CN117220287B - Generating capacity prediction method - Google Patents
Generating capacity prediction method Download PDFInfo
- Publication number
- CN117220287B CN117220287B CN202311486355.1A CN202311486355A CN117220287B CN 117220287 B CN117220287 B CN 117220287B CN 202311486355 A CN202311486355 A CN 202311486355A CN 117220287 B CN117220287 B CN 117220287B
- Authority
- CN
- China
- Prior art keywords
- power generation
- data
- door
- generation amount
- input
- 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
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000010248 power generation Methods 0.000 claims abstract description 62
- 230000006870 function Effects 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 6
- 238000005516 engineering process Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims abstract description 4
- 239000011159 matrix material Substances 0.000 claims description 28
- 239000013598 vector Substances 0.000 claims description 16
- 238000012163 sequencing technique Methods 0.000 claims description 12
- 238000011176 pooling Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 208000000044 Amnesia Diseases 0.000 claims description 3
- 208000031091 Amnestic disease Diseases 0.000 claims description 3
- 238000012935 Averaging Methods 0.000 claims description 3
- 230000006986 amnesia Effects 0.000 claims description 3
- 230000006835 compression Effects 0.000 claims description 3
- 238000007906 compression Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims 1
- 238000013527 convolutional neural network Methods 0.000 abstract description 14
- 230000000694 effects Effects 0.000 abstract 1
- 238000010801 machine learning Methods 0.000 description 4
- 230000003213 activating effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a power generation amount prediction method, which comprises the following steps: s1, acquiring original power generation data, and performing dimension reduction on the original power generation data by using a PCA technology to obtain main component data; s2, training a CNN-NLSTM-Attention power generation capacity prediction model by using principal component data to obtain a trained power generation capacity prediction model; s3, predicting power generation data by using a trained power generation amount prediction model to complete power generation amount prediction, decoupling and dimension reduction processing of variable data in the power generation amount prediction process is completed by using PCA, the generalization capability is improved by extracting multi-angle and deep features of a time sequence through a convolutional neural network CNN, the time sequence data is analyzed and calculated by using a long-short-term memory network, and the prediction effect is enhanced by using an Attention mechanism, so that the power generation amount prediction function of an algorithm is realized.
Description
Technical Field
The embodiment of the application relates to the technical field of photovoltaic power generation, in particular to a power generation amount prediction method.
Background
The traditional electric quantity prediction method comprises a time sequence method, a regression analysis method, a gray theory and the like, and the relation between the historical influence factor data and the generated energy is obtained by means of curve fitting, but the processing and the obtaining are difficult due to the large volume of the environmental data.
With the continuous development of intelligent algorithms, machine learning is widely applied to power generation amount prediction, common machine learning comprises support vector machines, decision trees, neural networks, bayesian learning and the like, and the relation between factor data of the environment where a new energy power station is located and power generation amount is obtained through machine learning. The machine learning can solve the nonlinear relation among data, but has certain limitation on the problem of time sequence information, and has the problems that effective information and potential relation contained among discontinuous data cannot be mined, the relevance between load data and related influence factors cannot be fully captured, the extraction of discontinuous data characteristics is not flexible enough and the like.
Disclosure of Invention
Aiming at the defects in the prior art, the generating capacity prediction method provided by the invention reserves a CNN structure in shortcut connection by replacing a main network in a residual block with a double-layer NLSTM network, thereby fully extracting generating capacity time sequence characteristics and ensuring layer jump transmission of useful data in a network layer, and solves the problems that the existing method cannot fully capture the relevance between load data and related influencing factors, the extraction of discontinuous data characteristics is not flexible enough and the like.
In order to achieve the aim of the invention, the invention adopts the following technical scheme: a power generation amount prediction method, comprising the steps of:
s1, acquiring original power generation data, and performing dimension reduction on the original power generation data by using a PCA technology to obtain main component data;
in the embodiment, the data of the Xinjiang Ham wind power station and the photovoltaic station are selected as the original power generation data, and PCA is used for reducing the dimension of multidimensional data such as wind speed, wind direction angle, illumination and the like;
s2, training a CNN-NLSTM-Attention power generation capacity prediction model by using principal component data to obtain a trained power generation capacity prediction model;
and S3, predicting the power generation data by using the trained power generation amount prediction model, and completing the prediction of the power generation amount.
Further: the step S1 comprises the following sub-steps:
s11, normalizing the original power generation data to obtain normalized data;
s12, calculating a covariance matrix of the standardized data;
s13, calculating eigenvalues of the covariance matrix, sequencing, and projecting according to the sequencing result to obtain main component data.
