CN112270434A - Wind-solar integrated power prediction method - Google Patents

Wind-solar integrated power prediction method Download PDF

Info

Publication number
CN112270434A
CN112270434A CN202011137621.6A CN202011137621A CN112270434A CN 112270434 A CN112270434 A CN 112270434A CN 202011137621 A CN202011137621 A CN 202011137621A CN 112270434 A CN112270434 A CN 112270434A
Authority
CN
China
Prior art keywords
power
wind
photovoltaic
prediction
data
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
CN202011137621.6A
Other languages
Chinese (zh)
Inventor
王金浩
李胜文
常潇
雷达
李慧蓬
南晓强
王锬
樊瑞
张世锋
赵军
张敏
肖莹
高枫
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.)
Beijing Qianrunhe Technology Co ltd
Electric Power Research Institute of State Grid Shanxi Electric Power Co Ltd
State Grid Shanxi Electric Power Co Ltd
Original Assignee
Beijing Qianrunhe Technology Co ltd
Electric Power Research Institute of State Grid Shanxi Electric Power Co Ltd
State Grid Shanxi 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 Beijing Qianrunhe Technology Co ltd, Electric Power Research Institute of State Grid Shanxi Electric Power Co Ltd, State Grid Shanxi Electric Power Co Ltd filed Critical Beijing Qianrunhe Technology Co ltd
Priority to CN202011137621.6A priority Critical patent/CN112270434A/en
Publication of CN112270434A publication Critical patent/CN112270434A/en
Pending 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
    • 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
    • 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
    • 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)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Software Systems (AREA)
  • Tourism & Hospitality (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • General Business, Economics & Management (AREA)
  • Molecular Biology (AREA)
  • Artificial Intelligence (AREA)
  • Marketing (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Primary Health Care (AREA)
  • Water Supply & Treatment (AREA)
  • Public Health (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Wind Motors (AREA)

Abstract

The invention relates to the technical field of comprehensive energy, and particularly discloses a wind-solar integrated power prediction method, which comprises the following steps: step S1, selecting model input data; step S2, carrying out noise reduction processing on the photovoltaic and wind power original power data through DT-CWT dual-tree complex wavelets; s3, carrying out LSTM multi-port long-time memory network prediction model training on the processed data; step S4, predicting wind power P by using least square methody1Photovoltaic predicted power Py2Fitting the data to finally obtain the wind power predicted power Py1Photovoltaic predicted power Py2The prediction curve of (1). The invention can effectively reduce the influence caused by power noise in prediction and improve wind-light complementary power supplyThe power prediction accuracy of the power system provides technical support for wind-solar integrated power prediction and reasonable power grid scheduling.

