CN112948462A - Ultra-short-term wind speed prediction method based on improved singular spectrum analysis and Bp neural network - Google Patents

Ultra-short-term wind speed prediction method based on improved singular spectrum analysis and Bp neural network Download PDF

Info

Publication number
CN112948462A
CN112948462A CN202110220371.0A CN202110220371A CN112948462A CN 112948462 A CN112948462 A CN 112948462A CN 202110220371 A CN202110220371 A CN 202110220371A CN 112948462 A CN112948462 A CN 112948462A
Authority
CN
China
Prior art keywords
wind speed
singular
neural network
prediction
sequence
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
Application number
CN202110220371.0A
Other languages
Chinese (zh)
Inventor
邓长虹
杨秋玲
王学斌
甘嘉田
卢国强
傅国斌
丁玉杰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wuhan University WHU
Electric Power Research Institute of State Grid Qinghai Electric Power Co Ltd
Original Assignee
Wuhan University WHU
Electric Power Research Institute of State Grid Qinghai Electric Power Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Wuhan University WHU, Electric Power Research Institute of State Grid Qinghai Electric Power Co Ltd filed Critical Wuhan University WHU
Priority to CN202110220371.0A priority Critical patent/CN112948462A/en
Publication of CN112948462A publication Critical patent/CN112948462A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01W—METEOROLOGY
    • G01W1/00—Meteorology
    • G01W1/10—Devices for predicting weather conditions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2474—Sequence data queries, e.g. querying versioned data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • 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

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Software Systems (AREA)
  • Mathematical Physics (AREA)
  • Databases & Information Systems (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • General Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Environmental & Geological Engineering (AREA)
  • Probability & Statistics with Applications (AREA)
  • Marketing (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Fuzzy Systems (AREA)
  • Public Health (AREA)
  • Primary Health Care (AREA)
  • Water Supply & Treatment (AREA)
  • Atmospheric Sciences (AREA)
  • Biodiversity & Conservation Biology (AREA)
  • Ecology (AREA)
  • Environmental Sciences (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)

Abstract

The invention discloses an ultra-short-term wind speed prediction method based on improved singular spectrum analysis and a Bp neural network, which comprises the following steps of: s1, acquiring historical wind speed data; s2, decomposing the wind speed sequence into a series of subsequences by adopting improved singular spectrum analysis, wherein the improved singular spectrum analysis introduces singular entropy to judge and remove noise components of the wind speed sequence; s3, dividing the training data of each subsequence, respectively establishing a Bp neural network model and completing training; s4, respectively predicting the subsequences obtained by the analysis and decomposition of the improved singular spectrum; and S5, combining the prediction results of all the subsequences to complete wind speed prediction. The invention removes the noise influence of the original wind speed sequence through the improved singular spectrum analysis, and further improves the prediction precision of the wind speed.

