CN109840629B - Photovoltaic power prediction method based on wavelet transform-dendritic neuron model - Google Patents
Photovoltaic power prediction method based on wavelet transform-dendritic neuron model Download PDFInfo
- Publication number
- CN109840629B CN109840629B CN201910051839.0A CN201910051839A CN109840629B CN 109840629 B CN109840629 B CN 109840629B CN 201910051839 A CN201910051839 A CN 201910051839A CN 109840629 B CN109840629 B CN 109840629B
- Authority
- CN
- China
- Prior art keywords
- data
- photovoltaic power
- neuron model
- dendritic
- historical
- 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.)
- Active
Links
Images
Landscapes
- Photovoltaic Devices (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a photovoltaic power prediction method based on a wavelet transform-dendritic neuron model, which comprises the following steps of: acquiring historical photovoltaic power data and historical meteorological data at corresponding moments; decomposing the photovoltaic power historical data and the historical meteorological data into a plurality of high-frequency components and low-frequency components by utilizing wavelet transformation; feeding the decomposed signal matrixes of the photovoltaic power historical data and the historical meteorological data to a dendritic neuron for model training; reconstructing output data of the dendritic neurons by utilizing wavelet transformation, and adjusting the wavelet decomposition times and the number of hidden layers of the dendritic neuron model according to a reconstruction result and the relative error of the actual power generation power until the relative error is not higher than an expected error; finally, taking historical meteorological data and photovoltaic power generation power as test data, wherein the calculation result of the test data is photovoltaic power prediction output; according to the photovoltaic power prediction method, the accuracy and the convergence rate of photovoltaic power prediction are improved by combining wavelet transformation and the dendritic neuron model.
Description
Technical Field
The invention belongs to the field of new energy, and particularly relates to a photovoltaic power prediction method based on a wavelet transform-dendritic neuron model.
Background
Photovoltaic power generation is easily influenced by external factors such as weather indexes, and has different output characteristics in different environments. As more and more photovoltaic power generation power is incorporated into the grid, intermittent, periodic, and random fluctuations thereof have increasingly severe impacts on the safe and stable operation of the grid. If the output power of the photovoltaic power generation can be effectively and accurately predicted, the risk of the photovoltaic power generation when the photovoltaic power generation is incorporated into a power system can be greatly reduced, and the stable operation of the power system is ensured. The existing photovoltaic power generation prediction methods comprise a single model prediction method and a mixed model prediction method, wherein the single model has the advantages that the model is simple, but the prediction precision is low, the prediction precision of the mixed model is high, but the model is often too complex and the convergence is slow. How to make the prediction model have higher precision and the model itself is simpler becomes a dilemma.
Disclosure of Invention
Aiming at the problem that the prediction precision and the model simplicity cannot be combined together in the photovoltaic power prediction method in the prior art, the invention provides a photovoltaic power prediction method based on a wavelet transform-dendritic neuron model; the specific technical scheme is as follows:
a method for photovoltaic power prediction based on a wavelet transform-dendritic neuron model, the method comprising:
s1, extracting photovoltaic power historical data at the time of the specified position, acquiring historical meteorological data at the corresponding time, acquiring photovoltaic power data corresponding to the historical meteorological data, and forming sequence data by the historical meteorological data and the photovoltaic power data;
s2, decomposing the sequence data by wavelet transform, obtaining a specified number of low-frequency components after decomposing the sequence data by down-sampling of a low-pass filter, and obtaining a specified number of high-frequency components after decomposing the sequence data by down-sampling of a high-pass filter;
s3, constructing a dendritic neuron model, and selecting a designated part from the photovoltaic power historical data and the historical meteorological data to form an input signal matrix, wherein the signal matrix is used for training the dendritic neuron model; decomposing the input signal matrix by using wavelet transformation and feeding the decomposed result to the dendritic neuron model for training to obtain output data synchronized with the dendritic neuron model;
s4, reconstructing the output data by adopting wavelet transformation, taking the reconstructed result as photovoltaic power prediction data, calculating the relative error between the reconstructed result and the actual photovoltaic power data at a specified position, and adjusting the number of wavelet decomposition times and the number of hidden layers of the dendritic neuron model based on the relative error;
and S5, inputting the remaining photovoltaic power historical data and the historical meteorological data which are divided to form the signal matrix as test data, adjusting the number of the hidden layers of the dendritic neuron model, calculating an output result of the test data according to the set wavelet decomposition times, and outputting the output result as the prediction of the photovoltaic power.
