CN112200384B - EWT neural network-based short-time prediction method for power load - Google Patents
EWT neural network-based short-time prediction method for power load Download PDFInfo
- Publication number
- CN112200384B CN112200384B CN202011169559.9A CN202011169559A CN112200384B CN 112200384 B CN112200384 B CN 112200384B CN 202011169559 A CN202011169559 A CN 202011169559A CN 112200384 B CN112200384 B CN 112200384B
- Authority
- CN
- China
- Prior art keywords
- neural network
- column vector
- power load
- ewt
- load 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 14
- 239000013598 vector Substances 0.000 claims abstract description 76
- 238000003062 neural network model Methods 0.000 claims abstract description 23
- 238000012545 processing Methods 0.000 claims abstract description 22
- 239000011159 matrix material Substances 0.000 claims abstract description 20
- 238000010606 normalization Methods 0.000 claims abstract description 19
- 210000002569 neuron Anatomy 0.000 claims abstract description 14
- 238000012546 transfer Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 4
- 101100391172 Dictyostelium discoideum forA gene Proteins 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 230000008901 benefit Effects 0.000 abstract description 6
- 238000000354 decomposition reaction Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000000700 time series analysis Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- 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
-
- 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
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Business, Economics & Management (AREA)
- Mathematical Optimization (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Pure & Applied Mathematics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Human Resources & Organizations (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Algebra (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Public Health (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Cheminformatics (AREA)
Abstract
The invention discloses a short-time prediction method of electric load based on an EWT (EWT) neural network, which is characterized in that electric load data measured in each hour in a power supply area are formed into a column vector, the column vector is obtained by carrying out standardized processing on the column vector, the column vector is converted into D sub-signal vectors by using EWT, the normalization processing is carried out, each row vector in a new input matrix is taken as input, each row vector in a new output matrix is taken as output at the same time, an Elman neural network model is built, transfer functions of neurons in an intermediate layer are determined, the Elman neural network model is trained by using BP algorithm, the short-time prediction model of the electric load is built, and the electric load data of 168 electric load data in the last seven days are used for predicting the electric load data of one day in the future; the method has the advantages that the prediction of 24 hours of power load data in the future day is realized through the power load data in the first 7 days, and the accuracy of the prediction is improved.
Description
Technical Field
The invention relates to a power load prediction method, in particular to a power load short-time prediction method based on an EWT neural network.
Background
The short-time prediction of the power load has the effect of predicting the power load condition of one day in the future mainly through the power load data of a certain area or region, and the accuracy of the short-time prediction of the power load has important research significance on the safety and economic benefits of a power system. In particular in the capital free market where power competition is intense, the accuracy of short-term forecasting of power loads is critical to optimizing the power system management capability. However, the dynamic change of the power load is very complex, and the power load is affected by various factors such as weather data. Thus, a suitable short-term predictive model is built that takes into account the complex nature of the power load changes.
In recent years, researches on short-term prediction problems of electric loads have been most commonly conducted using techniques widely applied to time series prediction analysis such as autoregressive moving average, linear regression, and generalized exponential smoothing. But these methods have proven incapable of handling the random nonlinear variation characteristics of power load variations. Recently, neural network models have become pets in short-term predictive studies of power load, with the greatest advantage of a strong nonlinear fitting capability. Taking into account that the circuit load changes are differentiated by day, week, month, quarter, etc., a very pronounced periodic characteristic is exhibited, and fourier transforms and wavelet transforms (Wavelet Transformation, abbreviated: WT) have been applied to extract frequency domain change characteristics of the power load. Furthermore, a corresponding power load short-time prediction model is established by combining the WT and the radial basis function neural network model. In addition, empirical mode decomposition (EMPIRICAL MODE DECOMPOSITION, abbreviated EMD) may also be used to signal the electrical load data in order to decompose the varying characteristics of the electrical load data.
Empirical wavelet transform (EMPIRICAL WAVELET transform, abbreviated as EWT) is a very effective time series analysis tool that combines the advantages of EMD with wavelet transform. The EWT can decompose an original signal into a plurality of sub-signal sequences of different frequency bandwidths, and can automatically determine the number of the sub-signal sequences. In addition, given a plurality of sub-signal sequences with different frequency bandwidths, the measurement signal in the time domain can be reconstructed by reverse EWT. Furthermore, in the prior patent and scientific research literature, the Elman neural network has the capability of adapting to time-varying characteristics due to the existence of a feedback loop, and can directly and dynamically reflect the time-sequence characteristics of a process system. Thus, elman neural networks can also make predictions of power load. However, the conventional Elman neural network cannot realize the frequency characteristic decomposition of the power load data, and the accuracy of the power load prediction is still to be questionable.