Description

Ultra-short-term wind speed prediction method based on improved singular spectrum analysis and Bp neural network
Technical Field
The invention relates to the field of ultra-short-term wind speed prediction, in particular to an ultra-short-term wind speed prediction method based on improved singular spectrum analysis and a Bp neural network.
Background art:
wind power is a new energy power generation technology with zero emission and low operation cost, and is developed at a high speed in the global power production industry. However, the characteristics of fluctuation, indirection, low energy density and the like of the wind speed can improve the difficulty of wind power integration and reduce the reliability of the operation of the power system. Therefore, accurate and reliable wind speed prediction has important significance for wind power integration and safety and stability analysis of a power system.
The current research on wind speed prediction mainly comprises a physical method, a statistical method or a combination of the physical method and the statistical method. However, the physical method needs high-precision weather forecast data and is more suitable for medium-long term prediction of wind power plant construction early-stage evaluation, and a single statistical method cannot completely capture all characteristics of wind speed data, so that the prediction precision is difficult to guarantee. Therefore, in recent years, researchers have been working on a prediction method in which a plurality of single models are combined to form a multi-model to obtain better prediction performance. The machine learning model is widely applied to ultra-short-term wind speed prediction due to strong fitting capability, but due to the limitation of input data, the characteristics of an original wind speed sequence cannot be mined, so that the prediction accuracy is not high.
Disclosure of Invention
Aiming at the problem that wind speed data acquired by a wind power plant is influenced by more factors and noise information is easily introduced, an ultra-short-term wind speed prediction method based on improved singular spectrum analysis and a Bp neural network is provided, and the precision of the Bp neural network is effectively improved.
The technical scheme adopted by the invention is as follows:
an ultra-short-term wind speed prediction method based on improved singular spectrum analysis and a Bp neural network is characterized by comprising the following steps:
step 1: the method comprises the steps of collecting historical wind speed data, specifically, acquiring N data in total at the height of a hub (m/s) within D days in a certain place and at set time intervals.
Step 2: decomposing an original wind speed sequence by using improved singular spectrum analysis;
and step 3: dividing a training set for the subsequence obtained by decomposition, and establishing a Bp neural network model to finish training;
and 4, step 4: respectively using the trained models to carry out multi-step prediction on the subsequences, specifically: taking the latest n historical wind speed values at the predicted initial moment as input, and predicting the wind speed value in the first step according to the trained neural network; taking the wind speed predicted value of the first step as a known quantity, and simultaneously pushing the input forward one to be used as a new input of a neural network to predict the wind speed value of the second step; and the wind speed sequence can be predicted in multiple steps by sequentially predicting.
And 5: combining the multi-step prediction results of all the subsequences to complete multi-step prediction of the wind speed and evaluate the prediction results, wherein the evaluation indexes comprise:
root Mean Square Error (RMSE):
Figure BDA0002954558950000021
mean Absolute Error (MAE):
Figure BDA0002954558950000022
mean Absolute Percentage Error (MAPE):
Figure BDA0002954558950000023
in the above method for predicting an ultra-short-term wind speed based on an improved singular spectrum analysis and a Bp neural network, the improved singular spectrum analysis in step 2 specifically includes:
step 2.1, constructing a track matrix, and defining an original wind speed sequence x ═ x1,x2,…,xN]TWherein N is the wind speed sequence length, and the window length is defined as L (L is more than 1 and less than N/2); the trajectory matrix X is represented as follows;
Figure BDA0002954558950000024
step 2.2, singular value decomposition, to the matrix XXTPerforming singular value decomposition to obtain L eigenvalues lambda arranged from large to smalli(i ═ 1,2, …, L), i.e. λ1≥λ2≥…≥λLWhile the characteristic value lambda can be obtainediCorresponding left and right eigenvectors uiAnd vi(ii) a Definition of
Figure BDA0002954558950000025
