CN103903071A - Wind power forecast combination method and system - Google Patents
Wind power forecast combination method and system Download PDFInfo
- Publication number
- CN103903071A CN103903071A CN201410155445.7A CN201410155445A CN103903071A CN 103903071 A CN103903071 A CN 103903071A CN 201410155445 A CN201410155445 A CN 201410155445A CN 103903071 A CN103903071 A CN 103903071A
- Authority
- CN
- China
- Prior art keywords
- wind power
- model
- prediction
- neural network
- artificial neural
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a wind power forecast combination method and system. A difference ARIMA forecast model and a BP-ANN forecast model are established through a time series method and a BP artificial neural network method respectively, then a new BP-ANN forecast model is established according to obtained forecast results, and finally the forecast value of the wind electricity power is obtained. According to the wind power forecast combination method and system, the advantages and disadvantages of two single models are fully considered and are combined by a novel combination mode, the different single models have complementary advantages, and forecast precision is further improved.
Description
Technical field
The present invention relates to power technology field, particularly a kind of wind power combination forecasting method and system.
Background technology
Wind-powered electricity generation has intermittence, randomness and undulatory property.Output of wind electric field shakiness, brings series of problems to dispatching of power netwoks, peak regulation, safety etc., and wind power prediction is the effective ways that overcome the above problems exactly.
Setting up wind power forecasting system has great significance for the daily operation of wind energy turbine set.According to the experience of the European wind-powered electricity generation developed country such as Germany, Denmark, if maintenance of fan all completes in little wind phase or calm spell, the annual generated energy of wind energy turbine set will improve 2%.In addition, generated output curve before wind power forecasting system can provide accurately day to grid company wind energy turbine set, this makes dispatching of power netwoks can effectively utilize wind power resources, improves wind-powered electricity generation generating online hour number.The most important thing is that wind energy turbine set generated output is the key factor that electrical network cannot be received wind-powered electricity generation on a large scale with the unordered variation of wind speed, therefore setting up wind power forecasting method and system is the effective means addressing this problem.
Summary of the invention
The object of the present invention is to provide a kind of wind power combination forecasting method and system, to solve output of wind electric field shakiness, bring a series of problem to dispatching of power netwoks, peak regulation, safety etc.
For solving the problems of the technologies described above, the invention provides a kind of wind power combination forecasting method, comprising:
Obtain wind speed and generate the first wind power by the prediction of difference ARMA model;
Obtaining wind speed is predicted and is generated the second wind power by back propagation artificial neural network model;
Carry out the final wind power of back propagation artificial neural network model prediction generation according to described the first wind power and the second wind power.
Preferably, in described wind power combination forecasting method, described difference ARMA model prediction generates the first wind power by time series method.
Preferably, in described wind power combination forecasting method, described time series method specifically comprises:
(1) determine sample sequence;
(2) auto-correlation and the partial correlation functional value of calculating sample sequence;
(3) whether steady by the figure difference judgement sample sequence of auto-correlation and partial correlation function, if turn to (5), directly turn to if not (4);
(4) carry out stationarity processing, the sequence of non-stationary is carried out to difference processing, data after treatment turn to (2), rejudge stationarity;
(5) utilize AIC and BIC to determine rank criterion and determine model order;
(6) utilize air speed data to try to achieve each model parameter estimation value;
(7) whether the residual sequence that judges model of fit is a white noise sequence, and suitable the proceeding to of model (8) if so, checked, proceeds to (5) if not;
(8) static prediction with above-mentioned set up forecast model, sample data being lagged behind 48 hours.
Preferably, in described wind power combination forecasting method, described back propagation artificial neural network model prediction generates the second wind power by BP artificial neural network method.
Preferably, in described wind power combination forecasting method, described BP artificial neural network method specifically comprises:
(1) determine sample data;
(2) data are normalized;
(3) given input vector and object vector;
(4) ask the output of hidden layer, the each node of output layer;
(5) ask the deviation between desired value and actual value;
(6) calculate reverse error;
(7) calculate forward each weights and the impact of threshold value on total error from last one deck, weights and threshold value are adjusted; Then turn to (5), repeatedly calculate, until reach within the scope of the predicated error of setting;
(8) predict with above-mentioned institute established model, obtain 48 power prediction values.
Preferably, in described wind power combination forecasting method, describedly carry out described back propagation artificial neural network model prediction according to described the first wind power and the second wind power and generate final wind power and comprise:
(1) choose suitable sample, set input data, target output data, the test data of described sample;
(2) set up new described back propagation artificial neural network model forecast model;
(3) utilize the described back propagation artificial neural network model of setting up to carry out wind power prediction, generate final wind power predicted value.
