CN112329987A - Short-term wind power plant power prediction method based on Adaboost-EMD-SVM - Google Patents

Short-term wind power plant power prediction method based on Adaboost-EMD-SVM Download PDF

Info

Publication number
CN112329987A
CN112329987A CN202011100375.7A CN202011100375A CN112329987A CN 112329987 A CN112329987 A CN 112329987A CN 202011100375 A CN202011100375 A CN 202011100375A CN 112329987 A CN112329987 A CN 112329987A
Authority
CN
China
Prior art keywords
svm
imf
wind
emd
adaboost
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.)
Withdrawn
Application number
CN202011100375.7A
Other languages
Chinese (zh)
Inventor
张旭东
苏志伟
曹竣
刘悦
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.)
Nanjing University of Science and Technology
Original Assignee
Nanjing University of Science and Technology
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 Nanjing University of Science and Technology filed Critical Nanjing University of Science and Technology
Priority to CN202011100375.7A priority Critical patent/CN112329987A/en
Publication of CN112329987A publication Critical patent/CN112329987A/en
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Business, Economics & Management (AREA)
  • Mathematical Physics (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Operations Research (AREA)
  • Pure & Applied Mathematics (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Computational Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Medical Informatics (AREA)
  • Databases & Information Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Algebra (AREA)
  • Evolutionary Biology (AREA)
  • Geometry (AREA)
  • Computer Hardware Design (AREA)
  • Development Economics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Game Theory and Decision Science (AREA)
  • Probability & Statistics with Applications (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Computing Systems (AREA)
  • Wind Motors (AREA)
  • Control Of Eletrric Generators (AREA)

Abstract

The invention discloses a short-term wind power plant power prediction method based on Adaboost-EMD-SVM, which comprises the following steps: acquisition of a sample set: taking historical meteorological data or numerical weather forecast data of the wind power plant as a sample set; processing of the sample set: performing EMD decomposition on the wind speed sample, and performing normalization processing on the decomposition modulus; determination of the base learner: for IMF after EMD decompositioniSelecting different SVM kernel functions; obtaining a strong learning machine: selecting T groups of sub-SVM to use IMFiAnd corresponding input data with the wind direction and the air temperature as SVM are used for carrying out enhancement training on each sub-SVM by using Adaboost.RT to obtain a strong learning machine SVM (i); each IMFiComponent and residual rnSVM prediction value ofSuperposing to obtain a wind speed predicted value, and inputting the wind speed predicted value into a wind power conversion curve to obtain a predicted value of the electric field power; the invention reduces the influence of the learning machine parameters on the performance of the learning machine and improves the prediction precision.