The trajectory matrix X can be expressed as,
Figure BDA0002954558950000026
step 2.3, denoising, namely judging the noise of the sequence by using singular entropy; the singular entropy is to calculate the information quantity by substituting singular values into the information entropy, the specific formula of the singular entropy increment is,
Figure BDA0002954558950000031
where k denotes the order of the singular entropy, Δ EkAn increment representing the singular entropy at k-order;
based on the singular entropy theory, the extraction sequence of the noise depends on whether the increment of the singular entropy is reduced to a stable value; if the increment of the singular entropy becomes stable in the Mth order, the singular values from 1 order to the M order mainly contain effective information, and the subsequent singular values are related to noise; the invention is based on the above and all the
Figure BDA0002954558950000032
Dividing into M +1 groups, the first M group containing main information, and the M +1 group containing all noise and removing, i.e. less than lambdaMCorresponding to the characteristic value of
Figure BDA0002954558950000033
Directly removing; that is to say that the first and second electrodes,
Figure BDA0002954558950000034
wherein M is 1,2, …, M (M is less than or equal to L);
step 2.4, reconstructing the signal by using a diagonal averaging method
Figure BDA0002954558950000035
Converting the form of the SSA component, ym＝[y1,m,y2,m,…,yN,m]TThe obtained one-dimensional SSA component is the mth subsequence obtained after the wind speed SSA decomposition and denoising; the original wind speed sequence x after the noise is removed can be reconstructed by superposing all the subsequences; superposing all the sub-sequence prediction results to obtain a prediction result of the wind speed sequence;
Figure BDA0002954558950000036
wherein the content of the first and second substances,
Figure BDA0002954558950000037
respectively represent ymAnd the predicted value of x.
In the above-mentioned method for predicting wind speed in an ultra-short period based on the improved singular spectral analysis and Bp neural network, the Bp neural network of step 3 includes 3 portions of an input layer, a hidden layer and an output layer.
Compared with the prior art, the invention has the beneficial effects that: the improved singular spectrum analysis of the invention eliminates the correlation among the subsequences in the singular value decomposition process, and realizes the conversion from nonlinear complex signals to subsequences with different trend characteristics; noise components of the wind speed sequence are eliminated based on the singular entropy theory in the grouping process, the fluctuation caused by noise in the original wind speed is weakened, and the method has important significance for improving the performance of a prediction model and improving the prediction precision of the ultra-short-term wind speed.
Drawings
FIG. 1 is a flow chart of an ultra-short term wind speed prediction method based on improved singular spectrum analysis and Bp neural network
FIG. 2 is a schematic structural diagram of a Bp neural network in an ultra-short-term wind speed prediction method based on improved singular spectrum analysis and the Bp neural network
FIG. 3 is a plot of hub altitude wind velocity (m/s) data over a 20 day period at 15 minute intervals.
FIG. 4 is a training data partitioning for a neural network.
FIG. 5 is a multi-step prediction scheme.
Detailed Description
The present invention is further described below, and the following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
The invention relates to an ultra-short-term wind speed prediction method based on improved singular spectrum analysis and a Bp neural network, which comprises the following specific steps of:
step 1: collecting historical data of wind speed;
this step is to obtain the hub height wind speed (m/s) in 20 days at 15 min intervals in a certain place, and totally 1920 data are shown in fig. 3.
Step 2: decomposing an original wind speed sequence by using improved singular spectrum analysis;
this step is carried outStarting from an original wind speed sequence, the wind speed sequence is decomposed according to a wind speed change track, and the conversion from a nonlinear complex signal to a subsequence with different trend characteristics is realized. For the original wind speed sequence x ═ x1,x2,…,xN]TAnd N is 1910, and is the wind speed sequence length. The original wind speed sequence is decomposed into M-dimensional vectors by decomposition with improved singular spectrum analysis. The method comprises the following specific steps:
1) constructing a trajectory matrix