Accordingly, the present invention also provides a kind of wind power combined prediction system, comprising:
The first wind power generation module, generates the first wind power for obtaining wind speed by the prediction of difference ARMA model;
The second wind power generation module, generates the second wind power for obtaining wind speed by predicting by back propagation artificial neural network model;
Final wind power generation module, for carrying out being predicted and being generated final wind power by back propagation artificial neural network model according to described the first wind power and the second wind power.
Wind power combination forecasting method provided by the invention and system, there is following beneficial effect: the present invention sets up respectively difference autoregression moving average forecast model (ARIMA) and reverse transmittance nerve network forecast model (BP-ANN) by time series method and BP artificial neural network method, then utilize predicting the outcome of obtaining to set up again new BP-ANN forecast model, finally obtained the predicted value of wind power.The present invention has taken into full account the relative merits of two kinds of single models, and combines by new array mode, has not only realized the mutual supplement with each other's advantages of different single models, has also further improved precision of prediction.
Brief description of the drawings
Fig. 1 is the wind power combination forecasting method schematic diagram of the preferred embodiment of the present invention;
Embodiment
The wind power combination forecasting method and the system that the present invention are proposed below in conjunction with the drawings and specific embodiments are described in further detail.According to the following describes and claims, advantages and features of the invention will be clearer.It should be noted that, accompanying drawing all adopts very the form of simplifying and all uses non-ratio accurately, only in order to convenient, the object of the aid illustration embodiment of the present invention lucidly.
Please refer to Fig. 1, it is the wind power combination forecasting method schematic diagram of the preferred embodiment of the present invention.As shown in Figure 1, the invention provides a kind of wind power combination forecasting method, comprising:
Step 1: obtain wind speed and generate the first wind power by the prediction of difference ARMA model;
In this step, set up difference autoregression moving average forecast model (ARIMA) by time series method, generate the first wind power.
Specifically, described time series method comprises the following steps:
(1) determine sample sequence;
(2) auto-correlation and the partial correlation functional value of calculating sample sequence;
(3) whether steady by the figure difference judgement sample sequence of auto-correlation and partial correlation function, if turn to (5), directly turn to if not (4);
(4) carry out stationarity processing, the sequence of non-stationary is carried out to difference processing, data after treatment turn to (2), rejudge stationarity;
(5) utilize AIC and BIC to determine rank criterion and determine model order;
(6) utilize air speed data to try to achieve each model parameter estimation value;
(7) whether the residual sequence that judges model of fit is a white noise sequence, and suitable the proceeding to of model (8) if so, checked, proceeds to (5) if not;
(8) static prediction with above-mentioned set up forecast model, sample data being lagged behind 48 hours.
Step 2: obtain wind speed and predict and generate the second wind power by back propagation artificial neural network model;
In this step, set up reverse transmittance nerve network forecast model (BP-ANN) by BP artificial neural network method and generate the second wind power
Specifically, described BP artificial neural network method comprises the following steps:
(1) determine sample data;
(2) data are normalized;
(3) given input vector and object vector;
(4) ask the output of hidden layer, the each node of output layer;
(5) ask the deviation between desired value and actual value;
(6) calculate reverse error;
(7) calculate forward each weights and the impact of threshold value on total error from last one deck, weights and threshold value are adjusted; Then turn to (5), repeatedly calculate, until reach within the scope of the predicated error of setting;
(8) predict with above-mentioned institute established model, obtain 48 power prediction values.
Step 3: carry out the final wind power of back propagation artificial neural network model (BP-ANN) prediction generation according to described the first wind power and the second wind power.
Specifically, this step comprises the following steps:
(1) choose suitable sample, set input data, target output data, the test data of described sample;
(2) set up new BP-ANN forecast model;
(3) utilize the BP-ANN model of setting up to carry out wind power prediction, generate final wind power predicted value.
Accordingly, the present invention also provides a kind of wind power combined prediction system, comprising:
The first wind power generation module, generates the first wind power for obtaining wind speed by the prediction of difference ARMA model;
The second wind power generation module, generates the second wind power for obtaining wind speed by back propagation artificial neural network model (BP-ANN) prediction;
Final wind power generation module, for carrying out the final wind power of BP-ANN prediction generation according to described the first wind power and the second wind power.
Base this, the present invention sets up respectively difference autoregression moving average forecast model (ARIMA) and BP-ANN forecast model by time series method and BP artificial neural network method, then utilize predicting the outcome of obtaining to set up again new BP-ANN forecast model, finally obtained the predicted value of wind power.The present invention has taken into full account the relative merits of two kinds of single models, and combines by new array mode, has not only realized the mutual supplement with each other's advantages of different single models, has also further improved precision of prediction.