Further, step S1 includes: and acquiring the historical meteorological data by adopting a correlation analysis method, taking the historical photovoltaic power data and the historical meteorological data at the same moment as a group, and extracting data between adjacent groups according to a time period with a set specified length.
Further, the dendritic neuron model comprises four parts of synapses, branches, cell membranes and cell bodies, wherein:
x for synapse1,x2,...,xiRepresentation, used as input to said dendritic neuron model and made in said synapse
After operation, transmitting the branch to the appointed branch, wherein m represents the serial number of the branch, and k is a positive parameter; and thetaim、wimFor setting a judgment threshold value when xiWhen the value of (D) is greater than the judgment threshold, the synapse does
Calculating;
the branch is doing at the receiving of the synapse
And performing a cumulative multiplication operation after the operation result:
i represents the number of input variables;
performing an accumulation operation of all the branches on the cell membrane:
m represents the number of branches;
accumulation of said cell body on said cell membraneResult of addition operation is based on
Calculating to obtain a final result, wherein thetasomaTo set the threshold.
Further, the dendritic neuron model is a feed-forward multilayer network, and the synapses are functions of
Function of the branch
Function of the cell membrane
Function of the cell body
Are all differential functions.
Further, the calculating the connection type between the feedforward multilayer networks by using a back propagation method comprises the following steps:
formula is formed by
Calculating a least square error E of an actual output and a required output of the dendritic neuron model;
according to the gradient descent learning method, by the equation:
threshold θ for the synapseim、wimCorrection is made to reduce the error, where η is the learning efficiency and θ is the thresholdim、wimThe corresponding correction formula obtained after partial derivative calculation is as follows:
wim(t+1)＝wim(t)+Δwim
θim(t+1)＝θim(t)+Δθim。
in the embodiment of the present invention, in step S1, a correlation analysis method is used to determine the selection of the category and the quantity of the historical meteorological data; preferably, the covariance analysis method is adopted to obtain the type and quantity of the historical meteorological data, specifically through an expression
In the formula, X represents meteorological data, and Y represents photovoltaic power.
In the embodiment of the invention, the decomposition of historical meteorological data and photovoltaic power data is carried out by adopting discrete wavelet transform, specifically by an expression
Implementation, where T is the length of signal x (T); the scaling and translation parameters are functions of integer variables m and n, where a is 2mAnd b is n2mAnd t is a discrete time index.
Compared with the prior art, the photovoltaic power prediction method based on the wavelet transform-dendritic neuron model has the beneficial effects that: before input data enter a neural network, a photovoltaic power prediction model decomposes and reconstructs the input data by utilizing wavelet transformation, can effectively compress and reduce noise of the data, not only simplifies the data amount of training of a dendritic neuron model, but also reduces noise influencing prediction precision in the data; in addition, compared with the traditional neural network model, a single neuron of the dendritic neuron model can perform more complex nonlinear operation, and the number of the neurons is less under the condition of the same precision, so that the convergence rate of the model and the prediction precision can be further improved. Therefore, the model provided by the invention is simpler and has higher convergence speed.
Drawings
FIG. 1 is a flowchart illustration of a photovoltaic power prediction method based on wavelet transform-dendritic neuron model according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a dendritic neuron model according to an embodiment of the present invention;
FIG. 3 is a flow diagram illustration of a photovoltaic power prediction operation utilizing the method of the present invention;
FIG. 4 is a diagram illustrating a simulation comparison between the method of the present invention and the prior art.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention.
Example one
Referring to fig. 1, in an embodiment of the present invention, there is provided a photovoltaic power prediction method based on a wavelet transform-dendritic neuron model, the method including:
s1, extracting photovoltaic power historical data at the moment of the specified position, acquiring historical meteorological data at the corresponding moment, acquiring photovoltaic power data corresponding to the historical meteorological data, and forming sequence data by the historical meteorological data and the photovoltaic power data; the method comprises the steps of obtaining historical meteorological data by a correlation analysis method, taking the historical photovoltaic power data and the historical meteorological data at the same moment as a group, and extracting data between adjacent groups according to a time period with a set specified length.