Suppose the original wind speed sequence x ═ x1,x2,…,xN]TWherein N is the wind speed sequence length, and the window length is defined as L (L is more than 1 and less than N/2). The trajectory matrix X is represented as follows.
Figure BDA0002954558950000041
2) Singular value decomposition
For matrix XXTPerforming singular value decomposition to obtain L eigenvalues lambda arranged from large to smalli(i ═ 1,2, …, L), i.e. λ1≥λ2≥…≥λLWhile the characteristic value lambda can be obtainediCorresponding left and right eigenvectors uiAnd vi. Definition of
Figure BDA0002954558950000051
The trajectory matrix X can be expressed as,
Figure BDA0002954558950000052
3) de-noising
In order to accurately extract and remove noise information in the original wind speed sequence, the singular entropy is used for judging the noise of the sequence. The singular entropy is to calculate the information quantity by substituting singular values into the information entropy, the specific formula of the singular entropy increment is,
Figure BDA0002954558950000053
where k denotes the order of the singular entropy, Δ EkRepresenting the increment of the singular entropy at order k.
Based on the singular entropy theory, the order of noise extraction depends on whether the increment of the singular entropy decreases to a stable value. If the increment of the singular entropy becomes stable in the Mth order, it is shown that the singular values from 1 order to M order mainly contain valid information, while the subsequent singular values are related to noise. The invention is based on the above and all the
Figure BDA0002954558950000054
Dividing into M +1 groups, the first M group containing main information, and the M +1 group containing all noise and removing, i.e. less than lambdaMCorresponding to the characteristic value of
Figure BDA0002954558950000055
And directly removing. That is to say that the first and second electrodes,
Figure BDA0002954558950000056
wherein M is 1,2, …, and M (M is less than or equal to L).
4) Reconstructing a signal
By using a diagonal averaging method
Figure BDA0002954558950000057
Converting the form of the SSA component, ym＝[y1,m,y2,m,…,yN,m]TThe obtained one-dimensional SSA component, namely the mth subsequence obtained after the wind speed SSA decomposition and denoising. And reconstructing the original wind speed sequence x after the noise is removed by superposing all the subsequences. And superposing all the sub-sequence prediction results to obtain the prediction result of the wind speed sequence.
Figure BDA0002954558950000058
Wherein the content of the first and second substances,
Figure BDA0002954558950000059
respectively represent ymAnd the predicted value of x.
And step 3: dividing a training set for the subsequence obtained by decomposition, and establishing a Bp neural network model to finish training;
this step divides the training data for the first 1910 data of each sub-sequence as shown in fig. 4. The first 70% of the data is used as a training set to train the weights and thresholds of the neural network, and the last 30% is used to verify the validity of the model.
And 4, step 4: respectively using the trained models to carry out multi-step prediction on the subsequences;
this step uses the trained neural network model to perform multi-step prediction on the subsequences, as shown in fig. 5. Where y is the historical value in the sub-sequence,
Figure BDA0002954558950000062
indicating the predicted value. The multi-step prediction is that firstly, the latest 20 historical values are predicted as input, and the predicted value which is one step ahead is obtained. And then, taking the predicted value which leads one step as a historical value, and performing prediction which leads two steps. By analogy, the prediction of the step m ahead can be completed. In this document, m is 10.
And 5: and combining the multi-step prediction results of all the subsequences to complete multi-step prediction of the wind speed and evaluate the prediction results.
The step is to combine the multi-step prediction results of all the subsequences to complete the multi-step prediction of the wind speed, and evaluate the prediction results in 10 steps, wherein the evaluation indexes are shown in the following table 1.
TABLE 1 evaluation index of prediction result
Figure BDA0002954558950000061
The invention improves singular spectrum analysis based on singular entropy, eliminates noise components of the wind speed sequence based on the singular entropy theory in the grouping process, weakens the volatility caused by the noise in the original wind speed, simultaneously realizes the conversion of nonlinear complex signals to subsequences with different trend characteristics, and has important significance for improving the performance of a prediction model and improving the prediction precision of ultra-short-term wind speed.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.

