CN103996071A - Wind power plant wind speed prediction method based on Markov theory - Google Patents
Wind power plant wind speed prediction method based on Markov theory Download PDFInfo
- Publication number
- CN103996071A CN103996071A CN201410062993.5A CN201410062993A CN103996071A CN 103996071 A CN103996071 A CN 103996071A CN 201410062993 A CN201410062993 A CN 201410062993A CN 103996071 A CN103996071 A CN 103996071A
- Authority
- CN
- China
- Prior art keywords
- wind speed
- data
- wind
- speed data
- air speed
- 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
Abstract
The invention discloses a wind power plant wind speed prediction method, which comprises a short-term wind speed prediction module, a real-time data module, a historical wind speed data module, a management system, a data processing system and a graph user interface, wherein the short-term wind speed prediction module is used for predicting the wind speed value of a wind power plant within 0-48 hours; real-time wind speed data transmitted from a data acquisition system is received by the real-time data module according to socket communication, and can be used as the input value of the prediction model and the measurement criteria of the prediction precision of the prediction module; the historical wind speed data module can be connected with the prediction module to provide necessary information for prediction; the management system, the data processing system and the graph user interface can maintain various databases, and display the prediction information in a graph or a chart. According to the wind power plant wind speed prediction method disclosed by the invention, the historical wind speed data of the wind power plant can be fully utilized, the wind speed data of the next stage can be predicted by the processing analysis of the wind speed data of the latest stage, the technical problem of poor wind speed prediction precision can be effectively solved, and the wind power plant wind speed prediction method has strong practicality.
Description
Technical field
The invention belongs to forecasting wind speed technical field, relate to a kind of Forecasting Methodology of wind farm wind velocity.
Background technology
In recent years, wind-power electricity generation, as huge, pollution-free, the reproducible clean environment firendly new forms of energy of a kind of reserves, has obtained global concern, and increasing wind power plant builds up.But due to the strong randomness of wind energy, the strong characteristic such as intermittent, wind-power electricity generation must leave enough standby unit and peak, the stability of system when guaranteeing that fluctuating widely appears in wind energy.Therefore, in order to guarantee the utilization factor of wind energy, more and more wind-power electricity generation enterprise needs data acquisition system (DAS) and prognoses system accurately, by the historical data gathering, carries out wind resource prediction, so that a regulation and control minute power distributing amount.Data acquisition system (DAS) and prognoses system can effectively be carried out wind energy turbine set addressing accurately simultaneously, to better realize wind energy turbine set, detect.
With regard to forecasting wind speed, at present the prediction of wind speed is mainly depended on to physical prediction model, calculated amount is large, error accumulation rate is high, this model needs professional person to safeguard, in common wind-powered electricity generation enterprise, do not possess good popularized type, and the predicted value area coverage of this model is larger, can not makes prediction accurately.In recent years, persistence forecasting algorithm, Kalman filtering algorithm, time series algorithm and neural network algorithm rise gradually.The historical data that these forecasting wind speed algorithms only need produce wind-powered electricity generation is set up model and can be predicted.Above algorithm operating is simple, but has larger relatine error for prediction, and predicated error can reach 20% conventionally.Affected by the multinomial factors such as temperature, air pressure, there is very strong randomness.For reducing wind energy turbine set operating cost, must improve forecasting wind speed precision.Those skilled in the art possesses high-precision forecasting wind speed algorithm in research always, solves the large difficult problem of prediction of wind speed error, not yet succeeds so far.
Summary of the invention
Poor in order to solve above-mentioned forecasting wind speed precision, mainly rely on the series of problems such as physical model and later maintenance change trouble, the present invention proposes a kind of wind farm wind velocity multistep forecasting method.
The method is integrated historical time survey wind data, the survey wind data of current season and the real-time survey wind data in nearly stage same season as forecast model Data Source, utilize Algorithms of Wavelet Analysis that the air speed data in training book season is decomposed into trend subitem, cycle subitem certificate and random subitem.Utilize arma modeling algorithm to carry out independent analysis modeling and utilize the air speed data in nearly stage to predict above different subitem data, after each subitem data prediction value of weighted stacking, obtain predicted value one time.For eliminating the strong randomness of wind speed, utilize Markov switching matrix to carry out a predicted data correction, to guarantee the Accurate Prediction of following wind speed variation tendency.