Disclosure of Invention
The technical problems to be solved by the invention are as follows: how to convert the electric load data sequence into a plurality of different sequences by using the EWT, and on the basis, carrying out the modeling of the short-time prediction based on the Elman neural network in parallel, and realizing the short-time accurate prediction of the electric load by using the reverse EWT.
The technical scheme adopted for solving the technical problems is as follows: a power load short-time prediction method based on an EWT neural network comprises the following steps:
Step 1, forming a column vector x epsilon R N×1 from power load data measured in a power supply area every hour, and performing standardization processing on the power load data according to the following formula to obtain the column vector
In the above formula, μ and δ represent the average value and standard deviation of all elements in the column vector x, respectively, and R N×1 represents an n×1-dimensional real number vector;
Step 2, using EWT to vector columns Is converted into D sub-signal vectors z 1,z2,…,zD, and an input matrix X 1,X2,…,XD and an output matrix Y 1,Y2,…,YD are respectively constructed according to the following formula:
In the above formula, D ε {1,2, …, D }, z d (i) represents the i-th element in the sub-signal vector z d, i ε {1,2, …, N };
Step 3, respectively carrying out normalization processing on X d and Y d to obtain a new input matrix And new output matrix/>
Step 4, using new input matrixEach row vector in (a) is taken as an input, and a new output matrix/>Setting up an Elman neural network model with 168 neurons at an input layer, h neurons at an intermediate layer and 24 neurons at an output layer as output, and determining the transfer function of the neurons at the intermediate layer as f (x) =1/(1+e -x), wherein x represents a function argument;
step 5, training the d-th Elman neural network model by using BP algorithm, reserving an intermediate layer weight coefficient W d and a threshold value b d, connecting a connection weight value V d from the bearing layer to the intermediate layer and a threshold value a d, and outputting a layer weight coefficient And threshold/>
Step 6, repeating the steps 4 to 5 so as to train and finish the 1 st Elman neural network model, the 2 nd Elman neural network model until the D-th Elman neural network model, and reserving corresponding weight coefficients, connection weight values and threshold values;
And 7, establishing a short-term power load prediction model through the steps 1 to 6, and predicting power load data of one day in the future by using 168 pieces of power load data in the last seven days.
Further, the normalization in step 3 is to subtract the minimum value of the column vector from each column vector in X d and divide the column vector by the difference between the maximum value and the minimum value of the column vector.
Further, in step 7, the power load data prediction step for the future day is as follows:
Step (1): the 168 power load data of the latest 7 days are collected and recorded as y 1,y2,…,y168 in sequence according to the time sequence, and the column vector y= [ y 1,y2,…,y168]T is subjected to standardization processing in the same manner in the step 1 to obtain the column vector
Step (2): column vector is set according to EWT in step 2Conversion to D sub-signal vectors/>And then according to the formula/>Respectively constructing input vectors/>Wherein/>Representation/>Element 168 >/>Representation/>D e {1,2, …, D }, and so on;
Step (3): for a pair of After the normalization processing same as that in the step 3 is respectively implemented, new input vectors/>, are correspondingly obtained
Step (4): to be used forAs input vector, the d < th > Elman neural network model after training is utilized to calculate and obtain corresponding output estimated value/>Repeating the steps until an output estimated value/> isobtained
Step (5): respectively toPerforming inverse normalization processing to obtain D estimated signal vectors c 1,c2,…,cD;
Step (6): performing inverse transformation on the D estimated signal vectors c 1,c2,…,cD by using the EWT in the step 2 to obtain 24 time domain data e 1,e2,…,e24, and performing inverse normalization processing on e 1,e2,…,e24 according to the step (1) to obtain predicted data These 24 prediction data/>Respectively representing 24 hours of power load data for a future day.
Further, in the step (5), theThe inverse normalization process is performed as follows: will/>Sequentially multiplying the differences between the maximum and minimum values of the column vectors in X d, and adding the maximum value of the column vector in the corresponding column in X d.
Compared with the prior art, the invention has the advantages that: firstly, when the short-time prediction model of the power load is established, the method firstly utilizes EWT to convert the power load data into a plurality of sub-signals under the frequency domain, and then utilizes a plurality of Elman neural network models to predict the sub-signals. It can be said that the method of the invention fully combines the advantages of EWT and Elman neural networks. Secondly, the method of the invention realizes the prediction of the power load data of 24 hours in the future day through the power load data of the first 7 days, and improves the prediction accuracy.