Claims (3)

1. An ultra-short-term wind speed prediction method based on improved singular spectrum analysis and a Bp neural network is characterized by comprising the following steps:
step 1: acquiring historical data of wind speed, specifically acquiring the height wind speed (m/s) of a hub at set time intervals in a certain place within D days, wherein the total number of the data is N;
step 2: decomposing an original wind speed sequence by using improved singular spectrum analysis;
and step 3: dividing a training set for the subsequence obtained by decomposition, and establishing a Bp neural network model to finish training;
and 4, step 4: respectively using the trained models to carry out multi-step prediction on the subsequences, specifically: taking the latest n historical wind speed values at the predicted initial moment as input, and predicting the wind speed value in the first step according to the trained neural network; taking the wind speed predicted value of the first step as a known quantity, and simultaneously pushing the input forward one to be used as a new input of a neural network to predict the wind speed value of the second step; the wind speed sequence can be predicted in multiple steps by sequentially predicting;
and 5: combining the multi-step prediction results of all the subsequences to complete multi-step prediction of the wind speed and evaluate the prediction results, wherein the evaluation indexes comprise:
root Mean Square Error (RMSE):
Figure FDA0002954558940000011
mean Absolute Error (MAE):
Figure FDA0002954558940000012
mean Absolute Percentage Error (MAPE):
Figure FDA0002954558940000013
2. the ultra-short-term wind speed prediction method based on the improved singular spectrum analysis and the Bp neural network as claimed in claim 1, wherein the improved singular spectrum analysis in the step 2 specifically comprises:
step 2.1, constructing a track matrix, and defining an original wind speed sequence x ═ x1,x2,…,xN]TWherein N is the wind speed sequence length, and the window length is defined as L (L is more than 1 and less than N/2); the trajectory matrix X is represented as follows;
Figure FDA0002954558940000014
step 2.2, singular value decomposition, to the matrix XXTPerforming singular value decomposition to obtain L eigenvalues lambda arranged from large to smalli(i ═ 1,2, …, L), i.e. λ1≥λ2≥…≥λLWhile the characteristic value lambda can be obtainediCorresponding left and right eigenvectors uiAnd vi(ii) a Definition of
Figure FDA0002954558940000021
The trajectory matrix X can be expressed as,
Figure FDA0002954558940000022
step 2.3, denoising, namely judging the noise of the sequence by using singular entropy; the singular entropy is to calculate the information quantity by substituting singular values into the information entropy, the specific formula of the singular entropy increment is,
Figure FDA0002954558940000023
where k denotes the order of the singular entropy, Δ EkAn increment representing the singular entropy at k-order;
based on the singular entropy theory, the extraction sequence of the noise depends on whether the increment of the singular entropy is reduced to a stable value; if the increment of the singular entropy becomes stable in the Mth order, the singular values from 1 order to the M order mainly contain effective information, and the subsequent singular values are related to noise; the invention is based on the above and all the
Figure FDA0002954558940000024
Dividing into M +1 groups, the first M group containing main information, and the M +1 group containing all noise and removing, i.e. less than lambdaMCorresponding to the characteristic value of
Figure FDA0002954558940000025
Directly removing; that is to say that the first and second electrodes,
Figure FDA0002954558940000026
wherein M is 1,2, …, M (M is less than or equal to L);
step 2.4, reconstructing the signal by using a diagonal averaging method
Figure FDA0002954558940000027
Converting the form of the SSA component, ym＝[y1,m,y2,m,…,yN,m]TThe obtained one-dimensional SSA component is the mth subsequence obtained after the wind speed SSA decomposition and denoising; the original wind speed sequence x after the noise is removed can be reconstructed by superposing all the subsequences; superposing all the sub-sequence prediction results to obtain a prediction result of the wind speed sequence;
Figure FDA0002954558940000028
wherein the content of the first and second substances,
Figure FDA0002954558940000029
respectively represent ymAnd the predicted value of x.
3. The ultra-short term wind speed prediction method based on the improved singular spectral analysis and Bp neural network as claimed in claim 1, wherein the Bp neural network of step 3 comprises 3 parts of input layer, hidden layer and output layer.
CN202110220371.0A 2021-02-26 2021-02-26 Ultra-short-term wind speed prediction method based on improved singular spectrum analysis and Bp neural network Pending CN112948462A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110220371.0A CN112948462A (en) 2021-02-26 2021-02-26 Ultra-short-term wind speed prediction method based on improved singular spectrum analysis and Bp neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110220371.0A CN112948462A (en) 2021-02-26 2021-02-26 Ultra-short-term wind speed prediction method based on improved singular spectrum analysis and Bp neural network

Publications (1)

Publication Number Publication Date
CN112948462A true CN112948462A (en) 2021-06-11

Family

ID=76246639

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110220371.0A Pending CN112948462A (en) 2021-02-26 2021-02-26 Ultra-short-term wind speed prediction method based on improved singular spectrum analysis and Bp neural network

Country Status (1)

Country Link
CN (1) CN112948462A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115345367A (en) * 2022-08-16 2022-11-15 哈尔滨工业大学 Large-span bridge wind speed prediction method based on real-time denoising
CN115796231A (en) * 2023-01-28 2023-03-14 湖南赛能环测科技有限公司 Ultrashort-term wind speed prediction method based on temporal analysis

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104636823A (en) * 2015-01-23 2015-05-20 中国农业大学 Wind power prediction method
CN105224872A (en) * 2015-09-30 2016-01-06 河南科技大学 A kind of user's anomaly detection method based on neural network clustering
CN106650976A (en) * 2015-10-29 2017-05-10 深圳市综合交通运行指挥中心 Travel analysis and forecasting method and system, and travel analysis and forecasting method and system based on IC card
CN106779151A (en) * 2016-11-14 2017-05-31 中南大学 A kind of line of high-speed railway wind speed multi-point multi-layer coupling prediction method
CN110083940A (en) * 2019-04-28 2019-08-02 东华大学 A kind of short-term wind speed forecasting method based on SSA-HMD-CNNSVM model
CN110348632A (en) * 2019-07-11 2019-10-18 广东电网有限责任公司 A kind of wind power forecasting method based on singular spectrum analysis and deep learning
AU2020101854A4 (en) * 2020-08-17 2020-09-24 China Communications Construction Co., Ltd. A method for predicting concrete durability based on data mining and artificial intelligence algorithm