Preferably, the covariance analysis method is used for obtaining the type and the quantity of the historical meteorological data, and the expression is used for obtaining the type and the quantity of the historical meteorological data
In the formula, X represents meteorological data, and Y represents photovoltaic power; of course, this is only a preferred embodiment of the present invention, and in other embodiments, other correlation analysis methods may be selected according to actual situations to obtain the type and quantity of the historical meteorological data, and the present invention is not limited or fixed.
S2, decomposing the sequence data by wavelet transform, obtaining low frequency components of the sequence data after decomposing the sequence data by down-sampling of low pass filter, and obtaining sequence data components by down-sampling of high pass filterA specified number of high frequency components after the solution; in the embodiment of the present invention, the decomposition of the historical meteorological data and the photovoltaic power data can be performed by either continuous wavelet transform or discrete wavelet-elimination transform, but in actual operation, in order to avoid generating a large amount of redundant information by continuous scaling and mother wavelet transformation using the continuous wavelet transform, the method of the present invention preferably performs the decomposition of the historical meteorological data and the photovoltaic power data by using the discrete wavelet transform, specifically by using the expression
Implementation, where T is the length of signal x (T); the scaling and translation parameters are functions of integer variables m and n, where a is 2mAnd b is n2mAnd t is a discrete time index.
S3, constructing a dendritic neuron model, feeding a signal matrix of sequence data decomposed by wavelet transform to the dendritic neuron model for training, and acquiring synchronous output data of the dendritic neuron model; specifically, the dendritic neuron model comprises four parts of synapses, branches, cell membranes and cell bodies, wherein the synapses use x1,x2,...,xiRepresentation, used as input to a model of dendritic neurons and done in synapses
After operation, the data are transmitted to a specified branch, wherein m represents the serial number of the branch, k is a positive parameter, and thetaim、wimIndicating a set threshold; branch is done at received synapse
Using a function after the result of the operation
Performing a cumulative multiplication operation; the branches transmit the multiplication result to the cell membrane after multiplication operation, and the cell membrane adopts function
Performing accumulation operation of all branches, and transmitting the accumulation operation result to the cell body; specific cell body adopted function
Calculating to obtain a final result, wherein thetasomaTo set the threshold.
In particular embodiments, the dendritic neuron model is a feed-forward multi-layer network, and the function described above with respect to synapses
Function on branches
Function of cell membrane
And functions relating to cell bodies
Are all differential functions; based on the method, the invention adopts a back propagation method to calculate the connection type between the feedforward multilayer networks, and specifically comprises the following steps:
first formula of
Calculating a minimum square error E of the actual output and the required output of the dendritic neuron model; then according to the gradient descent learning method, by equation
Threshold for synapse θim、wimMaking a correction to reduce the error, where eta is learning efficiency and the threshold thetaim、wimThe corresponding correction formula obtained after partial derivative calculation is as follows:
wim(t+1)＝wim(t)+Δwimand thetaim(t+1)＝θim(t)+Δθim。
Wherein E is with respect to θim、wimThe formula for calculating the partial derivatives is as follows:
the threshold value theta can be obtained through the partial derivative solving processim、wimThe correction formula of (1):
wim(t+1)＝wim(t)+Δwimand thetaim(t+1)＝θim(t)+Δθim。
S4, reconstructing the output data by adopting wavelet transformation, calculating the reconstruction result and the relative error of the photovoltaic power data at the specified position, setting the wavelet decomposition times based on the relative error, and adjusting the number of hidden layers of the dendritic neuron model.
S5, selecting the designated historical meteorological data and the corresponding photovoltaic power generation power as test data, inputting the dendritic neuron model with the number of the hidden layers adjusted, calculating the output result of the test data according to the set wavelet decomposition times, and taking the output result as the prediction output of the photovoltaic power.