The present invention includes following operation steps:
(1) read in same season air speed data in historical time, obtain a step transition probability matrix of this area wind speed state;
(2) read in the air speed data of current season, to obtain the training sample sequence of forecasting wind speed model;
(3) utilize wavelet decomposition and reconstruct that training sample sequence transformation is become
nlayer low frequency air speed data and 1
Floor height is air speed data frequently, more than utilization
n+ 1 layer of training sample data are set up
n+ 1 arma modeling;
(4) read the air speed data in nearly stage, and utilize wavelet decomposition and reconstruct nearly the air speed data in stage be for conversion into
nlayer low frequency air speed data and 1 floor height frequently, utilize the arma modeling in step 3 respectively multilayer low-frequency data and high-frequency data to be carried out to hierarchical prediction, can obtain one time predicted value after hierarchical prediction result weighted stacking;
(5) utilize Markov one step transition probability matrix to carry out a predicted value correction, obtain finally and predict the outcome.
An aspect according to the proposed method, the described historical time with season air speed data, the air speed data of current season, the air speed data in nearly stage be all the air speed data at a certain specific predicted time interval, the described particular prediction time interval can be 15 ~ 60min.
An aspect according to the proposed method, the value of predicting the outcome of described Forecasting Methodology compares with the real-time Wind observation value gathering, and uses computing formula
, calculate the error between described real-time Wind observation value and forecasting wind speed value, wherein
for
treal-time Wind observation value constantly,
for
tforecasting wind speed value constantly, when continuously certain specific times
nrelatively in, described forecast model value of predicting the outcome surpasses 10% with the average relative error of the real-time Wind observation value gathering, and adjusts the arma modeling of described Forecasting Methodology, rebuilds arma modeling forecast model.Wherein,
nfor natural number.
An aspect according to the proposed method, described step 1 specifically comprises: the same season air speed data information that reads the historical time is
x 0=
x 1,
x 2,
x t , with certain fixing wind speed interval
wcarry out the division of wind speed state space, i.e. wind speed state
nair speed data scope be [
w* (
n-1),
w*
n] (
wvariable).With above spatiality data construct wind speed state one step transition probability matrix
p ;
An aspect according to the proposed method, described step 3 specifically comprises: this air speed data that reads the current year is in season
v 0=
v 1,
v 2,
v n , choose discrete wavelet analysis this air speed data is carried out
nlayer Wavelet fast decomposition and reconstruct, and with after reconstruct
nlayer low frequency signal and 1 layer of high-frequency signal, as the training sample of arma modeling, obtain the ARMA forecast model of wind speed different layers data;
An aspect according to the proposed method, described step 4 specifically comprises: the air speed data that reads the nearly stage
μ 0=
μ 1,
μ 2,
μ n , choose discrete wavelet analysis right
μ 0carry out
nlayer Wavelet fast decomposition and reconstruct, and with after reconstruct
nlayer low frequency signal and 1 layer of high-frequency signal are as the input data of the arma modeling in step 3.
can obtain a predicted value of wind speed;
represent one time predicted value;
,
...,
,
represent 1 layer, 2 layers ...,
nfloor height is signal estimation value frequently,
nlayer low frequency signal predicted value;
An aspect according to the proposed method, described step 5 specifically comprises: utilize correction model (
s,
p,
π) read the state space of current wind speed
π i through a step, shift the state space that arrives a predicted value
π j probability so that judgement a predicted value validity and revise predicted value one time.Wherein
sthe non-NULL state set that forms of all possible states of this area current season wind energy turbine set, the i.e. state space of system.
π=[
π 1 π 2 . π n ] represent current wind speed state;
p=[
p ij (
t,
t+ 1)]
n x nthe state transition probability matrix of system,
p ij (
t,
t+ 1)=
p{
x t+ 1
=
j|
x t =
i,
i,
j∈
sexpression system exists
tconstantly in state
i,through 1 step state transitions in state
jprobability;
Advantage of the present invention:
(1) by wavelet decomposition, the training sample value that the air speed value by the nearly stage is formed is decomposed into different levels, makes trend term, and periodic term and random entry decompose, and for the independent analysis and prediction of every one deck, can provide forecast model precision of prediction;
(2) use Markov one step transition probability correction forecasting wind speed value, effectively improve forecasting accuracy steady, non-stationary wind velocity signal, possess the ability of good tracking wind velocity signal stochastic variable.