Description

Short-term wind power plant power prediction method based on Adaboost-EMD-SVM
Technical Field
The invention relates to the technical field of wind power prediction, in particular to a short-term wind power plant power prediction method based on Adaboost-EMD-SVM.
Background
The wind energy has the characteristics of cleanness and reproducibility, and is an ideal new energy source. Because wind power generation has volatility, randomness and intermittence, the safe and stable operation of a power system is influenced by the fact that large-capacity wind power generation is connected into a power grid. The high-precision wind power prediction can reduce the adverse effects of the fluctuation and the intermittency of wind power on the quality of electric energy, improve the capability of the wind power to cross the power limit, and enhance the safety and the stability of the operation of a power system. When the prediction reaches enough precision, the reserve capacity of the system for balancing the power fluctuation can be reduced, so that the operation cost of the power grid is reduced, and the system is favorable for realizing the effective consumption of large-scale wind power. Therefore, the method for researching the short-term prediction of the power of the wind power plant has important theoretical and practical significance.
The short-term prediction of the wind power mainly comprises a physical method and a statistical method. The former needs to consider weather data and fan parameters, adopts a micro-meteorology theory or a computational fluid mechanics method for calculation, and has lower practicability compared with the former relying on weather forecast data; the relation between the meteorological environment of the wind power plant and the output of the wind power plant is found out through historical statistical data, and the output power of the wind power plant is predicted according to actually measured meteorological data of the wind power plant. The common wind power prediction method comprises a time sequence method, an extreme learning machine, a spatial correlation analysis method, a combined prediction method based on weighted average and the like. Most of the methods focus on the aspects of data preprocessing, data training (or data mining), algorithm optimization and the like, and the influences of prediction errors of single samples in a sample set on the overall weight cannot be reflected.
Disclosure of Invention
The invention aims to provide a short-term wind power plant power prediction method based on Adaboost-EMD-SVM, which has high precision and good generalization capability and can effectively solve the problem of influence of single sample prediction error in a sample set on the overall weight, thereby effectively improving the performance of the short-term wind power plant power prediction method.
The technical scheme for realizing the purpose of the invention is as follows: a short-term wind power plant power prediction method based on Adaboost-EMD-SVM comprises the following steps:
step 1, obtaining a sample set: taking historical meteorological data or numerical weather forecast data of a wind power plant as a sample set, taking wind speed, wind direction and air temperature as input values, and taking actual output power of the wind power plant as an output value;
step 2, processing of the sample set: EMD decomposition is carried out on the wind speed sample in the step 1 to obtain intrinsic modulus IMF1，IMF2，……，IMFnAnd a residual rnNormalization processing is carried out on the intrinsic modulus;
step 3, determining a base learner: for IMF after EMD decompositioniSelecting different SVM kernel functions, and determining a base learner, i is 1, 2, … …, n;
step 4, obtaining a strong learning machine: selecting T groups of sub-SVM by component IMF based on the SVM kernel function selected in the step 3iAnd performing enhancement training on each sub-SVM by using AdaboostnThe strong learning machine of (1);
step 5, wind power prediction output: obtaining each IMF by a strong learning machineiComponent and residual rnPredicted value of SVM of (1), each IMFiComponent and residual rnAnd superposing the SVM predicted values to obtain a wind speed predicted value, and inputting the wind speed predicted value into a wind power conversion curve to obtain a predicted value of the short-time wind power plant power.
Further, the EMD decomposition step in step 2 specifically includes:
1) let the original wind speed signal be x (t), x1(t) ═ x (t), initialization variable k ═ 1, i ═ 1;
2) identifying a wind speed signal xk(t) fitting all maximum and minimum points in the envelope curve, and calculating to obtain average value m of upper and lower envelope curvesk(t)；
3) Subtracting the average value of the envelope curve from the wind speed signal to obtain hi(t), i.e. hi(t)＝xk(t)-mk(t); if mk(t)>0.1, mixing hi(t) as a new velocity signal xk+1(t) repeating step 2); otherwise, hi(t) is the i-th IMF component of the wind speed signal, IMFiIs denoted by ci(t) which contains the shortest periodic component of the original wind speed sample signal;