Description

Wind-solar integrated power prediction method
Technical Field
The invention relates to the technical field of comprehensive energy, in particular to a wind and light integrated power prediction method.
Background
In recent years, due to the rapid construction of wind power generation and photovoltaic power generation, randomness, intermittence and sudden fluctuation caused by simultaneous wind and light grid connection bring greater challenges to the stability of a power grid and the guarantee of the power quality, and huge pressure is brought to power grid dispatchers and operation and maintenance maintainers to arrange planned power failure and load transfer. In order to solve the problem of large-scale distributed micro-grid energy access, research teams in China also develop deep discussion aiming at the problems.
The national scholars also make a lot of researches on wind and light integrated power prediction methods, and the problems that the power noise influence is large and various types of power variables cannot be caused exist in the design of a power prediction model of a wind and light complementary power supply system at present, and certain limitations exist. Most prediction methods lack consideration of multidimensional environmental factors, so that the accuracy cannot be guaranteed in practical engineering application. With the wide application of the wind and light integrated system, the power prediction research of the wind and light combined power supply system is of great significance.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a wind and light integrated power prediction method aiming at the defects in the prior art, so that the influence caused by power noise in prediction can be effectively reduced, the power prediction accuracy of a wind and light complementary power supply system is improved, and a technical support is provided for wind and light integrated power prediction and reasonable power grid scheduling.
In order to solve the technical problems, the technical scheme of the invention is as follows:
a wind-solar integrated power prediction method comprises the following steps:
step S1, selecting model input data;
step S2, carrying out noise reduction processing on the photovoltaic and wind power original power data through DT-CWT dual-tree complex wavelets;
s3, carrying out LSTM multi-port long-time memory network prediction model training on the processed data;
step S4, predicting wind power P by using least square methody1Photovoltaic predicted power Py2Fitting the data to finally obtain the wind power predicted power Py1Photovoltaic predicted power Py2The prediction curve of (1).
Preferably, in the step S1, the model input data includes wind power output power P1, photovoltaic output power P1Output power P2 and wind speed f1Ambient temperature f2Humidity f3Installed capacity f of the apparatus4And the operating voltage f of the apparatus5Total radiation g1Ambient temperature g2Temperature g of the component3Altitude g4Angle of incidence g5。
Preferably, in the step S1, the DT-CWT dual-tree complex wavelet includes two parallel trees, an upper branch and a lower branch, the upper branch is a real part of the even-numbered complex wavelet transform, the lower branch is an imaginary part of the DT-CWT, and the real part and the imaginary part are respectively composed of two high-pass filters and two low-pass filters.
Preferably, in step S3, the LSTM multiport long-and-short memory network prediction model is:
ft＝σ(Wfg[ht-1,xt]+bf)
it＝σ(Wig[ht-1,xt]+bi)
gt＝tanh(Wgg[ht-1,xt]+bg)
ot＝σ(Wog[ht-1,xt]+bo)
ct＝ft*ct-1+it*gt
ht＝ot*tanh(ct)
wherein f ist,it,gt,otOutput values of the forgetting gate, the input gate, the update gate and the output gate, Wf、Wi、Wg、WoLink weights, b, for respective doorsf、bi、bg、boThe inputs of the four gates comprise the output value ht-1 at the t-1 moment of the LSTM and the input value x at the current moment respectively for the offset of each gatet，CtAnd a temporary information storage unit, wherein the sigma is a sigmoid excitation function. The LSTM belongs to a recurrent neural network, the internal structure of the LSTM mainly eliminates or increases corresponding information through a gate, and the LSTM gets rid of the existence of the traditional Recurrent Neural Network (RNN) trainingThe method comprises the steps of establishing a storage unit inside, storing a sequence of the storage unit through a memorability event structure, further extracting the sequence from a training model, adding a forgetting gate and an input gate to an LSTM (local side switch) in order to enhance the information storage inside the network, wherein the forgetting gate is used for eliminating invalid information components in the network, the input gate is used for adding corresponding effective new information, and the two are combined to form the network model with signal filtering and optimization.
Preferably, in the step S3, the method further includes reducing the noise-reduced wind power output power Ph1And the influencing factors form a wind power input matrix X1And the photovoltaic output power P after noise reductionh2And influencing factor photovoltaic input matrix X2，X1And X2The matrix model is:
X1＝[Ph1 f1 f2 f3 f4 f5]
X1＝[Ph2 g1 g2 g3 g4 g5]。
by adopting the technical scheme, the wind-solar integrated power prediction method provided by the invention decomposes wind power and photovoltaic output power signals after noise reduction by utilizing DT-CWT, combines all influence factors of wind power and photovoltaic as multi-port LSTM input, and finally obtains a prediction curve of wind-solar integrated output power through training of an LSTM network. By adopting the prediction model based on DT-CWT and LSTM, the noise reduction performance can be improved well, the prediction accuracy of wind-solar integrated output power can be improved, and guidance and support are provided for reasonable electric power scheduling, operation and maintenance.
Drawings
FIG. 1 is a flow chart of a wind-solar integrated power prediction method according to the present invention;
FIG. 2 is a schematic structural diagram of a DT-CWT dual tree complex wavelet in the present invention;
FIG. 3 is a schematic structural diagram of an LSTM multi-port long-and-short-term memory network prediction model in the invention.
Detailed Description
The following further describes embodiments of the present invention with reference to the drawings. It should be noted that the description of the embodiments is provided to help understanding of the present invention, but the present invention is not limited thereto. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
In the description of the present invention, it is to be understood that the terms "central," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," "circumferential," and the like are used in the orientations and positional relationships indicated in the drawings for convenience in describing the invention and to simplify the description, and are not intended to indicate or imply that the referenced devices or elements must have a particular orientation, be constructed and operated in a particular orientation, and are therefore not to be considered limiting of the invention.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
As shown in fig. 1, the flow chart of the wind-solar integrated power prediction method of the present invention includes the following steps: step S1, selecting model input data; step S2, carrying out noise reduction processing on the photovoltaic and wind power original power data through DT-CWT dual-tree complex wavelets; s3, carrying out LSTM multi-port long-time memory network prediction model training on the processed data; step S4, predicting wind power P by using least square methody1Photovoltaic predicted power Py2Fitting the obtained data to obtain the final productWind power prediction power Py1Photovoltaic predicted power Py2The prediction curve of (1). The wind-light integrated power prediction method based on dual-tree complex wavelet (DT-CWT) and long-and-short time memory network (LSTM) is designed on the basis of traditional load power prediction. Finally, model verification is carried out through field data of the wind and light complementary power supply system, and results show that the method can effectively reduce the influence caused by power noise in prediction and improve the power prediction accuracy of the wind and light complementary power supply system.
Specifically, step S1: determining model input data;
(1) wind farm influencing factor
According to statistical investigation, the factors influencing the output power of the wind turbine are mainly embodied in two aspects, one is environmental factor: wind speed, wind direction, ambient temperature, humidity, air pressure; two wind power installation positions and the influence factors of the equipment are as follows: generator type, installed capacity, equipment running state and the like. Selecting wind speed f by comprehensive consideration1Ambient temperature f2Humidity f3Installed capacity f of the apparatus4And the operating voltage f of the apparatus5These 5 key factors serve as the input set for the subsequent power prediction model.
(2) Photovoltaic power factor
The factors influencing the output power of the photovoltaic power generation mainly comprise: total radiation, direct radiation, scattered radiation, ambient temperature, component temperature, relative humidity, air pressure, wind speed, wind direction, cloud amount and other meteorological factors as well as longitude and latitude, altitude, incidence angle, photovoltaic power station capacity and other basic information, but the key influence factor mainly comprises total radiation g1Ambient temperature g2Temperature g of the component3Altitude g4Angle of incidence g5. In order not to influence the model calculation speed, key photovoltaic influence factors are selected: total radiation g1Environment, environmentTemperature g2Temperature g of the component3Altitude g4Angle of incidence g5As input to the subsequent LSTM model.
(3) Model input data determination
The model input data comprise wind power output power P1, photovoltaic output power P2 and wind speed f1Ambient temperature f2Humidity f3Installed capacity f of the apparatus4And the operating voltage f of the apparatus5Total radiation g1Ambient temperature g2Temperature g of the component3Altitude g4Angle of incidence g5。
Step S2: data noise reduction and preprocessing;
in the wind power and photovoltaic power prediction process, especially in the wind power prediction process, electric field high noise and low-dimensional historical data information are doped, furthermore, due to the influence of a plurality of factors, the measured data is abnormal, and the noise and abnormal data soybean milk influence the power prediction accuracy. The dual complex wavelet is based on the traditional wavelet analysis, can keep the original signal not to translate and carry out accurate reconstruction of the signal under the condition of no data distortion, carries out power data decomposition and reconstruction on the dual complex wavelet, and has the excellent characteristics of small redundancy, high calculation speed and the like. Dual tree complex wavelet is pair phih(t) and phig(t) to satisfy the Hilbert transform pair requirement,. psih(t) and ψg(t) as real and imaginary parts, having complex wavelet coefficients phi (t) ═ phih(t)+iφg(t)。
As can be seen from FIG. 2, the DT-CWT comprises two parallel trees of an upper branch and a lower branch, the upper branch is a real part of the even complex wavelet transform, the lower branch is an imaginary part of the DT-CWT, and the real part and the imaginary part are respectively composed of two high-pass filters and two low-pass filters. FIG. 2 is a decomposition process of the DT-CWT, and an inverse process of the decomposition process is a data reconstruction process.
The DT-CWT transformation process is as follows:
carrying out DT-CWT conversion of N layers on the power data signal, and decomposing the j layer data xi＝xi re+jxi imPerforming signal reconstruction, D of power under different scalesThe T-CWT transform will form a matrix of power data components, as shown in equation (1)
Figure BDA0002737195690000041
In the formula, Xi＝(xi(1),xi(2),…,xi(k) I ═ 1,2, …, m, and represents the power component of the i-th layer of the DT-CWT. The power data matrix Xn includes noisy power signals. XnThe composition formula is as follows:
Xn＝X+W (2)
in the formula, X does not include a noise power component matrix. W contains a noise component to the power data.
Wind power output power P by the method1And the photovoltaic output power P2Decomposing the data, and performing inverse DT-CWT (differential time constant-constant transform) on the wavelet transform coefficient obtained by decomposition to obtain two corresponding groups of wavelet components, wherein the components are the wind power output power P after noise reductionh1Photovoltaic output power Ph2。
Step S3: training an LSTM network model;
the Long Short-Term Memory network (LSTM) is a time-cycle neural network, which is specially designed to solve the Long-Term dependence problem of the general RNN (cyclic neural network), and all RNNs have a chain form of repeated neural network modules. In the standard RNN, this repeated structure block has only a very simple structure, e.g. one tanh layer. To minimize training errors, the Gradient descent method (Gradient device) is as follows: the application of a time-sequential reverse transfer algorithm can be used to modify the weight of each time according to the error. The major problem of gradient descent in Recurrent Neural Networks (RNNs) was first discovered in 1991, where the error gradient disappears exponentially with the length of time between events. When the LSTM tile is set, the error is also calculated with the rewind, from when output affects each gate in the input phase until this value is filtered out. Therefore, normal reciprocal transmission-like nerves is a method for effectively training the LSTM block to remember long-time numerical values. The LSTM belongs to a recurrent neural network, the internal structure of the LSTM is mainly removed or added with corresponding information through a gate, the problems of insufficient gradient and the like in the traditional Recurrent Neural Network (RNN) training are solved, a storage unit is arranged in the LSTM, the sequence of the LSTM is stored through a memorability event structure, the LSTM is further extracted from a training model, in order to enhance the information storage context in the network, a forgetting gate and an input gate are added to the LSTM, the forgetting gate is used for removing invalid information components in the network, the input gate is added with corresponding valid new information, the LSTM and the input gate are combined with each other to form a network model with signal filtering and optimization, and the LSTM algorithm model is as follows:
ft＝σ(Wfg[ht-1,xt]+bf)
it＝σ(Wig[ht-1,xt]+bi)
gt＝tanh(Wgg[ht-1,xt]+bg)
ot＝σ(Wog[ht-1,xt]+bo)
ct＝ft*ct-1+it*gt
ht＝ot*tanh(ct) (3)
wherein f ist,it,gt,otThe output values of the forgetting gate, the input gate, the updating gate and the output gate are respectively. Wf、Wi、Wg、WoLink weights, b, for respective doorsf、bi、bg、boRespectively, the offset of each door. The inputs to the four gates include the output value ht-1 at time t-1 of the LSTM and the input value x at the present timet。CtA temporary information storage unit. σ is the sigmoid excitation function. The LSTM structure is shown in fig. 3.
Wind power output power P after noise reductionh1Photovoltaic output power Ph2Wind power output power influencing factor (wind speed f)1Ambient temperature f2Humidity f3Installed capacity f of the apparatus4And the operating voltage f of the apparatus5) Photovoltaic output power influencing factor (total radiation g)1Ambient temperature g2Temperature g of the component3Altitude g4Angle of incidence g5) The wind power output power P after noise reduction is realizedh1And the influencing factors form a wind power input matrix X1And the photovoltaic output power P after noise reductionh2And influencing factor photovoltaic input matrix X2，X1And X2The matrix model is shown in equation (4). Mixing X1And X2Inputting the wind power into a long-time memory network for training a prediction model to finally obtain wind power prediction power Py1Photovoltaic predicted power Py2The data of (1).
X1＝[Ph1 f1 f2 f3 f4 f5]
X1＝[Ph2 g1 g2 g3 g4 g5](4)
Step S4: fitting a prediction curve
The method utilizes a least square method to predict the wind power Py1Photovoltaic predicted power Py2Fitting the data to finally obtain the wind power predicted power Py1Photovoltaic predicted power Py2The prediction curve of (1). Least squares (also known as the least squares method) is a mathematical optimization technique. It finds the best functional match of the data by minimizing the sum of the squares of the errors. Unknown data can be easily obtained by the least square method, and the sum of squares of errors between these obtained data and actual data is minimized. The least squares method can also be used for curve fitting. Other optimization problems may also be expressed in a least squares method by minimizing energy or maximizing entropy.
The least squares fit model is as follows:
Figure BDA0002737195690000051
Figure BDA0002737195690000052
Figure BDA0002737195690000053
in the formula aiIs a polynomial parameter, yiAverage value of power.
The method has the advantages that the design is reasonable, the structure is unique, the noise signal and the noise reduction power signal are decomposed by carrying out DT-CWT decomposition on the original data, the noise reduction power signal is combined with the photovoltaic power influence factor and the wind power influence factor of the wind-solar integrated power supply system to serve as the output of LSTM model training, the photovoltaic power prediction data and the wind power prediction data of the microgrid energy system are finally obtained, the curve fitting of the least square method is combined, the photovoltaic power prediction curve and the wind power prediction curve are obtained, the influence caused by power noise in prediction can be effectively reduced, the power prediction accuracy of the wind-solar complementary power supply system is improved, and the technical support is provided for the aspects of wind-solar integrated power prediction and reasonable scheduling of.
The embodiments of the present invention have been described in detail with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. It will be apparent to those skilled in the art that various changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, and the scope of protection is still within the scope of the invention.

Claims (5)

1. A wind-solar integrated power prediction method is characterized in that: the method comprises the following steps:
step S1, selecting model input data;
step S2, carrying out noise reduction processing on the photovoltaic and wind power original power data through DT-CWT dual-tree complex wavelets;
s3, carrying out LSTM multi-port long-time memory network prediction model training on the processed data;
step S4, predicting wind power P by using least square methody1Photovoltaic pre-stagePower measurement Py2Fitting the data to finally obtain the wind power predicted power Py1Photovoltaic predicted power Py2The prediction curve of (1).
2. The wind-solar integrated power prediction method according to claim 1, characterized in that: in the step S1, the model input data includes wind power output power P1, photovoltaic output power P2, and wind speed f1Ambient temperature f2Humidity f3Installed capacity f of the apparatus4And the operating voltage f of the apparatus5Total radiation g1Ambient temperature g2Temperature g of the component3Altitude g4Angle of incidence g5。
3. The wind-solar integrated power prediction method according to claim 1, characterized in that: in step S1, the DT-CWT dual-tree complex wavelet includes two parallel trees, an upper branch and a lower branch, where the upper branch is a real part of the dual-even-number complex wavelet transform, the lower branch is an imaginary part of the DT-CWT, and the real part and the imaginary part are respectively composed of two high-pass filters and two low-pass filters.
4. The wind-solar integrated power prediction method according to claim 2, characterized in that: in step S3, the LSTM multiport long-short term memory network prediction model is:
ft＝σ(Wfg[ht-1,xt]+bf)
it＝σ(Wig[ht-1,xt]+bi)
gt＝tanh(Wgg[ht-1,xt]+bg)
ot＝σ(Wog[ht-1,xt]+bo)
ct＝ft*ct-1+it*gt
ht＝ot*tanh(ct)
wherein f ist,it,gt,otOutput values of the forgetting gate, the input gate, the update gate and the output gate, Wf、Wi、Wg、WoLink weights, b, for respective doorsf、bi、bg、boThe inputs of the four gates comprise the output value ht-1 at the t-1 moment of the LSTM and the input value x at the current moment respectively for the offset of each gatet，CtAnd a temporary information storage unit, wherein the sigma is a sigmoid excitation function.
5. The wind-solar integrated power prediction method according to claim 2, characterized in that: in step S3, the method further includes reducing the noise of the wind power output power Ph1And the influencing factors form a wind power input matrix X1And the photovoltaic output power P after noise reductionh2And influencing factor photovoltaic input matrix X2，X1And X2The matrix model is:
X1＝[Ph1 f1 f2 f3 f4 f5]
X1＝[Ph2 g1 g2 g3 g4 g5]。
CN202011137621.6A 2020-10-22 2020-10-22 Wind-solar integrated power prediction method Pending CN112270434A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011137621.6A CN112270434A (en) 2020-10-22 2020-10-22 Wind-solar integrated power prediction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011137621.6A CN112270434A (en) 2020-10-22 2020-10-22 Wind-solar integrated power prediction method

Publications (1)

Publication Number Publication Date
CN112270434A true CN112270434A (en) 2021-01-26

Family

ID=74341472

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011137621.6A Pending CN112270434A (en) 2020-10-22 2020-10-22 Wind-solar integrated power prediction method

Country Status (1)

Country Link
CN (1) CN112270434A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113139817A (en) * 2021-04-28 2021-07-20 北京沃东天骏信息技术有限公司 Data classification method, data classification device, medium, and electronic apparatus

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113139817A (en) * 2021-04-28 2021-07-20 北京沃东天骏信息技术有限公司 Data classification method, data classification device, medium, and electronic apparatus

Similar Documents

Publication Publication Date Title
CN109492822B (en) Air pollutant concentration time-space domain correlation prediction method
CN109214592B (en) Multi-model-fused deep learning air quality prediction method
CN105184391B (en) Wind farm wind velocity and power forecasting method based on wavelet decomposition and support vector machines
CN108197648A (en) A kind of Fault Diagnosis Method of Hydro-generating Unit and system based on LSTM deep learning models
CN106920007B (en) PM based on second-order self-organizing fuzzy neural network2.5Intelligent prediction method
CN112434848B (en) Nonlinear weighted combination wind power prediction method based on deep belief network
CN114201924B (en) Solar irradiance prediction method and system based on transfer learning
CN111415010A (en) Bayesian neural network-based wind turbine generator parameter identification method
CN112712215A (en) Wind power plant power prediction method
CN112270434A (en) Wind-solar integrated power prediction method
CN112070272A (en) Method and device for predicting icing thickness of power transmission line
CN115603446A (en) Power distribution station area operation monitoring system based on convolution neural network and cloud edge synergistic effect
Qu et al. Research on short‐term output power forecast model of wind farm based on neural network combination algorithm
CN111488974B (en) Ocean wind energy downscaling method based on deep learning neural network
CN113985156A (en) Intelligent fault identification method based on transformer voiceprint big data
CN116565856A (en) Power distribution network state estimation method considering unknown topology change
CN116432359A (en) Variable topology network tide calculation method based on meta transfer learning
Wenting et al. Short-term wind power forecasting model based on random forest algorithm and TCN
Zhang et al. Short‐Term Power Prediction of Wind Power Generation System Based on Logistic Chaos Atom Search Optimization BP Neural Network
CN113705885B (en) VMD, XGBoost and TCN fused power distribution network voltage prediction method and system
Tan et al. Denoising convolutional neural networks based dust accumulation status evaluation of photovoltaic panel
CN113537573B (en) Wind power operation trend prediction method based on double space-time feature extraction
CN114741952A (en) Short-term load prediction method based on long-term and short-term memory network
CN112801815B (en) Power communication network fault early warning method based on federal learning
CN112232714B (en) Deep learning-based risk assessment method for distribution network under incomplete structural parameters

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20210126