalent power curves | |
CN102479339B (en) | Method and system for forecasting short-term wind speed of wind farm based on hybrid neural network | |
CN105356492B (en) | Energy management simulation system and method suitable for micro-grid | |
Kou et al. | Photovoltaic power forecasting based on artificial neural network and meteorological data | |
CN106979126B (en) | Wind power generating set high wind speed section effective wind speed estimation method based on SVR | |
CN105654207A (en) | Wind power prediction method based on wind speed information and wind direction information | |
CN105809293A (en) | Multi-model combined prediction method for short-term power of wind farm | |
CN110880789A (en) | Economic dispatching method for wind power and photovoltaic combined power generation system | |
CN104992248A (en) | Microgrid photovoltaic power station generating capacity combined forecasting method | |
CN105303250A (en) | Wind power combination prediction method based on optimal weight coefficient | |
CN105404937A (en) | Photovoltaic plant short-term power prediction method and system | |
CN103295077B (en) | A kind of wind power plant cluster dispatching method considering forecast error distribution character | |
CN105787594A (en) | Irradiation prediction method based on multivariate time series and regression analysis | |
CN106446440A (en) | Short-term photovoltaic generation power prediction method based on online sequential extreme learning machine | |
CN107045574A (en) | The low wind speed section effective wind speed method of estimation of wind power generating set based on SVR | |
CN104102951A (en) | Short-term wind power prediction method based on EMD historical data preprocessing | |
CN106611243A (en) | Residual correction method for wind speed prediction based on GARCH (Generalized ARCH) model | |
CN114021848A (en) | Generating capacity demand prediction method based on LSTM deep learning | |
Luo et al. | Short-term photovoltaic generation forecasting based on similar day selection and extreme learning machine | |
Zhang et al. | Short-term wind power prediction based on EMD-LSTM combined model | |
Xiaolan et al. | One-month ahead prediction of wind speed and output power based on EMD and LSSVM | |
CN110147908A (en) | A kind of wind power forecasting method based on three-dimensional optimal similarity and improvement cuckoo algorithm | |
Xiaojuan et al. | Short-time wind speed prediction for wind farm based on improved neural network | |
CN112836876A (en) | Power distribution network line load prediction method based on deep learning |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170322 |