4) calculating a residual component ri(t)＝x(t)-ci(t); if i ═ n stops the iteration, the remaining IMF components and residual r can be obtainednOtherwise, with the residual component ri(t) as a new raw wind speed signal x (t), execute 1 st), when i is not initialized anymore, i ═ i + 1.
Further, the method for determining the basis learner in step 3 includes: and adopting a radial basis kernel function for the IMF component with the fluctuation frequency higher than the sample frequency mean value, adopting a polynomial kernel function for the middle-low frequency IMF component with stable change, and adopting a linear kernel function for the rest.
Further, the step 4 of obtaining the strong learning machine includes:
(a) by component IMFiThe corresponding wind direction and air temperature are used as training input data of the SVM, the iteration number of Adaboost.RT is T, and meanwhile a threshold phi epsilon (0,1) is determined;
(b) t represents the t training, when t is 0, the error rate epsilon is takentWhen the value is 0, all the weights of i 1, 2, …, n are initialized
Figure BDA0002725113010000031
(c) Calling weak learning machine SVMi，tEstablishing a corresponding regression model ft(xi)→yi，ft(xi) Is a predicted value of the ith point, xiAs input variables of the regression model, yiThe measured value of the ith point is; the error for each sample is calculated according to the following formula:
AREt＝|(ft(xi)-yi)/yi|
Figure BDA0002725113010000032
is provided with
Figure BDA0002725113010000033
Wherein p is a power coefficient, p is 1, 2 or 3, and the weight is updated by using the following formula:
Figure BDA0002725113010000034
wherein Zt is a normalization factor;
adding 1 to the training times T, and repeating the step (c) until T is T;
(d) when T is T, T base learners are obtained, and the result SVM (i), namely f, is outputfin(x):
Figure BDA0002725113010000035
Further, the power coefficient p is 1.
Compared with the prior art, the invention has the following remarkable advantages:
(1) the EMD algorithm is adopted to carry out IMF decomposition on the wind speed sample, so that the non-stationarity of the sample data can be effectively weakened;
(2) the SVM algorithm has the common characteristics of the neural network algorithm, the learning speed is higher, the generalization capability is stronger, the SVM algorithm can approach any function with any precision theoretically, and the nonlinear characteristics of the wind power can be effectively processed;
(3) the quality of the kernel function and the parameters directly influences the quality of the SVM performance, and the AdaBoost. RT algorithm is adopted, so that the dependence of the SVM on kernel function and parameter selection can be reduced, and the training speed and prediction accuracy of the SVM algorithm are greatly improved;
(4) the prediction precision of the Adaboost.RT-EMD-SVM algorithm is greatly improved, and the method has certain advancement in the aspect of short-term wind power plant power prediction.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is a flow chart of EMD decomposition in the present invention.
FIG. 3 is a flowchart of SVM prediction steps in the present invention.
Fig. 4 is a strong learning machine training flow chart based on adaboost.
FIG. 5 is a diagram of IMF components after EMD decomposition of wind speed in an embodiment of the present invention.
Fig. 6 is a schematic diagram of a wind power prediction result of the adaboost.
Detailed Description
The following detailed description of the embodiments of the present invention will be given in conjunction with the accompanying drawings to make it clear to those skilled in the art how to practice the present invention. It should be understood that while the invention has been described in conjunction with the preferred specific embodiments thereof, that these embodiments are intended to illustrate and not limit the scope of the invention.
As shown in FIG. 1, a short-term wind power plant power prediction method based on Adaboost-EMD-SVM comprises the following steps:
step 1, obtaining a sample set: the method comprises the steps of utilizing historical meteorological data or numerical weather forecast data of a wind power plant as a sample set, wherein wind speed, wind direction and air temperature are taken as predicted input quantities;
step 2, processing a sample set: EMD decomposition is carried out on the wind speed sample in the step 1 to obtain intrinsic modulus IMF1，IMF2，……，IMFnAnd a residual rnNormalization processing is carried out on the intrinsic modulus; with reference to fig. 2, the specific steps are as follows:
1) let the original wind speed signal be x (t), x1(t) ═ x (t), initialization variable k ═ 1, i ═ 1;
2) identifying a wind speed signal xk(t) fitting all maximum and minimum points in the envelope curve, and calculating to obtain average value m of upper and lower envelope curvesk(t)；
3) Subtracting the average value of the envelope curve from the wind speed signal to obtain hi(t), i.e. hi(t)＝xk(t)-mk(t); if mk(t)>0.1, mixing hi(t) as a new velocity signal xk+1(t) repeating step 2); otherwise, hi(t) is the i-th IMF component of the wind speed signal, IMFiIs denoted by ci(t) which contains the shortest periodic component of the original wind speed sample signal;
4) calculating a residual component ri(t)＝x(t)-ci(t)；