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104636823A (en) * 2015-01-23 2015-05-20 中国农业大学 Wind power prediction method
CN105224872A (en) * 2015-09-30 2016-01-06 河南科技大学 A kind of user's anomaly detection method based on neural network clustering
CN106650976A (en) * 2015-10-29 2017-05-10 深圳市综合交通运行指挥中心 Travel analysis and forecasting method and system, and travel analysis and forecasting method and system based on IC card
CN106779151A (en) * 2016-11-14 2017-05-31 中南大学 A kind of line of high-speed railway wind speed multi-point multi-layer coupling prediction method
CN110083940A (en) * 2019-04-28 2019-08-02 东华大学 A kind of short-term wind speed forecasting method based on SSA-HMD-CNNSVM model
CN110348632A (en) * 2019-07-11 2019-10-18 广东电网有限责任公司 A kind of wind power forecasting method based on singular spectrum analysis and deep learning
AU2020101854A4 (en) * 2020-08-17 2020-09-24 China Communications Construction Co., Ltd. A method for predicting concrete durability based on data mining and artificial intelligence algorithm

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
张文;胡从川;阙波;滕明尧;钱海;杨昊;: "一种实时校正的改进BP神经网络超短期风速预测模型", 《电网与清洁能源》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115345367A (en) * 2022-08-16 2022-11-15 哈尔滨工业大学 Large-span bridge wind speed prediction method based on real-time denoising
CN115796231A (en) * 2023-01-28 2023-03-14 湖南赛能环测科技有限公司 Ultrashort-term wind speed prediction method based on temporal analysis
CN115796231B (en) * 2023-01-28 2023-12-08 湖南赛能环测科技有限公司 Temporal analysis ultra-short term wind speed prediction method

Similar Documents

Publication Publication Date Title
CN112149316B (en) Aero-engine residual life prediction method based on improved CNN model
CN111582551B (en) Wind power plant short-term wind speed prediction method and system and electronic equipment
CN112948462A (en) Ultra-short-term wind speed prediction method based on improved singular spectrum analysis and Bp neural network
CN110059845B (en) Metering device clock error trend prediction method based on time sequence evolution gene model
CN114239718B (en) High-precision long-term time sequence prediction method based on multi-element time sequence data analysis
CN111242377A (en) Short-term wind speed prediction method integrating deep learning and data denoising
CN112634608B (en) Regional traffic flow prediction method and system
CN110598955A (en) Maximum instantaneous wind speed probability prediction method for high-speed train
CN113537649A (en) Power grid load prediction method and system based on neural network and dynamic mode decomposition
CN113344288A (en) Method and device for predicting water level of cascade hydropower station group and computer readable storage medium
CN112712215A (en) Wind power plant power prediction method
CN116451035A (en) Data characteristic engineering processing method for improving distributed photovoltaic prediction precision
CN110766215B (en) Wind power climbing event prediction method based on feature adaptive selection and WDNN
CN114037002A (en) Cable-stayed bridge girder abnormal deflection monitoring method based on deep learning model
CN106777466B (en) Dynamic evolution modeling method of high-sulfur natural gas purification process based on ST-UPFNN algorithm
CN116665483A (en) Novel method for predicting residual parking space
CN113609766B (en) Soft measurement method based on depth probability hidden model
CN113627685B (en) Wind driven generator power prediction method considering wind power internet load limit
CN116388184B (en) Ultra-short-term wind speed revising method and system based on wind speed daily fluctuation characteristics
CN111275090B (en) GNSS ultra-fast clock error forecasting method
CN116739168A (en) Runoff prediction method based on gray theory and codec
CN107563573B (en) Solar power generation capacity prediction method based on self-adaptive learning hybrid model
CN113537573B (en) Wind power operation trend prediction method based on double space-time feature extraction
CN116388148A (en) Wind power prediction method, device, equipment and storage medium
CN115296298A (en) Wind power plant power prediction method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20210611