Example two
With reference to fig. 2 and fig. 3, the method of the present invention is described in detail by using a specific embodiment, specifically, decomposing the received air temperature, photovoltaic panel temperature, wind speed, solar irradiance and previous output power data and the photovoltaic output power at the corresponding time by using wavelet transform; in this embodiment, three high-frequency and one low-frequency signals can be obtained from the original photovoltaic power data sequence by using a Mallat multi-resolution analysis method and selecting three-level decomposition; since the decomposition involves filtering and downsampling, the wavelet reconstruction involves three upsampling and filtering steps; the Daubechies type wavelet function selected by the present embodiment is used as the mother wavelet.
Then designing a dendritic neuron model, determining the input number of the prediction model according to the number of meteorological factors, wherein x1, x2, x3, x4 and x5 respectively represent a matrix B obtained by wavelet decomposition of air temperature, photovoltaic panel temperature, wind speed, solar irradiance and output power data at the previous momentiRepresenting the number of hidden layers, in this embodiment, B is specifically usediThe description is given for 5; synapses, branches, cell membranes and cell bodies are respectively referred to by Synaptic, Branch, Membrane and Soma, and then an input vector is substituted into the synapse part of the model to perform an operation, and a specific operation process may refer to step S3 in the first embodiment, which is not described herein again; the process then continues to steps S4-S5 to obtain the final predicted result.
Based on the specific process steps of the method, the advantages of the method over the prior art are illustrated by using the actual data of a certain photovoltaic power plant in Lanzhou, Gansu to perform simulation, specifically, by comparing a prediction model based on a wavelet-radial basis (WT + RBF) neural network with a photovoltaic power prediction model based on a single dendritic Neuron (NBDM) model; the time of the data sample in the training stage is from 6 months to 8 months in 2012, the interval between adjacent data is 15 minutes, and 8000 groups are counted; the time for testing the data samples is in 6 months in 2013, 200 groups of data in total of about 48 hours are selected, and five input variables are air temperature, panel temperature, solar irradiance, air humidity and photovoltaic power at the previous moment. The algorithm development and simulation environment is Visual C + +6.0, the computer memory is 2gb, the AMD CPU is 2.2ghz, and the operating system is Window 7; the graphical curve drawing tool is Matlab2012 a; selecting a model with the highest prediction precision by controlling the same training time in the simulation; referring to fig. 4 in particular, as can be seen from the simulation comparison curve, the curve of the photovoltaic power prediction model based on the wavelet-dendritic neuron model (WT + NBDM) provided by the present invention is closest to the actual photovoltaic power generation curve; by further calculating the relative error, the error based on the wavelet-radial basis (WT + RBF) neural network is 11.63%, the error based on the single dendritic Neuron (NBDM) model is 9.96%, and the photovoltaic power prediction model based on the wavelet-dendritic neuron model (WT + NBDM) of the present invention is 9.15%; compared with the prior art, the method for combining wavelet transformation with the dendritic neuron model can exert the advantages of two algorithms, and effectively improves the photovoltaic power prediction precision.
Compared with the prior art, the photovoltaic power prediction method based on the wavelet transform-dendritic neuron model has the beneficial effects that: before input data enter a neural network, a photovoltaic power prediction model decomposes and reconstructs the input data by utilizing wavelet transformation, can effectively compress and reduce noise of the data, not only simplifies the data amount of training of a dendritic neuron model, but also reduces noise influencing prediction precision in the data; in addition, compared with the traditional neural network model, a single neuron of the dendritic neuron model can perform more complex nonlinear operation, and the number of the neurons is less under the condition of the same precision, so that the convergence rate of the model and the prediction precision can be further improved. Therefore, the model provided by the invention is simpler and has higher convergence speed.
Although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments described in the foregoing detailed description, or equivalent arrangements may be substituted for those skilled in the art. All equivalent structures made by using the contents of the specification and the attached drawings of the invention can be directly or indirectly applied to other related technical fields, and all the equivalent structures are within the protection scope of the invention.