Accompanying drawing explanation
Below in conjunction with drawings and Examples, the invention will be further described, in accompanying drawing:
Fig. 1 is the structural representation that the present invention is based on the method for forecasting of Markov.
Fig. 2 is the schematic flow sheet that the present invention is based on the method for forecasting of Markov.
Fig. 3 is the schematic flow sheet that the present invention is based on method for forecasting first embodiment of Markov.
Fig. 4 is northeast wind energy turbine set actual measurement wind speed time series schematic diagram in the specific embodiment of the invention.
Fig. 5 is the prediction effect schematic diagram of the wind speed in the specific embodiment of the invention.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with accompanying drawing and embodiment, the present invention is described in further detail, concrete embodiment described herein only, in order to explain the present invention, does not limit the present invention.
As shown in Figure 1, the block diagram of a kind of method for forecasting based on Markov theory of the present invention, should comprise read module, decomposition and reconstruction mould by the method for forecasting based on Markov theory
Piece, modeling and forecasting module, variable determination module, Markov correcting module, predicted value module and the value of predicting the outcome module.Wherein variable determination module, decomposition and reconstruction module are connected with modeling and forecasting module, and read module is connected with decomposed and reconstituted module, and Markov correcting module is connected with a predicted value module.The all illustrated annexations of the present invention are explain information reciprocal process only, is annexation in logic, should not only limit to physical connection.
In example work, read module from wind energy turbine set data acquisition and monitoring system with socket agreement according to the air speed data in the current season in historical time of the desired predicted time interval acquiring of prognoses system, this year when season air speed data, thereby obtain the wind speed one step transition probability matrix in current season and the training sample sequence that forecasting wind speed model needs.Read module is connected with the air speed data in nearly stage of multilayer decomposition and reconstruction with decomposed and reconstituted module simultaneously.Thus, arma modeling can be set up input and output variable, can carry out model training according to above input/output variable.Decomposed and reconstituted module is also connected with nearly stage wind speed module, and the nearly stage air speed data after decomposing is as to set up the input value of model.
Wavelet analysis and arma modeling can be realized the prediction of wind speed, but still possess larger predicated error on some time point, in order further to improve the precision of above forecasting wind speed model, the present invention proposes a kind of wind speed forecasting method based on Markov theory.The present invention also can utilize Markov correcting module to revise improvement for the primary air velocity predicted value of wavelet analysis and arma modeling.
Fig. 2 shows the flow process of the method for forecasting embodiment that the present invention is based on Markov theory, the system architecture of the method flow process based on shown in Fig. 1, and concrete steps process is as follows:
S1: according to a specific predicted time interval, read in same season air speed data in historical time, obtain a step transition probability matrix of this area wind speed state, this predicted time variable spaced.
In real work, can obtain Markov one step transition probability matrix according to following formula:
If same season air speed data information in historical time is
x 0=
x 1,
x 2,
x t-1
, wherein
x t be illustrated in
tair speed data constantly, time at intervals is a specific predicted time interval.
With certain fixing wind speed interval
wcarry out the division of air speed data state space, can arrange
nindividual wind speed state space air speed data scope be [
w* (
n-1),
w*
n] (
wvariable).
Build the wind speed one step transition probability matrix of wind energy turbine set in this season
p
p ij definition wind speed is by state space
π i through a step, be transferred to state space
π j probability,
,
Wherein,
be illustrated in all air speed datas, wind speed is by state
π i one step shifts and reaches
π j number of times;
nexpression state
π i the number of times occurring.
S2: according to a specific predicted time interval, read in the air speed data of current season, to obtain the training sample sequence of forecasting wind speed model.This predicted time variable spaced.
S3: utilize wavelet decomposition and reconstruct that training sample is for conversion into
nlayer low frequency air speed data and 1 floor height be air speed data frequently, utilizes above data to set up
n+ 1 arma modeling.
In real work, according to following formula, carry out respectively wavelet decomposition and reconstruct, arma modeling training;
If the air speed data in nearly stage
v 0=
v 1,
v 2,
v n be
nthe wind speed time series of individual time period, each air speed value is the mean wind speed of corresponding time period.
Choosing discrete wavelet analysis carries out the air speed data in nearly stage
nlayer Wavelet fast decomposition, to obtain
nlayer low frequency signal and 1 layer of high-frequency signal.