Drawings
FIG. 1 is a schematic flow chart of the method of the present invention;
FIG. 2 is a graph of the predicted future day power load results of the method of the present invention.
Detailed Description
The invention is described in further detail below with reference to the embodiments of the drawings.
A power load short-time prediction method based on an EWT neural network comprises the following steps:
Step 1, forming a column vector x epsilon R N×1 from power load data measured in a power supply area every hour, and performing standardization processing on the power load data according to the following formula to obtain the column vector
In the above formula, μ and δ represent the average value and standard deviation of all elements in the column vector x, respectively, and R N×1 represents an n×1-dimensional real number vector;
step 2, data processing is carried out through MATLAB tools, and the EWT tools in the MATLAB tools are utilized to carry out column vector processing Is converted into D sub-signal vectors z 1,z2,…,zD, and an input matrix X 1,X2,…,XD and an output matrix Y 1,Y2,…,YD are respectively constructed according to the following formula:
In the above formula, D ε {1,2, …, D }, z d (i) represents the i-th element in the sub-signal vector z d, i ε {1,2, …, N };
Step 3, respectively carrying out normalization processing on X d and Y d to obtain a new input matrix And new output matrix/>The normalization process is to divide each column vector in X d by the difference between the maximum value and the minimum value of the column vector after subtracting the minimum value of the column vector.
Step 4, using new input matrixEach row vector in (a) is taken as an input, and a new output matrix/>An Elman neural network model with 168 neurons at an input layer, h neurons at an intermediate layer and 24 neurons at an output layer is built by taking each row vector of the model as output, and the transfer function of the neurons at the intermediate layer is determined to be f (x) =1/(1+e -x), and the activation function zeta (x) of the neurons at the output layer in the model is a linear function; the x represents a function argument;
Step 5, training the d-th Elman neural network model by using the existing BP algorithm (Back-Propagation), reserving the intermediate layer weight coefficient W d and the threshold value b d, accepting the layer-to-intermediate layer connection weight value V d and the threshold value a d, and outputting the layer weight coefficient And threshold/>
Step 6, repeating the steps 4 to 5 so as to train and finish the 1 st Elman neural network model, the 2 nd Elman neural network model until the D-th Elman neural network model, and reserving corresponding weight coefficients, connection weight values and threshold values;
And 7, establishing a short-term power load prediction model through the steps 1 to 6, and predicting power load data of one day in the future by using 168 pieces of power load data in the last seven days.
The specific steps of predicting the power load data of the future day in the step 7 are as follows:
Step (1): the 168 power load data of the latest 7 days are collected and recorded as y 1,y2,…,y168 in sequence according to the time sequence, and the column vector y= [ y 1,y2,…,y168]T is subjected to standardization processing in the same manner in the step 1 to obtain the column vector
Step (2): column vector is set according to EWT in step 2Conversion to D sub-signal vectors/>And then according to the formula/>Respectively constructing input vectors/>Wherein/>Representation/>Element 168 >/>Representation/>D e {1,2, …, D }, and so on;
Step (3): for a pair of After the normalization processing same as that in the step 3 is respectively implemented, new input vectors/>, are correspondingly obtained
Step (4): to be used forAs input vector, the d < th > Elman neural network model after training is utilized to calculate and obtain corresponding output estimated value/>Repeating the steps until an output estimated value/> isobtained
Step (5): respectively toPerforming inverse normalization processing to obtain D estimated signal vectors c 1,c2,…,cD; wherein pair/>The inverse normalization process is performed as follows: will/>Sequentially multiplying the differences between the maximum and minimum values of the column vectors in X d, and adding the maximum value of the column vector in the corresponding column in X d.
Step (6): performing inverse transformation on the D estimated signal vectors c 1,c2,…,cD by using the EWT tool in the step 2 to obtain 24 time domain data e 1,e2,…,e24, and performing inverse normalization processing on e 1,e2,…,e24 according to the step (1) to obtain predicted dataThese 24 prediction data/>And respectively representing 24 hours of power load data of a future day, thereby completing prediction of the power load of the future day.
After the power load data of the future day is acquired in real time, the actual power load data and the predicted data are plotted in fig. 2. As can be seen from comparison of FIG. 2, the power load data predicted by the method of the invention has higher fitting degree with the actual power load data, and the feasibility and reliability of the method of the invention are verified.