Foregoing description is only the description to preferred embodiment of the present invention, the not any restriction to the scope of the invention, and any change, modification that the those of ordinary skill in field of the present invention does according to above-mentioned disclosure, all belong to the protection domain of claims.
Claims (7)
1. a wind power combination forecasting method, is characterized in that, comprising:
Obtain wind speed and generate the first wind power by the prediction of difference ARMA model;
Obtaining wind speed is predicted and is generated the second wind power by back propagation artificial neural network model;
Carry out the final wind power of back propagation artificial neural network model prediction generation according to described the first wind power and the second wind power.
2. wind power combination forecasting method as claimed in claim 1, is characterized in that, described difference ARMA model prediction generates the first wind power by time series method.
3. wind power combination forecasting method as claimed in claim 2, is characterized in that, described time series method specifically comprises:
(1) determine sample sequence;
(2) auto-correlation and the partial correlation functional value of calculating sample sequence;
(3) whether steady by the figure difference judgement sample sequence of auto-correlation and partial correlation function, if turn to (5), directly turn to if not (4);
(4) carry out stationarity processing, the sequence of non-stationary is carried out to difference processing, data after treatment turn to (2), rejudge stationarity;
(5) utilize AIC and BIC to determine rank criterion and determine model order;
(6) utilize air speed data to try to achieve each model parameter estimation value;
(7) whether the residual sequence that judges model of fit is a white noise sequence, and suitable the proceeding to of model (8) if so, checked, proceeds to (5) if not;
(8) static prediction with above-mentioned set up forecast model, sample data being lagged behind 48 hours.
4. wind power combination forecasting method as claimed in claim 1, is characterized in that, described back propagation artificial neural network model prediction generates the second wind power by BP artificial neural network method.
5. wind power combination forecasting method as claimed in claim 4, is characterized in that, described BP artificial neural network method specifically comprises:
(1) determine sample data;
(2) data are normalized;
(3) given input vector and object vector;
(4) ask the output of hidden layer, the each node of output layer;
(5) ask the deviation between desired value and actual value;
(6) calculate reverse error;
(7) calculate forward each weights and the impact of threshold value on total error from last one deck, weights and threshold value are adjusted; Then turn to (5), repeatedly calculate, until reach within the scope of the predicated error of setting;
(8) predict with above-mentioned institute established model, obtain 48 power prediction values.
6. wind power combination forecasting method as claimed in claim 1, is characterized in that, describedly carries out back propagation artificial neural network model prediction according to described the first wind power and the second wind power and generates final wind power and comprise:
(1) choose suitable sample, set input data, target output data, the test data of described sample;
(2) set up new reverse transmittance nerve network and survey model;
(3) utilize the back propagation artificial neural network model of setting up to carry out wind power prediction, generate final wind power predicted value.
7. a wind power combined prediction system, is characterized in that, comprising:
The first wind power generation module, generates the first wind power for obtaining wind speed by the prediction of difference ARMA model;
The second wind power generation module, is predicted and is generated the second wind power by back propagation artificial neural network model for obtaining wind speed;
Final wind power generation module, for carrying out the final wind power of back propagation artificial neural network model prediction generation according to described the first wind power and the second wind power.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410155445.7A CN103903071A (en) | 2014-04-17 | 2014-04-17 | Wind power forecast combination method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410155445.7A CN103903071A (en) | 2014-04-17 | 2014-04-17 | Wind power forecast combination method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103903071A true CN103903071A (en) | 2014-07-02 |
Family
ID=50994381
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410155445.7A Pending CN103903071A (en) | 2014-04-17 | 2014-04-17 | Wind power forecast combination method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103903071A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104217260A (en) * | 2014-09-19 | 2014-12-17 | 南京信息工程大学 | Combined filling system for measured wind speed loss values of multiple neighboring wind motors in wind field |
| CN104239962A (en) * | 2014-08-07 | 2014-12-24 | 河海大学 | Regional power grid total wind power short-term prediction method based on correlation analysis |