Further: the formula for obtaining the standardized data in the step S11 is as follows:
wherein,for standardized data, ++>For the original power generation data, < > a->And (5) averaging each variable in the original power generation data.
Further: the covariance matrix of the normalized data is calculated in the step S12EThe formula of (2) is:
wherein,is a constant in the matrix, +.>For covariance +.>For variance->、/>Is->、/>Correlation coefficient of individual variables->Is->Correlation coefficients of the individual variables.
Further: the step S13 includes the following sub-steps:
s131, calculating eigenvalues of a covariance matrix, and sequencing the eigenvalues larger than 1, wherein the sequencing method comprises the following steps:
wherein,is the firstkA characteristic value of greater than 1,ka total number of eigenvalues greater than 1;
s132, projecting the standardized data tokAnd obtaining the main component data on the feature vectors corresponding to the feature values.
Further: the trained generating capacity prediction model in the step S2 comprises a CNN module and an NLSTM module which are sequentially connected;
the CNN module comprises an input layer, a first convolution layer, a first pooling layer, a second convolution layer, a second pooling layer, a compression layer and an output layer which are sequentially connected.
Further: the CNN module is used for sharing weight and local connection, reducing complexity of a model and modeling parameters, and the filtering formula of the CNN module on the node value of the identity matrix is as follows:
wherein,g(i) Is the identity matrixiThe value of the individual node(s),and->Is node%x,y,z) Weight value of->Is a bias parameter.
Further: the NLSTM module is used for carrying out normalization processing on the data and dividing and adding labels to the data according to time steps, and the NLSTM model utilizes the last moment to output a stateAnd current time input +.>Calculating amnesia door->Input door->And forget door->The method comprises the steps of carrying out a first treatment on the surface of the And combine with amnestic door->And input door->Refresh memory cell->；
Through forgetting doorControlling to forget the previous moment information;
by controlling the input doorMemory cell retention/>The required information;
through the output doorTransmitting the time status to an output status;
the NLSTM module also comprises a candidate state between the external layer and the internal layerAnd->Through an external input door->And forget door->The updated hidden state parameters and inputs are used as internal hierarchy and enter the internal hierarchy calculation.
Further: the operation formula in the NLSTM module comprises:
wherein,、/>、/>、/>and +.>、/>、/>、/>Corresponding weight matrixes and bias items for the modules; />Candidate cell states at a certain moment; />Is the updated cell state; />Is an input value; />Is a hidden state vector; />Activating a function for Sigmoid->、/>And->Is a weight matrix.
Further: in the process of predicting the generated energy, the attention distribution of the input characteristics is calculated through a score function, and the formula is as follows:
wherein,Yis thatNThe vector is output in a dimension that is,for scoring function, ++>For inquiring the vector +.>For each attention distribution on the query vector, < +.>Scoring the attention function->Representing probability distribution case +.>、/>、/>Is a matrix of learnable parameters.
The beneficial effects of the invention are as follows: first, the power generation amount prediction model uses PCA to influence the power generation amountnMapping of dimensional features tokDimensionally, thiskThe dimension is brand new orthogonal feature also called principal component data, in originalnReconstructed on the basis of dimensional characteristicskDimension characteristics, thereby realizing dimension reduction processing of the multidimensional data of the generated energy influencing factors; and secondly, the generating capacity prediction model is used for solving the effective information and the potential relation which are contained in discontinuous data and can not be mined by combining CNN (convolutional neural network), attention (Attention mechanism) and NLSTM (long and short time memory network), so that the relevance between generating capacity data and various related factors in the generating capacity prediction process is fully mined, and the generating capacity data is flexibly extracted.
Drawings
Fig. 1 is a flowchart of a power generation amount prediction method described in the present application.
Fig. 2 is a diagram of the CNN module structure.
Fig. 3 is a construction diagram of NLSTM modules.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and all the inventions which make use of the inventive concept are protected by the spirit and scope of the present invention as defined and defined in the appended claims to those skilled in the art.
As shown in fig. 1, in one embodiment of the present invention, there is provided a power generation amount prediction method including the steps of:
s1, acquiring original power generation data, and performing dimension reduction on the original power generation data by using a PCA technology to obtain main component data;
s2, training a CNN-NLSTM-Attention power generation capacity prediction model by using principal component data to obtain a trained power generation capacity prediction model;
and S3, predicting the power generation data by using the trained power generation amount prediction model, and completing the prediction of the power generation amount.