The foregoing embodiments are merely preferred embodiments of the present invention, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it is possible for a person skilled in the art to make several variants and modifications without departing from the inventive concept, which fall within the scope of protection of the present invention.
Claims (4)
1. The electric load short-time prediction method based on the EWT neural network is characterized by comprising the following steps of:
Step 1, forming a column vector x epsilon R N×1 from power load data measured in a power supply area every hour, and performing standardization processing on the power load data according to the following formula to obtain the column vector
In the above formula, μ and δ represent the average value and standard deviation of all elements in the column vector x, respectively, and R N×1 represents an n×1-dimensional real number vector;
Step 2, using EWT to vector columns Is converted into D sub-signal vectors z 1,z2,…,zD, and an input matrix X 1,X2,…,XD and an output matrix Y 1,Y2,…,YD are respectively constructed according to the following formula:
In the above formula, D ε {1,2, …, D }, z d (i) represents the i-th element in the sub-signal vector z d, i ε {1,2, …, N };
Step 3, respectively carrying out normalization processing on X d and Y d to obtain a new input matrix And new output matrix/>
Step 4, using new input matrixEach row vector in (a) is taken as an input, and a new output matrix/>Setting up an Elman neural network model with 168 neurons at an input layer, h neurons at an intermediate layer and 24 neurons at an output layer as output, and determining the transfer function of the neurons at the intermediate layer as f (x) =1/(1+e -x), wherein x represents a function argument;
step 5, training the d-th Elman neural network model by using BP algorithm, reserving an intermediate layer weight coefficient W d and a threshold value b d, connecting a connection weight value V d from the bearing layer to the intermediate layer and a threshold value a d, and outputting a layer weight coefficient And threshold/>
Step 6, repeating the steps 4 to 5 so as to train and finish the 1 st Elman neural network model, the 2 nd Elman neural network model until the D-th Elman neural network model, and reserving corresponding weight coefficients, connection weight values and threshold values;
And 7, establishing a short-term power load prediction model through the steps 1 to 6, and predicting power load data of one day in the future by using 168 pieces of power load data in the last seven days.
2. The method of claim 1, wherein the normalization in step 3 is to subtract the minimum value of each column vector in X d from the maximum value of the column vector and then divide the column vector by the difference between the maximum value and the minimum value of the column vector.
3. The short-term prediction method of electric load based on EWT neural network according to claim 1, wherein the step of predicting the electric load data of a future day in step 7 is:
Step (1): the 168 power load data of the latest 7 days are collected and recorded as y 1,y2,…,y168 in sequence according to the time sequence, and the column vector y= [ y 1,y2,…,y168]T is subjected to standardization processing in the same manner in the step 1 to obtain the column vector
Step (2): column vector is set according to EWT in step 2Conversion to D sub-signal vectors/>And then according to the formulaRespectively constructing input vectors/>Wherein/>Representation/>Element 168 >/>Representation/>D e {1,2, …, D }, and so on;
Step (3): for a pair of After the normalization processing same as that in the step 3 is respectively implemented, new input vectors/>, are correspondingly obtained
Step (4): to be used forAs input vector, the d < th > Elman neural network model after training is utilized to calculate and obtain corresponding output estimated value/>Repeating the steps until an output estimated value/> isobtained
Step (5): respectively toPerforming inverse normalization processing to obtain D estimated signal vectors c 1,c2,…,cD;
Step (6): performing inverse transformation on the D estimated signal vectors c 1,c2,…,cD by using the EWT in the step 2 to obtain 24 time domain data e 1,e2,…,e24, and performing inverse normalization processing on e 1,e2,…,e24 according to the step (1) to obtain predicted data These 24 prediction data/>Respectively representing 24 hours of power load data for a future day.