If i ═ n stops the iteration, the residual IMF component and residual (residual component) r are obtainednOtherwise, with the residual component ri(t) repeating the above steps as a new original wind speed signal x (t), i ═ i + 1.
With reference to fig. 3 and 4, the method for establishing the regression model of adaboost.
Step 3, determining a base learner: aiming at the characteristics of IMF after EMD decomposition, different SVM kernel functions are selected, a radial basis kernel function is adopted for IMF components with fluctuation frequency higher than the sample frequency mean value, a polynomial kernel function is adopted for middle-low frequency IMF components with stable change, and linear kernel functions are adopted for the rest IMF components;
step 4, obtaining a strong learning machine: selecting different kernel functions of T groups based on the SVM kernel function selected in the step 3 according to different characteristics of IMF components to obtain T groups of sub-SVM (namely T groups of weak learning machines), and IMF (IMF) with the ith IMF componentiAnd taking the wind direction and the air temperature corresponding to the ith IMF component as input data of the SVM, and performing enhancement training on each sub-SVM by using Adaboost.
(a) By component IMFiThe corresponding wind direction and air temperature are used as training input data of the SVM, the iteration number of Adaboost.RT is also T, and meanwhile, a threshold value phi epsilon (0,1) is determined;
(b) t represents the T training, T is less than or equal to T, and when T is 0, the error rate epsilon is takentWhen the value is 0, all the weights of i 1, 2, …, n are initialized
Figure BDA0002725113010000051
(c) Calling weak learning machine SVMi，tEstablishing a corresponding regression model ft(xi)→yi，ft(xi) Is the predicted value of the ith point, yiIs the measured value of the ith point, xiTo return toThe error for each sample is calculated according to the following equation, given as the input variables to the model:
AREt＝|(ft(xi)-yi)/yi|
Figure BDA0002725113010000052
is provided with
Figure BDA0002725113010000053
Where p is the power coefficient, p is 1, 2 or 3, and the weights are updated using the following formula
Figure BDA0002725113010000054
Wherein Zt is a normalization factor.
Adding 1 to the training times T, and repeating the step (c) until T is T;
(d) and when T is equal to T, namely after T rounds of training, obtaining T base learners, and outputting a result SVM (i), namely ffin(x):
Figure BDA0002725113010000061
And 5, wind power prediction output: selecting T groups of sub-SVM based on the SVM kernel function selected in the step 3, and combining each IMF component and residual rnThe predicted values of the SVM are superposed to obtain a wind speed predicted value, and the predicted value is input into a wind power conversion curve to obtain a predicted value of the short-time wind power plant power.
Examples
The method comprises the steps of taking continuous 360 points from measured data of a No. 2 unit (the rated power of the unit is 800kW) of a certain wind power plant as a sample set, wherein the first 180 points are training samples, the last 180 points are testing samples, the training samples comprise wind speed, wind direction and wind power plant temperature, the wind speed of the training data is decomposed according to an EMD decomposition algorithm, and the decomposition result is shown in an attached figure 5; c 1-c 5 are data in sequenceIMF component from high frequency to low frequency after decomposition, r6For residual, the power coefficient takes 1.
In order to verify the prediction effect and precision of the method provided by the invention, the invention adopts the normalized absolute average error eNMAEAnd normalized root mean square error eNRMSEAs the prediction error index, the error when the iteration number is t can be respectively expressed as:
(1) normalized absolute mean error:
Figure BDA0002725113010000062
(2) normalized root mean square error:
Figure BDA0002725113010000063
wherein n is the number of samples, ft(xi) Is the predicted value of the ith point, yiIs the measured value of the ith point.
As can be seen from FIG. 6, the method provided by the invention has higher prediction accuracy, thereby verifying the effectiveness of the prediction model provided by the invention.
The method also predicts the power of the wind power plant through a BP neural network, the SVM and an EMD-SVM prediction model, and the result and the prediction error of the Adaboost.
TABLE 1 prediction error comparison table for each prediction method
Figure BDA0002725113010000071
As can be seen from table 1, the prediction method based on adaboost, rt-EMD-SVM, provided by the invention, effectively improves the accuracy of wind power prediction, and reduces the prediction error to below 10% of the international universal standard, thereby having certain advancement.
While the foregoing is directed to the preferred embodiment of the present invention, the present invention is not limited in any way by the foregoing examples, and it is understood that various changes and modifications may be made by one skilled in the art without departing from the spirit and scope of the invention.