In one embodiment of the invention, the step S1 comprises the following sub-steps:
s11, normalizing the original power generation data to obtain normalized data;
s12, calculating a covariance matrix of the standardized data;
s13, calculating eigenvalues of the covariance matrix, sequencing, and projecting according to the sequencing result to obtain main component data.
In one embodiment of the present invention, the formula for obtaining the normalized data in the step S11 is:
wherein,for standardized data, ++>For the original power generation data, < > a->And (5) averaging each variable in the original power generation data.
In one embodiment of the present invention, the step S12 calculates a covariance matrix of the normalized dataEThe formula of (2) is:
wherein,is a constant in the matrix, +.>For covariance +.>For variance->、/>Is->、/>Correlation coefficient of individual variables->Is->Correlation coefficients of the individual variables.
In one embodiment of the present invention, the step S13 includes the following sub-steps:
s131, calculating eigenvalues of a covariance matrix, and sequencing the eigenvalues larger than 1, wherein the sequencing method comprises the following steps:
wherein,is the firstkA characteristic value of greater than 1,ka total number of eigenvalues greater than 1;
in this embodiment, the category with covariance feature value greater than 1 is selected,ktotally 5 kinds;
s132, projecting the standardized data tokAnd obtaining the main component data on the feature vectors corresponding to the feature values.
As shown in fig. 2, in one embodiment of the present invention, the power generation amount prediction model trained in the step S2 includes a CNN module and an NLSTM module that are sequentially connected;
the CNN module comprises an input layer, a first convolution layer, a first pooling layer, a second convolution layer, a second pooling layer, a compression layer and an output layer which are sequentially connected.
In one embodiment of the present invention, the CNN module is configured to share weights and local connections, reduce complexity of a model, and reduce modeling parameters, where a filtering formula of the CNN module on a node value of an identity matrix is:
wherein,g(i) Is the identity matrixiThe value of the individual node(s),and->Is node%x,y,z) Weight value of->Is a bias parameter.
As shown in FIG. 3, in one embodiment of the invention, anThe NLSTM module is used for carrying out normalization processing on the data and dividing and adding labels to the data according to time steps, and the NLSTM model utilizes the last moment to output a stateAnd current time input +.>Calculating amnesia door->Input door->And forget door->The method comprises the steps of carrying out a first treatment on the surface of the And combine with amnestic door->And input door->Refreshing a memory cell；
Through forgetting doorControlling to forget the previous moment information;
by controlling the input doorMemory cell retention->The required information;
through the output doorTransmitting the time status to an output status;
the NLSTM module further comprises a layer between the outer layer and the inner layerCandidate stateAnd->Through an external input door->And forget door->The updated hidden state parameters and inputs are used as internal hierarchy and enter the internal hierarchy calculation.
In one embodiment of the present invention, the operational formula in the NLSTM module includes:
wherein,、/>、/>、/>and +.>、/>、/>、/>Corresponding weight matrixes and bias items for the modules; />Candidate cell states at a certain moment; />Is the updated cell state; />Is an input value; />Is a hidden state vector; />Activating a function for Sigmoid->、/>And->Is a weight matrix.
In one embodiment of the present invention, the attention distribution of the input feature is calculated by a scoring function in the power generation amount prediction process, and the formula is:
wherein,Yis thatNThe vector is output in a dimension that is,for scoring function, ++>For inquiring the vector +.>For each attention distribution on the query vector, < +.>To pay attention toForce scoring function, < ->Representing probability distribution case +.>、/>、/>Is a matrix of learnable parameters.
And calculating the attention distribution of the input features in the power generation amount prediction process through the score function, converting the index function and the score function, and summing under the limitation of the weight coefficient. And finally, selectively extracting information from the input vector according to the obtained attention distribution to obtain and output an attention value, and giving variable weight by applying an attention mechanism to enhance the self-adjusting capacity of the power generation capacity prediction model.
In the description of the present invention, it should be understood that the terms "center," "thickness," "upper," "lower," "horizontal," "top," "bottom," "inner," "outer," "radial," and the like indicate or are based on the orientation or positional relationship shown in the drawings, merely to facilitate description of the present invention and to simplify the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be configured and operated in a particular orientation, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be interpreted as indicating or implying a relative importance or number of technical features indicated. Thus, a feature defined as "first," "second," "third," or the like, may explicitly or implicitly include one or more such feature.