4. The method for short-term prediction of electrical load based on EWT neural network according to claim 3, wherein step (5) is performed onThe inverse normalization process is performed as follows: will/>Sequentially multiplying the differences between the maximum and minimum values of the column vectors in X d, and adding the maximum value of the column vector in the corresponding column in X d.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011169559.9A CN112200384B (en) | 2020-10-28 | 2020-10-28 | EWT neural network-based short-time prediction method for power load |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011169559.9A CN112200384B (en) | 2020-10-28 | 2020-10-28 | EWT neural network-based short-time prediction method for power load |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112200384A CN112200384A (en) | 2021-01-08 |
CN112200384B true CN112200384B (en) | 2024-05-17 |
Family
ID=74011753
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011169559.9A Active CN112200384B (en) | 2020-10-28 | 2020-10-28 | EWT neural network-based short-time prediction method for power load |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112200384B (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104636822A (en) * | 2015-01-21 | 2015-05-20 | 广州市香港科大霍英东研究院 | Residential load prediction method of elman-based neural network |
CN107491412A (en) * | 2017-07-10 | 2017-12-19 | 华北电力大学 | A kind of user power utilization load characteristic extracting method based on experience wavelet transformation |
CN110490369A (en) * | 2019-07-25 | 2019-11-22 | 国网安徽省电力有限公司 | A kind of Short-Term Load Forecasting Method based on EWT and LSSVM model |
CN111697621A (en) * | 2020-06-23 | 2020-09-22 | 石家庄铁道大学 | Short-term wind power prediction method based on EWT-PDBN combination |
-
2020
- 2020-10-28 CN CN202011169559.9A patent/CN112200384B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104636822A (en) * | 2015-01-21 | 2015-05-20 | 广州市香港科大霍英东研究院 | Residential load prediction method of elman-based neural network |
CN107491412A (en) * | 2017-07-10 | 2017-12-19 | 华北电力大学 | A kind of user power utilization load characteristic extracting method based on experience wavelet transformation |
CN110490369A (en) * | 2019-07-25 | 2019-11-22 | 国网安徽省电力有限公司 | A kind of Short-Term Load Forecasting Method based on EWT and LSSVM model |
CN111697621A (en) * | 2020-06-23 | 2020-09-22 | 石家庄铁道大学 | Short-term wind power prediction method based on EWT-PDBN combination |
Non-Patent Citations (4)
Title |
---|
Teng,Z等.Application of empirical wavelet transform in power load forecasting.Chinese Automation Congress (CAC).2018,4009-4013. * |
一种基于小波分析和Elman动态神经网络的中长期电力负荷预测方法;赵智勇等;《山西电力》(第01期);1-5 * |
含高渗透率分布式电源的节点负荷预测方法;林梅芬等;《河南工程学院学报(自然科学版)》;第32卷(第03期);39-44 * |
基于改进快速集合经验模态分解和 Elman-Adaboost 的短期风速预测方法;王凯等;《电力科学与工程》;第36卷(第5期);32-39 * |
Also Published As
Publication number | Publication date |
---|---|
CN112200384A (en) | 2021-01-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Zhang et al. | Short term electricity load forecasting using a hybrid model | |
CN108256697B (en) | Prediction method for short-term load of power system | |
Nazaripouya et al. | Univariate time series prediction of solar power using a hybrid wavelet-ARMA-NARX prediction method | |
Reis et al. | Feature extraction via multiresolution analysis for short-term load forecasting | |
Xiao et al. | BP neural network with rough set for short term load forecasting | |
Ji et al. | Prediction of hourly solar radiation using a novel hybrid model of ARMA and TDNN | |
CN110738010A (en) | Wind power plant short-term wind speed prediction method integrated with deep learning model | |
CN111553543B (en) | TPA-Seq2 Seq-based power load prediction method and related components | |
CN111079989B (en) | DWT-PCA-LSTM-based water supply amount prediction device for water supply company | |
CN111105321A (en) | Power load prediction method based on EEMD secondary decomposition | |
Anaghi et al. | A model for stock price forecasting based on ARMA systems | |
CN114358389A (en) | Short-term power load prediction method combining VMD decomposition and time convolution network | |
Pindoriya et al. | Forecasting of short-term electric load using application of wavelets with feed-forward neural networks | |
CN113449904A (en) | Multi-energy load prediction method, device and equipment | |
Sfetsos et al. | Time series forecasting of averaged data with efficient use of information | |
CN112200384B (en) | EWT neural network-based short-time prediction method for power load | |
Mourad et al. | A hybrid wavelet transform and ANFIS model for short term electric load prediction | |
CN116432860A (en) | Short-term power load prediction method, system and medium based on EWT-TCN | |
CN105243436A (en) | Wind power real-time prediction method based on rank set pair analysis | |
Xu et al. | Application of GMDH to short-term load forecasting | |
CN114925940A (en) | Holiday load prediction method and system based on load decomposition | |
Xia et al. | Research on short-term load forecasting model based on wavelet decomposition and neural network | |
Chaparro et al. | A method for the monthly electricity demand forecasting in Colombia based on wavelet analysis and a nonlinear autoregressive model | |
Xu et al. | A combination model based on EEMD-PE and echo state network for chaotic time series prediction | |
CN114169416A (en) | Short-term load prediction method under small sample set based on transfer learning |
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 |