Claims (5)

1. A short-term wind power plant power prediction method based on Adaboost-EMD-SVM is characterized by comprising the following steps:
step 1, obtaining a sample set: taking historical meteorological data or numerical weather forecast data of a wind power plant as a sample set, taking wind speed, wind direction and air temperature as input values, and taking actual output power of the wind power plant as an output value;
step 2, processing of the sample set: EMD decomposition is carried out on the wind speed sample in the step 1 to obtain intrinsic modulus IMF1，IMF2，……，IMFnAnd a residual rnNormalization processing is carried out on the intrinsic modulus;
step 3, determining a base learner: for IMF after EMD decompositioniSelecting different SVM kernel functions, and determining a base learner, i is 1, 2, … …, n;
step 4, obtaining a strong learning machine: selecting T groups of sub-SVM by component IMF based on the SVM kernel function selected in the step 3iAnd performing enhancement training on each sub-SVM by using AdaboostnThe strong learning machine of (1);
step 5, wind power prediction output: obtaining each IMF by a strong learning machineiComponent and residual rnPredicted value of SVM of (1), each IMFiComponent and residual rnAnd superposing the SVM predicted values to obtain a wind speed predicted value, and inputting the wind speed predicted value into a wind power conversion curve to obtain a predicted value of the short-time wind power plant power.
2. The method for predicting the power of the short-term wind power plant based on Adaboost-EMD-SVM according to claim 1 is characterized in that: the EMD decomposition step in the step 2 specifically comprises the following steps:
1) let the original wind speed signal be x(t)，x1(t) ═ x (t), initialization variable k ═ 1, i ═ 1;
2) identifying a wind speed signal xk(t) fitting all maximum and minimum points in the envelope curve, and calculating to obtain average value m of upper and lower envelope curvesk(t)；
3) Subtracting the average value of the envelope curve from the wind speed signal to obtain hi(t), i.e. hi(t)＝xk(t)-mk(t); if mk(t)>0.1, mixing hi(t) as a new velocity signal xk+1(t) repeating step 2); otherwise, hi(t) is the i-th IMF component of the wind speed signal, IMFiIs denoted by ci(t) which contains the shortest periodic component of the original wind speed sample signal;
4) calculating a residual component ri(t)＝x(t)-ci(t); if i ═ n stops the iteration, the remaining IMF components and residual r can be obtainednOtherwise, with the residual component ri(t) as a new raw wind speed signal x (t), execute 1 st), when i is not initialized anymore, i ═ i + 1.
3. The method for predicting the power of the short-term wind power plant based on Adaboost-EMD-SVM according to claim 1 is characterized in that: the method for determining the base learner in the step 3 comprises the following steps: and adopting a radial basis kernel function for the IMF component with the fluctuation frequency higher than the sample frequency mean value, adopting a polynomial kernel function for the middle-low frequency IMF component with stable change, and adopting a linear kernel function for the rest.
4. The method for predicting the power of the short-term wind power plant based on Adaboost-EMD-SVM according to claim 1 is characterized in that: the strong learning machine in the step 4 comprises the following steps:
(a) by component IMFiThe corresponding wind direction and air temperature are used as training input data of the SVM, the iteration number of Adaboost.RT is T, and meanwhile a threshold phi epsilon (0,1) is determined;
(b) t represents the t training, when t is 0, the error rate epsilon is takentWhen the value is 0, all the weights of i 1, 2, …, n are initialized
Figure FDA0002725111000000021
(c) Calling weak learning machine SVMi，tEstablishing a corresponding regression model ft(xi)→yi，ft(xi) Is a predicted value of the ith point, xiAs input variables of the regression model, yiThe measured value of the ith point is; the error for each sample is calculated according to the following formula:
AREt＝|(ft(xi)-yi)/yi|
Figure FDA0002725111000000022
is provided with
Figure FDA0002725111000000023
Wherein p is a power coefficient, p is 1, 2 or 3, and the weight is updated by using the following formula:
Figure FDA0002725111000000024
wherein Zt is a normalization factor;
adding 1 to the training times T, and repeating the step (c) until T is T;
(d) when T is T, T base learners are obtained, and the result SVM (i), namely f, is outputfin(x):
Figure FDA0002725111000000025
。
5. The Adaboost-EMD-SVM based short-term wind farm power prediction method according to claim 4, characterized in that: the power coefficient p takes 1.
CN202011100375.7A 2020-10-15 2020-10-15 Short-term wind power plant power prediction method based on Adaboost-EMD-SVM Withdrawn CN112329987A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011100375.7A CN112329987A (en) 2020-10-15 2020-10-15 Short-term wind power plant power prediction method based on Adaboost-EMD-SVM