Claims (7)
1. A method of predicting power generation, comprising the steps of:
s1, acquiring original power generation data, and performing dimension reduction on the original power generation data by using a PCA technology to obtain main component data;
the original power generation data are wind power and photovoltaic power generation data;
s2, training a CNN-NLSTM-Attention power generation capacity prediction model by using principal component data to obtain a trained power generation capacity prediction model;
the trained generating capacity prediction model in the step S2 comprises a CNN module and an NLSTM module which are sequentially connected;
the CNN module comprises an input layer, a first convolution layer, a first pooling layer, a second convolution layer, a second pooling layer, a compression layer and an output layer which are sequentially connected;
the NLSTM module is used for carrying out normalization processing on the data and dividing and adding labels to the data according to time steps, and the NLSTM model utilizes the last moment to output a stateAnd current time input +.>Calculating amnesia door->Input door->And an output doorThe method comprises the steps of carrying out a first treatment on the surface of the And combine with amnestic door->And input door->Refresh memory cell->；
Through forgetting doorControlling to forget the previous moment information;
by controlling the input doorMemory cell retention->The required information;
through the output doorTransmitting the time status to an output status;
candidate states are also included between the external hierarchy and the internal hierarchy in the NLSTM moduleAnd->Through an external input door->And an output door->After updating, the hidden state parameters and the input are used as internal hierarchy and enter internal hierarchy calculation;
in the process of predicting the generated energy, the attention distribution of the input characteristics is calculated through a score function, and the formula is as follows:
wherein,Yis thatNThe vector is output in a dimension that is,for scoring function, ++>For inquiring the vector +.>For each attention distribution on the query vector, < +.>Scoring the attention function->Representing probability distribution case +.>、/>、/>Is a matrix of learnable parameters;
and S3, predicting the power generation data by using the trained power generation amount prediction model, and completing the prediction of the power generation amount.
2. The power generation amount prediction method according to claim 1, characterized in that the step S1 includes the sub-steps of:
s11, normalizing the original power generation data to obtain normalized data;
s12, calculating a covariance matrix of the standardized data;
s13, calculating eigenvalues of the covariance matrix, sequencing, and projecting according to the sequencing result to obtain main component data.
3. The method according to claim 2, wherein the formula for obtaining the standardized data in step S11 is:
wherein,for standardized data, ++>For the original power generation data, < > a->And (5) averaging each variable in the original power generation data.
4. The power generation amount prediction method according to claim 2, characterized in that the covariance matrix of the normalized data is calculated in step S12EThe formula of (2) is:
wherein,is a constant in the matrix, +.>For covariance +.>For variance->Is->Correlation coefficients of the individual variables.
5. The power generation amount prediction method according to claim 2, characterized in that the step S13 includes the sub-steps of:
s131, calculating eigenvalues of a covariance matrix, and sequencing the eigenvalues larger than 1, wherein the sequencing method comprises the following steps:
wherein,is the firstkA characteristic value of greater than 1,ka total number of eigenvalues greater than 1;
s132, projecting the standardized data tokAnd obtaining the main component data on the feature vectors corresponding to the feature values.
6. The method for predicting power generation according to claim 1, wherein the CNN module is configured to share weights and local connections, reduce complexity of a model, and reduce modeling parameters, and a filtering formula of the CNN module on a node value of an identity matrix is:
wherein,g(i) Is the identity matrixiThe value of the individual node(s),and->Is node%x,y,z) Weight value of->Is a bias parameter.