| CN104376388A (en) * | 2014-12-08 | 2015-02-25 | 国家电网公司 | Ultra-short period wind power prediction method based on wind speed factor control model |
| CN106295908A (en) * | 2016-08-24 | 2017-01-04 | 上海电机学院 | A kind of SVM wind power forecasting method |
| CN106682754A (en) * | 2015-11-05 | 2017-05-17 | 阿里巴巴集团控股有限公司 | Event occurrence probability prediction method and device |
| CN109614742A (en) * | 2018-12-25 | 2019-04-12 | 中国海洋大学 | A kind of sea level height duration prediction algorithm |
| CN109961315A (en) * | 2019-01-29 | 2019-07-02 | 河南中烟工业有限责任公司 | A kind of monthly Method for Sales Forecast method of cigarette based on nonlinear combination model |
| CN112217858A (en) * | 2020-08-28 | 2021-01-12 | 北京思特奇信息技术股份有限公司 | Method and system for elastic expansion and contraction of cloud computing resources |
-
2014
- 2014-04-17 CN CN201410155445.7A patent/CN103903071A/en active Pending
Non-Patent Citations (2)
| Title |
|---|
| 方江晓: "短期风速和风电功率预测模型的研究", 《中国优秀硕士学位论文全文数据库（工程科技Ⅱ辑）》 * |
| 马蕊 等: "基于时间序列分析和神经网络的风电功率预测方法研究", 《大功率变流技术》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104239962A (en) * | 2014-08-07 | 2014-12-24 | 河海大学 | Regional power grid total wind power short-term prediction method based on correlation analysis |
| CN104217260A (en) * | 2014-09-19 | 2014-12-17 | 南京信息工程大学 | Combined filling system for measured wind speed loss values of multiple neighboring wind motors in wind field |
| CN104217260B (en) * | 2014-09-19 | 2017-08-22 | 南京信息工程大学 | A kind of wind field measures the combination fill system of wind speed defect value adjacent to many typhoon motors |
| CN104376388A (en) * | 2014-12-08 | 2015-02-25 | 国家电网公司 | Ultra-short period wind power prediction method based on wind speed factor control model |
| CN106682754A (en) * | 2015-11-05 | 2017-05-17 | 阿里巴巴集团控股有限公司 | Event occurrence probability prediction method and device |
| CN106295908A (en) * | 2016-08-24 | 2017-01-04 | 上海电机学院 | A kind of SVM wind power forecasting method |
| CN109614742A (en) * | 2018-12-25 | 2019-04-12 | 中国海洋大学 | A kind of sea level height duration prediction algorithm |
| CN109961315A (en) * | 2019-01-29 | 2019-07-02 | 河南中烟工业有限责任公司 | A kind of monthly Method for Sales Forecast method of cigarette based on nonlinear combination model |
| CN112217858A (en) * | 2020-08-28 | 2021-01-12 | 北京思特奇信息技术股份有限公司 | Method and system for elastic expansion and contraction of cloud computing resources |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103903071A (en) | Wind power forecast combination method and system | |
| CN105426956B (en) | A kind of ultra-short term photovoltaic prediction technique | |
| Saunders | Point estimate method addressing correlated wind power for probabilistic optimal power flow | |
| Brekken et al. | Optimal energy storage sizing and control for wind power applications | |
| Bossoufi et al. | Rooted tree optimization for the backstepping power control of a doubly fed induction generator wind turbine: dSPACE implementation | |
| CN103746370B (en) | A kind of wind energy turbine set Reliability Modeling | |
| CN104167765B (en) | A kind of maximum wind installed capacity computational methods based on the distribution of the ability of receiving | |
| CN106355344A (en) | Method for robustly and optimally operating micro-grids on basis of orthogonal arrays | |
| CN105429129A (en) | Evaluation method of intermittent energy generating capacity confidence considering network constraint | |
| CN105243259A (en) | Extreme learning machine based rapid prediction method for fluctuating wind speed | |
| CN104037755A (en) | Optimization method for solving Pareto solution sets of wind-storage-thermal joint operation system in multiple time periods | |
| Zahedi | Sustainable power supply using solar energy and wind power combined with energy storage | |
| CN104915788B (en) | A method of considering the Electrical Power System Dynamic economic load dispatching of windy field correlation | |
| CN106611243A (en) | Residual correction method for wind speed prediction based on GARCH (Generalized ARCH) model | |
| CN105956708A (en) | Grey correlation time sequence based short-term wind speed forecasting method | |
| CN114638396A (en) | Photovoltaic power prediction method and system based on neural network instantiation | |
| Prakash et al. | Modified interval‐based generator scheduling for PFR adequacy under uncertain PV generation | |
| Vohra et al. | End-to-end learning with multiple modalities for system-optimised renewables nowcasting | |
| Zhou et al. | Cost–benefit analysis of pumped hydro storage using improved probabilistic production simulation method | |
| Salcedo-Sanz et al. | A review of recent evolutionary computation-based techniques in wind turbines layout optimization problems | |
| Liang et al. | Study on economic operation for micro-grid based on scenario and PSO | |
| HAJIBASHI et al. | Markovian approach applied to reliability modeling of a wind farm | |
| Mohseni et al. | A novel method of maximum power point tracking for a SRG based wind power generation system using AI | |
| bin Ahmad et al. | Application of hybrid GMDH and Least Square Support Vector Machine in energy consumption forecasting | |
| Masoumi et al. | Wind speed forecasting using back propagation artificial neural networks in North of Iran |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140702 |
|
| WD01 | Invention patent application deemed withdrawn after publication |