Claims (5)
1. The photovoltaic power prediction method based on the wavelet transform-dendritic neuron model is characterized by comprising the following steps of:
s1, extracting photovoltaic power historical data at the time of the specified position, acquiring historical meteorological data at the corresponding time, acquiring photovoltaic power data corresponding to the historical meteorological data, and forming sequence data by the historical meteorological data and the photovoltaic power data;
s2, decomposing the sequence data by wavelet transformation, obtaining the low frequency component with the appointed number after decomposing the sequence data by the down-sampling of the low-pass filter, and obtaining the high frequency component with the appointed number after decomposing the sequence data by the down-sampling of the high-pass filter;
s3, constructing a dendritic neuron model, and selecting a designated part from the photovoltaic power historical data and the historical meteorological data to form an input signal matrix, wherein the signal matrix is used for training the dendritic neuron model; decomposing the input signal matrix by using wavelet transformation and feeding the decomposed result to the dendritic neuron model for training to obtain output data synchronized with the dendritic neuron model;
s4, reconstructing the output data by adopting wavelet transformation, taking the reconstructed result as photovoltaic power prediction data, calculating the relative error between the reconstructed result and actual photovoltaic power data at a specified position, and adjusting the number of wavelet decomposition times and the number of hidden layers of the dendritic neuron model based on the relative error;
and S5, inputting the remaining photovoltaic power historical data and the historical meteorological data which are divided to form the signal matrix as test data, adjusting the number of the hidden layers of the dendritic neuron model, calculating an output result of the test data according to the set wavelet decomposition times, and outputting the output result as the prediction of the photovoltaic power.
2. The method for photovoltaic power prediction based on wavelet transform-dendritic neuron model according to claim 1, wherein step S1 comprises: and acquiring the historical meteorological data by adopting a correlation analysis method, taking the historical photovoltaic power data and the historical meteorological data at the same moment as a group, and extracting data between adjacent groups according to a time period with a set specified length.
3. The wavelet transform-dendritic neuron model-based photovoltaic power prediction method according to claim 1, wherein the dendritic neuron model comprises four parts, namely synapses, branches, cell membranes and cell bodies, wherein:
x for synapse1,x2,...,xiRepresentation, used as input to said dendritic neuron model and made in said synapse
After operation, transmitting the branch to the appointed branch, wherein m represents the serial number of the branch, and k is a positive parameter; and thetaim、wimFor setting a judgment threshold value when xiWhen the value of (D) is greater than the judgment threshold, the synapse does
Calculating;
the branch is doing at the receiving of the synapse
And performing multiplication operation after the operation result:
i represents the number of input variables;
performing an additive operation of all the branches on the cell membrane:
m represents the number of branches;
the cell body makes the accumulated operation result on the cell membrane according to
Calculating to obtain a final result, wherein thetasomaTo set the threshold.
4. The method for photovoltaic power prediction based on wavelet transform-dendritic neuron model according to claim 3, wherein the dendritic neuron model is a feed-forward multilayer network, and the function of synapses is a function of the number of synapses
Function of the branch
Function of the cell membrane
Function of the cell body
Are all differential functions.