7. The power generation amount prediction method according to claim 1, wherein the operation formula in the NLSTM module includes:
wherein,、/>、/>、/>representing a weight matrix in the network,>、/>、/>、/>representing a bias term; />Candidate cell states at a certain moment; />Is the updated cell state; />Is an input value; />Is a hidden state vector; ,/>、/>And->Is a weight matrix; />Representing a calculation forgetting door, a->Representing Sigmoid activation functions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311486355.1A CN117220287B (en) | 2023-11-09 | 2023-11-09 | Generating capacity prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311486355.1A CN117220287B (en) | 2023-11-09 | 2023-11-09 | Generating capacity prediction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117220287A CN117220287A (en) | 2023-12-12 |
| CN117220287B true CN117220287B (en) | 2024-02-06 |
Family
ID=89046708
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311486355.1A Active CN117220287B (en) | 2023-11-09 | 2023-11-09 | Generating capacity prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117220287B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110956312A (en) * | 2019-11-15 | 2020-04-03 | 国网浙江省电力有限公司电力科学研究院 | Photovoltaic power distribution network voltage prediction method based on EMD-CNN deep neural network |
| CN111127146A (en) * | 2019-12-19 | 2020-05-08 | 江西财经大学 | Information recommendation method and system based on convolutional neural network and noise reduction self-encoder |
| CN114970952A (en) * | 2022-04-14 | 2022-08-30 | 楚能新能源股份有限公司 | Photovoltaic output short-term prediction method and system considering environmental factors |
| CN115496263A (en) * | 2022-08-24 | 2022-12-20 | 苏州中立志科技有限公司 | Short-term photovoltaic power generation prediction method based on space-time genetic attention network |
| CN115936254A (en) * | 2022-12-30 | 2023-04-07 | 中国人民解放军海军航空大学 | Intelligent prediction method for key state parameters of aviation equipment |
-
2023
- 2023-11-09 CN CN202311486355.1A patent/CN117220287B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110956312A (en) * | 2019-11-15 | 2020-04-03 | 国网浙江省电力有限公司电力科学研究院 | Photovoltaic power distribution network voltage prediction method based on EMD-CNN deep neural network |
| CN111127146A (en) * | 2019-12-19 | 2020-05-08 | 江西财经大学 | Information recommendation method and system based on convolutional neural network and noise reduction self-encoder |
| CN114970952A (en) * | 2022-04-14 | 2022-08-30 | 楚能新能源股份有限公司 | Photovoltaic output short-term prediction method and system considering environmental factors |
| CN115496263A (en) * | 2022-08-24 | 2022-12-20 | 苏州中立志科技有限公司 | Short-term photovoltaic power generation prediction method based on space-time genetic attention network |
| CN115936254A (en) * | 2022-12-30 | 2023-04-07 | 中国人民解放军海军航空大学 | Intelligent prediction method for key state parameters of aviation equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117220287A (en) | 2023-12-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111832825B (en) | Wind power prediction method and system integrating long-term memory network and extreme learning machine | |
| CN113227808A (en) | Learning device, estimation device, learning method, estimation method, learning program, and estimation program | |
| CN111353631A (en) | Thermal power plant condenser vacuum degree prediction method based on multilayer LSTM | |
| CN108876044B (en) | Online content popularity prediction method based on knowledge-enhanced neural network | |
| CN114219181B (en) | Wind power probability prediction method based on transfer learning | |
| CN111079989A (en) | Water supply company water supply amount prediction device based on DWT-PCA-LSTM | |
| CN113361803A (en) | Ultra-short-term photovoltaic power prediction method based on generation countermeasure network | |
| CN112784920A (en) | Cloud-side-end-coordinated dual-anti-domain self-adaptive fault diagnosis method for rotating part | |
| CN111525587A (en) | Reactive load situation-based power grid reactive voltage control method and system | |
| CN114065653A (en) | Construction method of power load prediction model and power load prediction method | |
| CN111008726A (en) | Class image conversion method in power load prediction | |
| Kathirgamanathan et al. | Performance comparison of single and ensemble CNN, LSTM and traditional ANN models for short‐term electricity load forecasting | |
| Liu et al. | Enhancing Short-Term Power Load Forecasting for Industrial and Commercial Buildings: A Hybrid Approach Using TimeGAN, CNN, and LSTM | |
| CN111680818B (en) | Short-term reactive load prediction method and system | |
| Da et al. | Hourly load prediction based feature selection scheme and hybrid CNN‐LSTM method for building's smart solar microgrid | |
| CN114239945B (en) | Short-term power load prediction method, device, equipment and storage medium | |
| CN117220287B (en) | Generating capacity prediction method | |
| CN118316033A (en) | Power load prediction method based on CNN-GRU and ARIMA model | |
| CN115952928B (en) | Short-term power load prediction method, device, equipment and storage medium | |
| CN117786321A (en) | Method for predicting residual service life of proton exchange membrane fuel cell | |
| CN117113086A (en) | Energy storage unit load prediction method, system, electronic equipment and medium | |
| CN117713238A (en) | Random optimization operation strategy combining photovoltaic power generation and energy storage micro-grid | |
| Jiao et al. | Improved particle filter using clustering similarity of the state trajectory with application to nonlinear estimation: theory, modeling, and applications | |
| CN115081551A (en) | RVM line loss model building method and system based on K-Means clustering and optimization | |
| Tao et al. | Intelligent grid load forecasting based on BERT network model in low-carbon economy |
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 |