The formula that discrete wavelet decomposes is:
Wherein,
nthe number that represents list entries;
it is the low frequency component after decomposing;
it is the high fdrequency component after decomposing;
jrepresented
jlevel wavelet decomposition;
it is the scale coefficient of multiresolution analysis;
it is the wavelet coefficient of multiresolution analysis.
The n layer low frequency signal and the 1 layer of high-frequency signal that after decomposing, obtain are carried out respectively to wavelet reconstruction, can obtain and original signal
v 0the hierarchical signal of same scale.
The ARMA forecast model of layering low frequency signal and layering high-frequency signal builds following formula;
Wherein,
,
be respectively parameter undetermined;
be expressed as predicted value,
for known signal value;
for Gaussian sequence just too;
p,
qautoregressive and moving average exponent number for model.
The Autoregressive of model and moving average exponent number determined by AIC criterion, that is:
S4: the wind measuring device air speed data that reads the nearly stage, and utilize wavelet decomposition and reconstruct nearly the air speed data in stage be for conversion into N layer low frequency air speed data and 1 floor height air speed data frequently, utilize the arma modeling in S3 respectively multilayer low-frequency data and high-frequency data to be carried out to hierarchical prediction, after hierarchical prediction result weighted stacking, can obtain one time predicted value.
In real work, wavelet decomposition and the reconstruction formula in step 3 deferred in nearly stage air speed data wavelet decomposition and the reconstruct of in wind measuring device, reading, utilize formula in step 3 can obtain and the N layer low frequency air speed data of raw data same scale and 1 floor height air speed data frequently, and above multilayer air speed data is inputted respectively to each layer of arma modeling can obtain each layer of forecasting wind speed value, in each layer of predicted value weighted stacking, adopt following formula can obtain predicted value one time:
In formula:
represent one time predicted value;
,
...,
, represent 1 layer, 2 layers ...,
nfloor height is signal estimation value frequently,
nlayer low frequency signal predicted value.
S5: utilize Markov one step transition probability matrix to carry out a predicted value correction, obtain finally and predict the outcome.
Markov correction model can be expressed as (
s,
p,
π), wherein:
sthe non-NULL state set that forms of all possible states of this area current season wind energy turbine set, the i.e. state space of system.As the state space of wind energy turbine set
s=1,2,
x,
n, wherein 1 represents primary state, N represents final state.
xstate can represent wind speed range [
w* (
x-1),
w*
x],
wfor the wind speed size interval between state space, variable.
p=[
p ij (
t,
t+ 1)]
n x nthe state transition probability matrix of system,
p ij (
t,
t+ 1)=
p{
x t+ 1
=
j|
x t =
i,
i,
j∈
sexpression system exists
tconstantly in state
i,through 1 step state transitions in state
jprobability.
π=[π 1 π 2 . π n ] represent the current wind speed state of wind energy turbine set.
The correction model of Markov theory judges the state space of current wind speed
π i through a step state transitions, arrive the state space of a predicted value
π j probability carry out the validity judgement of a predicted value.In real work, Forecasting Methodology is with the wind speed of nearest time point
the state space being positioned
π i as original state, utilize matrix
pjudge the residing state space of forecasting wind speed value of its next time point
π j whether reasonable.If there is certain rationality, current state is described effectively and retains current data.If this probability is less than 0.05, illustrate that this transition probability is less, be small probability event, method is now sought distance automatically
π j nearest state space carries out the replacement correction of prediction of wind speed.
In addition, described forecast model value of predicting the outcome compares with the real-time Wind observation value gathering, and uses computing formula
, calculate the error between described real-time Wind observation value and forecasting wind speed value, wherein
for
treal-time Wind observation value constantly,
for
tforecasting wind speed value constantly, when continuously certain specific times
nrelatively in, described forecast model value of predicting the outcome surpasses 10% with the average relative error of the real-time Wind observation value gathering, and adjusts described forecasting wind speed model, wherein,
nfor natural number.
As shown in Figure 3, take certain large-scale wind power field as example, adopt the air speed data of this wind energy turbine set, carry out the predictions in 15 hours in advance that predicted time is spaced apart 1 hour, the validity of the method for forecasting of checking based on Markov theory, specific implementation process is as follows:
(1), the system prediction time interval is 1 hour, reads the air speed data in the current month in historical time according to predicted time space requirement, and then builds the wind farm wind velocity state one step transition probability matrix in current season;
(2), the system prediction time interval is 1 hour, reads the air speed data in current season according to predicted time space requirement, builds the training sample of forecast model;
(3), choosing the wavelet decomposition number of plies is 3 layers, adopt the western small echo of many shellfishes to utilize wavelet decomposition and reconstruct that training sample is for conversion into 3 layers of low frequency air speed data and 1 floor height frequency air speed data, for above 3 layers of low frequency air speed data and 1 floor height frequency air speed data, set up respectively arma modeling separately;
(4), read the air speed data in nearly stage, and utilize wavelet decomposition and reconstruct nearly the air speed data in stage be for conversion into 3 layers of low frequency air speed data and 1 floor height air speed data frequently, utilize the arma modeling in step 3 respectively multilayer low-frequency data and high-frequency data to be carried out to hierarchical prediction, after hierarchical prediction result weighted stacking, can obtain one time predicted value;
(5), utilize Markov one step transition probability matrix to carry out a predicted value correction, obtain finally and predict the outcome.
In order to test the accuracy of this method for forecasting based on Markov, adopt average error, square absolute error, root-mean-square error as evaluation criterion,
Average error is defined as:
Square absolute error is defined as:
Root-mean-square error is defined as:
In formula, represent the air speed value of actual measurement,
the forecasting wind speed value that represents the same time, errors is less, illustrates to predict the outcome better, and precision of prediction is higher.In this example, L=15, resulting statistics is as table 1.
The comparison of table 1 method test performance
| Forecasting Methodology | Method one | Method two | Method three |
| Average error | 0.1745 | -0.1313 | -0.0671 |
| Square absolute error | 2.9246 | 0.6704 | 0.4690 |
| Root-mean-square error | 3.6142 | 0.9195 | 0.5819 |
High-frequency signal in table after algorithm one use wavelet analysis and the prediction algorithm of arma modeling; Multilayer signal after algorithm two expression use wavelet analysises and the prediction algorithm of arma modeling; Algorithm three is the improvement forecasting wind speed algorithm based on Markov theory.It is as shown in the table, the short-term wind speed forecasting method based on Markov theory proposed by the invention, and precision of prediction is greatly improved, and has illustrated that this method for forecasting based on Markov theory has higher accuracy and reliability.
The foregoing is only preferred embodiments of the present invention, be not limited to the present invention, all any modifications of making within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.
Claims (2)
1. the method for forecasting based on Markov theory, is characterized in that comprising the following steps:
According to a specific predicted time interval, read in same season air speed data in historical time, obtain a step transition probability matrix of this area wind speed state, this predicted time variable spaced;
According to a specific predicted time interval, read in the air speed data of current season, to obtain the training sample sequence of forecasting wind speed model; This predicted time variable spaced;
Utilize wavelet decomposition and reconstruct that training sample is for conversion into
nlayer low frequency air speed data and 1 floor height be wind speed training sample data frequently, more than utilization
n+ 1 layer of training sample data are set up
n+ 1 arma modeling;
Read the air speed data in nearly stage, and utilize wavelet decomposition and reconstruct nearly the air speed data in stage be for conversion into
nlayer low frequency air speed data and 1 floor height be air speed data frequently, utilizes the arma modeling of having set up respectively multilayer low-frequency data and high-frequency data to be carried out to hierarchical prediction, can obtain one time predicted value after hierarchical prediction result weighted stacking;
Utilize Markov one step transition probability matrix to carry out a predicted value correction, obtain finally and predict the outcome.
2. the method for forecasting of Markov theory according to claim 1, characterized by further comprising:
Using the historical time same season air speed data as a step transition probability matrix of wind speed state of current season of wind energy turbine set, establish same season air speed data in historical time
x 0=
x 1,
x 2,
x t , wherein
x t be illustrated in
tair speed data constantly;
With certain fixing wind speed interval
wcarry out the division of air speed data state space, can arrange
nindividual wind speed state space air speed data scope be [
w* (
n-1),
w*
n] (
wvariable);
And then can build the wind speed state one step transition probability matrix of wind energy turbine set in this season
p ;
With wind speed state, one step transition probability matrix P carries out the correction of a predicted value, Markov model can be expressed as (
s,
p,
π);
sthe non-NULL state set that all possible states of this area wind energy turbine set form, the i.e. state space of system;
p=[
p ij (
t,
t+ 1)]
n x nthe state transition probability matrix of system,
p ij (
t,
t+ 1)=
p{
x t+ 1
=
j|
x t =
iexpression system exists
tconstantly in state
i,through 1 step state transitions in state
jprobability;
π=[
π 1 π 2 . π n ] represent the current wind speed state of wind energy turbine set; The correction model of Markov theory passes through
The state space of current wind speed
π i through a step state transitions, arrive the state space of a predicted value
π j probable value size carry out the validity judgement of a predicted value, and according to this probable value, carry out judgement and the correction of a predicted value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410062993.5A CN103996071A (en) | 2014-02-25 | 2014-02-25 | Wind power plant wind speed prediction method based on Markov theory |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410062993.5A CN103996071A (en) | 2014-02-25 | 2014-02-25 | Wind power plant wind speed prediction method based on Markov theory |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103996071A true CN103996071A (en) | 2014-08-20 |
Family
ID=51310231
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410062993.5A Pending CN103996071A (en) | 2014-02-25 | 2014-02-25 | Wind power plant wind speed prediction method based on Markov theory |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103996071A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182914A (en) * | 2014-09-05 | 2014-12-03 | 国家电网公司 | Wind power output time series modeling method based on fluctuation characteristics |
| CN105825040A (en) * | 2015-12-29 | 2016-08-03 | 海南电力技术研究院 | Short-term power load prediction method |
| CN107038279A (en) * | 2017-03-08 | 2017-08-11 | 北京航空航天大学 | The Forecasting Methodology and device of a kind of turbulence signal |
| CN107313898A (en) * | 2017-08-15 | 2017-11-03 | 华北电力大学 | The method of wind generator system control based on Markov saltus step rules |
| WO2018218832A1 (en) * | 2017-06-02 | 2018-12-06 | 北京金风科创风电设备有限公司 | Method and apparatus for correcting predicted wind speed of wind farm |
| CN110659827A (en) * | 2019-09-23 | 2020-01-07 | 珠海格力电器股份有限公司 | Energy scheduling method, child node system, scheduling system, and storage medium |
| CN112381420A (en) * | 2020-11-17 | 2021-02-19 | 平安普惠企业管理有限公司 | Risk early warning method, device and equipment based on desert wind speed and storage medium |
| CN113408591A (en) * | 2021-06-01 | 2021-09-17 | 上海自动化仪表有限公司 | Process trend analysis and prediction method based on intelligent instrument |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102129511A (en) * | 2011-02-21 | 2011-07-20 | 北京航空航天大学 | System for forecasting short-term wind speed of wind power station based on MATLAB |
| CN102338808A (en) * | 2011-08-26 | 2012-02-01 | 天津理工大学 | Online hybrid forecasting method for short-term wind speed of wind power field |
| CN102682207A (en) * | 2012-04-28 | 2012-09-19 | 中国科学院电工研究所 | Ultrashort combined predicting method for wind speed of wind power plant |
-
2014
- 2014-02-25 CN CN201410062993.5A patent/CN103996071A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102129511A (en) * | 2011-02-21 | 2011-07-20 | 北京航空航天大学 | System for forecasting short-term wind speed of wind power station based on MATLAB |
| CN102338808A (en) * | 2011-08-26 | 2012-02-01 | 天津理工大学 | Online hybrid forecasting method for short-term wind speed of wind power field |
| CN102682207A (en) * | 2012-04-28 | 2012-09-19 | 中国科学院电工研究所 | Ultrashort combined predicting method for wind speed of wind power plant |
Non-Patent Citations (5)
| Title |
|---|
| 李东福 等: "基于多尺度小波分解和时间序列法的风电场风速预测", 《华北电力大学学报》 * |
| 李玲玲 等: "基于小波分析和时间序列的风速预测", 《陕西电力》 * |
| 焦娇 等: "基于ARIMA和马尔可夫链的风速中期预测模型", 《2011年电磁测量技术及仪器学术年会论文集》 * |
| 贺军 等: "灰色一马尔科夫模型在风电场风速预测中的应用", 《发电与空调 》 * |
| 马海云: "《面向过程的软件设计及优化技术》", 30 June 2013, 国防工业出版社 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182914A (en) * | 2014-09-05 | 2014-12-03 | 国家电网公司 | Wind power output time series modeling method based on fluctuation characteristics |
| CN104182914B (en) * | 2014-09-05 | 2017-06-23 | 国家电网公司 | A kind of wind power output time series modeling method based on wave characteristic |
| CN105825040A (en) * | 2015-12-29 | 2016-08-03 | 海南电力技术研究院 | Short-term power load prediction method |
| CN107038279A (en) * | 2017-03-08 | 2017-08-11 | 北京航空航天大学 | The Forecasting Methodology and device of a kind of turbulence signal |
| WO2018218832A1 (en) * | 2017-06-02 | 2018-12-06 | 北京金风科创风电设备有限公司 | Method and apparatus for correcting predicted wind speed of wind farm |
| US11208985B2 (en) | 2017-06-02 | 2021-12-28 | Beijing Goldwind Science & Creation Windpower Equipment Co., Ltd. | Correction method and apparatus for predicted wind speed of wind farm |
| CN107313898A (en) * | 2017-08-15 | 2017-11-03 | 华北电力大学 | The method of wind generator system control based on Markov saltus step rules |
| CN110659827A (en) * | 2019-09-23 | 2020-01-07 | 珠海格力电器股份有限公司 | Energy scheduling method, child node system, scheduling system, and storage medium |
| CN110659827B (en) * | 2019-09-23 | 2020-12-15 | 珠海格力电器股份有限公司 | Energy scheduling method, child node system, scheduling system, and storage medium |
| CN112381420A (en) * | 2020-11-17 | 2021-02-19 | 平安普惠企业管理有限公司 | Risk early warning method, device and equipment based on desert wind speed and storage medium |
| CN113408591A (en) * | 2021-06-01 | 2021-09-17 | 上海自动化仪表有限公司 | Process trend analysis and prediction method based on intelligent instrument |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103996071A (en) | Wind power plant wind speed prediction method based on Markov theory | |
| Dong et al. | Wind power day-ahead prediction with cluster analysis of NWP | |
| CN111626506B (en) | Regional photovoltaic power probability prediction method based on federal learning and cooperative regulation and control system thereof | |
| Wang et al. | Echo state network based ensemble approach for wind power forecasting | |
| US10443577B2 (en) | Systems and methods for improved wind power generation | |
| Tascikaraoglu et al. | A review of combined approaches for prediction of short-term wind speed and power | |
| CN102663513B (en) | Utilize the wind power combined prediction modeling method of grey relational grade analysis | |
| Bhaskar et al. | AWNN-assisted wind power forecasting using feed-forward neural network | |
| Foley et al. | Current methods and advances in forecasting of wind power generation | |
| Hu et al. | A hybrid approach based on the Gaussian process with t-observation model for short-term wind speed forecasts | |
| Zhang et al. | GEFCom2014 probabilistic solar power forecasting based on k-nearest neighbor and kernel density estimator | |
| CN102478584B (en) | Wind power station wind speed prediction method based on wavelet analysis and system thereof | |
| CN105868853B (en) | Method for predicting short-term wind power combination probability | |
| CN102236795B (en) | Method for forecasting wind speed in wind power station | |
| CN103268366A (en) | Combined wind power prediction method suitable for distributed wind power plant | |
| CN102102626A (en) | Method for forecasting short-term power in wind power station | |
| CN111488896B (en) | Distribution line time-varying fault probability calculation method based on multi-source data mining | |
| CN102184453A (en) | Wind power combination predicting method based on fuzzy neural network and support vector machine | |
| CN110766200A (en) | Method for predicting generating power of wind turbine generator based on K-means mean clustering | |
| CN102426674A (en) | Power system load prediction method based on Markov chain | |
| CN104077632A (en) | Wind power field power prediction method based on deep neural network | |
| Hu et al. | Daily condition monitoring of grid-connected wind turbine via high-fidelity power curve and its comprehensive rating | |
| CN105243502A (en) | Hydropower station scheduling risk assessment method and system based on runoff interval prediction | |
| Jiang et al. | Parallel processing of probabilistic models-based power supply unit mid-term load forecasting with apache spark | |
| CN102509173A (en) | Markov chain based method for accurately forecasting power system loads |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140820 |