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011100375.7A CN112329987A (en) 2020-10-15 2020-10-15 Short-term wind power plant power prediction method based on Adaboost-EMD-SVM

Publications (1)

Publication Number Publication Date
CN112329987A true CN112329987A (en) 2021-02-05

Family

ID=74313543

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011100375.7A Withdrawn CN112329987A (en) 2020-10-15 2020-10-15 Short-term wind power plant power prediction method based on Adaboost-EMD-SVM

Country Status (1)

Country Link
CN (1) CN112329987A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113344270A (en) * 2021-06-03 2021-09-03 上海交通大学 Wind resource prediction method and system based on integrated extreme learning machine

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113344270A (en) * 2021-06-03 2021-09-03 上海交通大学 Wind resource prediction method and system based on integrated extreme learning machine

Similar Documents

Publication Publication Date Title
Haque et al. A hybrid intelligent model for deterministic and quantile regression approach for probabilistic wind power forecasting
CN105654207A (en) Wind power prediction method based on wind speed information and wind direction information
CN105631558A (en) BP neural network photovoltaic power generation system power prediction method based on similar day
CN105354620A (en) Method for predicting fan generation power
CN110365053B (en) Short-term wind power prediction method based on delay optimization strategy
CN112186761B (en) Wind power scene generation method and system based on probability distribution
CN114792156A (en) Photovoltaic output power prediction method and system based on curve characteristic index clustering
CN113822418A (en) Wind power plant power prediction method, system, device and storage medium
CN113379164B (en) Load prediction method and system based on deep self-attention network
CN109272258B (en) Regional wind and solar power generation resource evaluation method based on K-means clustering
CN112819238A (en) Short-term wind power prediction method based on chaotic chicken flock optimization algorithm
CN115660887A (en) Photovoltaic output prediction method and system based on limited weather forecast information
CN114970952A (en) Photovoltaic output short-term prediction method and system considering environmental factors
CN112001537A (en) Short-term wind power prediction method based on gray model and support vector machine
CN111784543A (en) System model for predicting short-term photovoltaic power generation system power generation based on hidden Markov model
CN112329987A (en) Short-term wind power plant power prediction method based on Adaboost-EMD-SVM
CN108694475B (en) Short-time-scale photovoltaic cell power generation capacity prediction method based on hybrid model
CN114266416A (en) Photovoltaic power generation power short-term prediction method and device based on similar days and storage medium
CN113205228B (en) Method for predicting short-term wind power generation output power
CN112580844A (en) Meteorological data processing method, device, equipment and computer readable storage medium
CN114638396A (en) Photovoltaic power prediction method and system based on neural network instantiation
CN114298132A (en) Wind power prediction method and device and electronic equipment
CN116896093A (en) Online analysis and optimization method for grid-connected oscillation stability of wind farm
CN111476437A (en) Method and device for predicting ultra-short term interval of photovoltaic power generation power
CN116565840A (en) High-precision wind speed soft measurement method for wind power prediction of wind power plant

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
WW01 Invention patent application withdrawn after publication
WW01 Invention patent application withdrawn after publication

Application publication date: 20210205