5. The wavelet transform-dendritic neuron model-based photovoltaic power prediction method according to claim 4, wherein the calculating of the connection type between the feedforward multilayer networks by using a back propagation method comprises the following steps:
formula is formed by
Calculating a least squares difference E of actual output and desired output of the dendritic neuron model;
according to the gradient descent learning method, by the equation:
threshold θ for the synapseim、wimMaking a correction to reduce the error, where eta is learning efficiency and the threshold thetaim、wimThe corresponding correction formula obtained after partial derivative calculation is as follows:
wim(t+1)＝wim(t)+Δwim
θim(t+1)＝θim(t)+Δθim。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910051839.0A CN109840629B (en) | 2019-01-21 | 2019-01-21 | Photovoltaic power prediction method based on wavelet transform-dendritic neuron model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910051839.0A CN109840629B (en) | 2019-01-21 | 2019-01-21 | Photovoltaic power prediction method based on wavelet transform-dendritic neuron model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109840629A CN109840629A (en) | 2019-06-04 |
| CN109840629B true CN109840629B (en) | 2022-07-15 |
Family
ID=66883970
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910051839.0A Active CN109840629B (en) | 2019-01-21 | 2019-01-21 | Photovoltaic power prediction method based on wavelet transform-dendritic neuron model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109840629B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114004424A (en) * | 2021-12-21 | 2022-02-01 | 南方电网科学研究院有限责任公司 | Photovoltaic power prediction method, system, equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105139264A (en) * | 2015-06-06 | 2015-12-09 | 安徽工程大学 | Photovoltaic generation capacity prediction method based on particle swarm algorithm wavelet neural network |
| CN108038580A (en) * | 2017-12-30 | 2018-05-15 | 国网江苏省电力公司无锡供电公司 | The multi-model integrated Forecasting Methodology of photovoltaic power based on synchronous extruding wavelet transformation |
| CN108537359A (en) * | 2018-02-28 | 2018-09-14 | 北京信息科技大学 | Photovoltaic power generation power prediction based on wavelet analysis and BP neural network |
-
2019
- 2019-01-21 CN CN201910051839.0A patent/CN109840629B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105139264A (en) * | 2015-06-06 | 2015-12-09 | 安徽工程大学 | Photovoltaic generation capacity prediction method based on particle swarm algorithm wavelet neural network |
| CN108038580A (en) * | 2017-12-30 | 2018-05-15 | 国网江苏省电力公司无锡供电公司 | The multi-model integrated Forecasting Methodology of photovoltaic power based on synchronous extruding wavelet transformation |
| CN108537359A (en) * | 2018-02-28 | 2018-09-14 | 北京信息科技大学 | Photovoltaic power generation power prediction based on wavelet analysis and BP neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109840629A (en) | 2019-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109214575B (en) | Ultrashort-term wind power prediction method based on small-wavelength short-term memory network | |
| Xiong et al. | Short-term wind power forecasting based on Attention Mechanism and Deep Learning | |
| CN110738010A (en) | Wind power plant short-term wind speed prediction method integrated with deep learning model | |
| Duan et al. | A combined short-term wind speed forecasting model based on CNN–RNN and linear regression optimization considering error | |
| Zhang et al. | A novel hybrid model for wind speed prediction based on VMD and neural network considering atmospheric uncertainties | |
| CN104792522B (en) | Intelligent gear defect analysis method based on fractional wavelet transform and BP neutral network | |
| CN111461463B (en) | Short-term load prediction method, system and equipment based on TCN-BP | |
| CN107292446B (en) | Hybrid wind speed prediction method based on component relevance wavelet decomposition | |
| CN104951836A (en) | Posting predication system based on nerual network technique | |
| CN107909209A (en) | The wind speed forecasting method of complete overall experience mode decomposition and depth belief network | |
| CN111582551A (en) | Method and system for predicting short-term wind speed of wind power plant and electronic equipment | |
| CN116316591A (en) | Short-term photovoltaic power prediction method and system based on hybrid bidirectional gating cycle | |
| CN112990553A (en) | Wind power ultra-short-term power prediction method using self-attention mechanism and bilinear fusion | |
| CN114036850A (en) | Runoff prediction method based on VECGM | |
| Wang et al. | Hybrid deep learning model for short-term wind speed forecasting based on time series decomposition and gated recurrent unit | |
| Li et al. | A deep learning framework for non-stationary time series prediction | |
| CN109840629B (en) | Photovoltaic power prediction method based on wavelet transform-dendritic neuron model | |
| CN115511657A (en) | Wind power output and photovoltaic output evaluation method based on combined prediction model | |
| CN113361782B (en) | Photovoltaic power generation power short-term rolling prediction method based on improved MKPLS | |
| Sun et al. | Short-term power load prediction based on VMD-SG-LSTM | |
| Phan et al. | Application of a new Transformer-based model and XGBoost to improve one-day-ahead solar power forecasts | |
| Su et al. | A LSTM based wind power forecasting method considering wind frequency components and the wind turbine states | |
| CN116797204A (en) | Primary air quantity fault early warning method for coal mill based on wavelet decomposition and reconstruction and TCN-GRU-Self-Attention | |
| CN113761777B (en) | HP-OVMD-based ultra-short-term photovoltaic power prediction method | |
| Zhao et al. | A dendritic neuron based very short-term prediction model for photovoltaic power |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |