CN114511747A - Unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN - Google Patents
Unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN Download PDFInfo
- Publication number
- CN114511747A CN114511747A CN202210188027.2A CN202210188027A CN114511747A CN 114511747 A CN114511747 A CN 114511747A CN 202210188027 A CN202210188027 A CN 202210188027A CN 114511747 A CN114511747 A CN 114511747A
- Authority
- CN
- China
- Prior art keywords
- load data
- vae
- preprocessing
- type identification
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000007781 pre-processing Methods 0.000 title claims abstract description 27
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 28
- 238000012549 training Methods 0.000 claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 9
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 6
- 230000006870 function Effects 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 8
- 230000014509 gene expression Effects 0.000 claims description 6
- 238000011176 pooling Methods 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
-
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides an unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN, which comprises the following steps: step 1, collecting load data; step 2, preprocessing the load data; step 3, a variational self-encoder is used for carrying out balancing processing on the load data of a few classes; step 4, converting the load data into a two-dimensional threshold-free recursive graph by using a recursive graph algorithm; and 5, building a two-dimensional convolutional neural network, training and optimizing the two-dimensional convolutional neural network to obtain a load data type identification model, and inputting the recursive graph into the load data type identification model to obtain a classification result of the load data.
Description
Technical Field
The invention belongs to the field of power load classification and identification, and particularly relates to an unbalanced load data type identification method based on VAE preprocessing and RP-2 DCNN.
Background
In recent years, with the rapid development of the intellectualization of the power internet of things, more and more advanced measurement systems are put into operation, so that a large amount of power utilization data of users are accumulated. Valuable potential information is mined and extracted from mass load data, a reasonable and effective load classification algorithm is researched, a personalized power utilization strategy is favorably formulated, and the method has important significance for reasonably regulating and controlling power resources, improving the energy utilization rate and improving the economic benefits of enterprises.
Currently, load classification methods can be classified into unsupervised clustering and supervised classification. The unsupervised clustering is to divide data according to a specified rule under the label of an unknown sample, but the method has the problems of complex parameter adjustment, sensitivity to data and the like. With the rapid increase of the load data volume, unsupervised clustering usually requires a large amount of running time, and wastes part of labeled data, so that the requirement of rapid classification under the background of massive load data is hard to meet.
The supervised classification method can effectively give consideration to both classification speed and classification precision, but is affected by unbalanced data types in the training process, so that the classification result is poor. In addition, most of the existing classifiers are from a sequence perspective, and the deep features of the existing classifiers are difficult to extract, so that the final load identification result is influenced.
Disclosure of Invention
The present invention is made to solve the above problems, and an object of the present invention is to provide an unbalanced load data type identification method based on VAE preprocessing and RP-2 DCNN.
The invention provides an unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN, which is characterized by comprising the following steps: step 1, collecting load data;
and 5, building a two-dimensional convolutional neural network, training and optimizing the two-dimensional convolutional neural network to obtain a load data type identification model, and inputting the recursive graph into the load data type identification model to obtain a classification result of the load data.
The unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN provided by the invention can also have the following characteristics: wherein, step 2 includes the following substeps:
step 2-1, filling missing values existing in the load data by using a multi-order Lagrange interpolation method, wherein the formula is as follows:
step 2-2, the load data is normalized by adopting the maximum and minimum values to eliminate the influence of the load magnitude under each category on the classification result, and the formula is as follows:
in the formula (1), the first and second groups,
correction value, x, representing abnormal load data collected at time t for the ith sample pointi,t-a、xi,t+bRepresenting the sample points taken forward and backward, a, respectively1、 b1The number of samples taken for forward and backward is selected to be 4-6,
in the formula (2), x'iRepresenting normalized load data, xmaxRepresenting the maximum value, x, in the load dataminRepresenting the minimum value in the load data.
The unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN provided by the invention can also have the following characteristics: wherein, the step 3 comprises the following substeps:
step 3-1, building a network structure of a variational self-encoder;
step 3-2, the minority class of load data is used as the input of the variational self-encoder, and the distribution characteristics of the minority class of load data are learned through training the variational self-encoder;
step 3-3, sampling hidden variables from the standard normal distribution N (0,1) and inputting the hidden variables into a generator of a variational self-encoder to generate samples with specified quantity;
and 3-4, repeating the step 3-2 to the step 3-3, merging the generated data of the variational self-encoder and the original data to obtain expanded load data, and finishing the balance processing of a few types of load data.
The unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN provided by the invention can also have the following characteristics: wherein, step 4 comprises the following substeps:
step 4-1, load data xiReconstructing to a two-dimensional space, and the formula is as follows:
Xi＝(xi,xi+τ,…,xi+(m-1)τ) (3)
in equation (3), τ is the delay time, m represents the embedding dimension,
step 4-2, drawing a recursion graph according to the recursion matrix, wherein each element R in the recursion graphijCalculated by the following formula:
Rij＝θ(ε-Eij) (4)
in the formula (4), EijFor reconstructing vector XiAnd XjThe Euclidean distance between the two, epsilon is a threshold function, and theta (x) is a Heaviside function, and the expression is as follows:
the unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN provided by the invention can also have the following characteristics: wherein, the step 5 comprises the following substeps:
step 5-1, dividing the recursion graph into a training set and a test set according to the ratio of 1.5:1, and designing an input layer, a convolutional layer, a pooling layer, a flattening layer and a classification layer of the two-dimensional convolutional neural network;
step 5-2, inputting the training set into a two-dimensional convolutional neural network, performing iterative learning, judging whether the network is converged by calculating cross entropy loss between an output result and a true value of the two-dimensional convolutional neural network, and storing the two-dimensional convolutional neural network as a load data type identification model after optimizing parameters of the convolutional neural network;
and 5-3, inputting the test set into the load data type identification model, and obtaining the classification result of the load data.
Action and Effect of the invention
According to the unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN, the problem of unbalanced class in the load is solved by using a variational self-encoder for the preprocessed load data, the load data is encoded into a two-dimensional gray scale map by using recursive graph calculation, and finally a two-dimensional convolutional neural network is constructed to train to obtain a load data type identification model for carrying out classification identification on the obtained gray scale map.
Drawings
FIG. 1 is a flowchart of a method for identifying unbalanced load data types based on VAE preprocessing and RP-2DCNN according to an embodiment of the present invention;
FIG. 2 is a flow chart of the pre-processing of minority class data by the variational self-encoder in an embodiment of the invention;
FIG. 3 is a graph of seven types of load curve data selected in an embodiment of the present invention;
FIG. 4 is a graph comparing the center curve of a third class load curve generated using VAE and GAN with the original curve in an embodiment of the present invention;
FIG. 5 is a graph comparing the center curve of a fifth class of load curves generated using VAE and GAN with the original curve in an embodiment of the present invention;
FIG. 6 is a graph of recall results for various types of load data in an embodiment of the present invention;
fig. 7 is a non-threshold recursive graph corresponding to the central curve of the seven types of load curve data selected in the embodiment of the present invention.
Detailed Description
In order to make the technical means and functions of the present invention easily understood, the present invention will be specifically described below with reference to the embodiments and the accompanying drawings.
< example >
FIG. 1 is a flowchart illustrating a method for identifying unbalanced load data types based on VAE preprocessing and RP-2DCNN according to an embodiment of the present invention.
As shown in fig. 1, the unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN of the present embodiment includes the following steps: step 1, collecting load data.
And 2, preprocessing the load data.
The step 2 comprises the following substeps:
step 2-1, filling missing values in the load data by using a multi-order Lagrange interpolation method, wherein the formula is as follows:
step 2-2, the load data is normalized by adopting the maximum and minimum values to eliminate the influence of the load magnitude under each category on the classification result, and the formula is as follows:
in the formula (1), the first and second groups,
correction value, x, representing abnormal load data collected at time t for the ith sample pointi,t-a、xi,t+bRepresenting the sample points taken forward and backward, a, respectively1、 b1The number of samples taken for forward and backward is selected to be 4-6,
in the formula (2), x'iRepresenting normalized load data, xmaxRepresenting the maximum value, x, in the load dataminRepresenting the minimum value in the load data.
And 3, balancing the load data of the minority class by using a variational self-encoder (VAE).
FIG. 2 is a flow diagram of the pre-processing of minority class data by a variational auto-encoder in an embodiment of the present invention;
as shown in fig. 2, step 3 includes the following substeps:
step 3-1, building a network structure of a variational self-encoder;
in this embodiment, the structures of the encoder and the decoder in the variational self-encoder need to be symmetrical, taking the encoder as an example, the variational self-encoder includes 3 fully-connected layers, the number of the neurons is 50, 100, and 50, respectively, and a BN layer optimization network is used to accelerate the training speed.
3-2, using the load data of the minority class as the input of a variational self-encoder, and learning the distribution characteristics of the load data of the minority class by training the variational self-encoder;
step 3-3, sampling hidden variables from the standard normal distribution N (0,1) and inputting the hidden variables into a generator of a variational self-encoder to generate samples with specified quantity;
and 3-4, repeating the step 3-2 to the step 3-3, merging the generated data of the variational self-encoder and the original data to obtain expanded load data, and finishing the balance processing of a few types of load data.
And 4, converting the load data into a two-dimensional threshold-free recursive graph by using a recursive graph algorithm.
step 4-1, load data xiReconstructing to a two-dimensional space, and the formula is as follows:
Xi＝(xi,xi+τ,…,xi+(m-1)τ) (3)
in equation (3), τ is the delay time, m represents the embedding dimension,
step 4-2, drawing a recursion graph according to the recursion matrix, wherein each element R in the recursion graphijCalculated by the following formula:
Rij＝θ(ε-Eij) (4)
in the formula (4), EijFor reconstructing vector XiAnd XjThe Euclidean distance between the two, epsilon is a threshold function, and theta (x) is a Heaviside function, and the expression is as follows:
in the embodiment, when the threshold epsilon is not properly selected, a large amount of characteristic information can be lost in the recursion graph, and based on the characteristic information, the daily load curve is converted into the corresponding non-threshold recursion graph.
And 5, building a two-dimensional convolutional neural network, training and optimizing the two-dimensional convolutional neural network to obtain a load data type identification model, and inputting the recursive graph into the load data type identification model to obtain a classification result of the load data.
The step 5 comprises the following substeps:
step 5-1, dividing the recursion graph into a training set and a test set according to the ratio of 1.5:1, and designing an input layer, a convolutional layer, a pooling layer, a flattening layer and a classification layer of a two-dimensional convolutional neural network;
in this embodiment, the number of convolution kernels in the pooling layer is 3, the number of filters is set to 3, the size of the pooling layer is 2, the step length is 1, the filling mode is "same", the activation function is "relu", the classification layer adopts a full-link layer, and a softmax algorithm is adopted to predict the result.
Step 5-2, inputting the training set into a two-dimensional convolutional neural network, performing iterative learning, judging whether the network is converged by calculating cross entropy loss between an output result and a true value of the two-dimensional convolutional neural network, and storing the two-dimensional convolutional neural network as a load data type identification model after optimizing parameters of the convolutional neural network;
and 5-3, inputting the test set into the load data type identification model, and obtaining the classification result of the load data.
In this embodiment, seven types of load curve data are selected for classification testing, and fig. 3 is a data diagram of the seven types of load curve data selected in the embodiment of the present invention.
As shown in fig. 3, in the seven types of load data, the third type and the fifth type are small samples, the number of the small samples is 50 and 25, respectively, and the number of the remaining five types of load data is 100.
FIG. 4 is a graph comparing the center curve of a third class load curve generated using VAE and GAN with the original curve in an embodiment of the present invention; FIG. 5 is a graph comparing the center curve of a fifth class load curve generated using VAE and GAN with the original curve in an embodiment of the present invention.
As shown in fig. 4 and 5, in the case of a small number of samples, the generation curve of the variational self-encoder (VAE) is closer to the original load curve than to the GAN, and even in the case of the fifth type of load curve with large fluctuation, the expanded curve still remains close to the cumulative probability distribution curve of the original curve, which indicates that the performance of the variational self-encoder (VAE) for generating the load curve is better.
FIG. 6 is a chart illustrating recall results of various types of load data in an embodiment of the present invention.
As shown in fig. 6, the recall rate of the GAN oversampled processed data in the fifth category can be slightly increased compared to that in the non-processed data, but the increase is not large, but the data in the third category tends to decrease. And the recall rate of the data after oversampling processing by the variational self-encoder (VAE) is obviously improved in the third class and the fifth class compared with that without processing, and the recall rates of other classes are not reduced, which shows that the problem of class imbalance can be effectively improved by oversampling by the variational self-encoder (VAE), the recall rate of a few classes is improved, and the classification accuracy of other classes is not influenced. In the classification test of the seven selected types of load data, the classification accuracy of the invention (VAE + RP-2DCNN) and the rest classifiers is shown in Table 1.
TABLE 1 Classification accuracy of the present invention with the remaining classifiers
| Accuracy of classification | |
| VAE+XGBOOST | 0.968 |
| VAE+SVM | 0.981 |
| VAE+1DCNN | 0.974 |
| VAE+RP-2DCNN | 0.991 |
Fig. 7 is a non-threshold recursive graph corresponding to the central curve of the seven types of load curve data selected in the embodiment of the present invention.
As shown in fig. 7, after the load curve is encoded by the recursive graph algorithm, it has different feature expressions, and it can be seen by combining table 1 that the accuracy of classification using the two-dimensional convolutional neural network (2DCNN) is higher than that using the sequence input. The reason is that the load curve can effectively enhance the feature expression after being processed by a recursion graph, and then the classification accuracy of the model is improved by utilizing the strong feature extraction capability of a two-dimensional convolutional neural network (2 DCNN).
In summary, according to the unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN, the problem of class imbalance can be effectively improved through variational self-encoder (VAE) oversampling, the recall rate of minority classes is improved, the classification accuracy of other classes is not affected, the load curve can effectively enhance the feature expression after being processed by the recursive graph, and the classification accuracy of the model can be effectively improved through the strong feature extraction capability of the two-dimensional convolutional neural network (2 DCNN).
Effects and effects of the embodiments
According to the unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN, the problem of unbalanced class in the load is solved by using a variational self-encoder for the preprocessed load data, the load data is encoded into a two-dimensional gray scale map by using recursive graph calculation, and finally a two-dimensional convolutional neural network is constructed and trained to obtain a load data type identification model for carrying out classification and identification on the obtained gray scale map.
The above embodiments are preferred examples of the present invention, and are not intended to limit the scope of the present invention.
Claims (5)
1. A VAE preprocessing and RP-2 DCNN-based unbalanced load data type identification method is characterized by comprising the following steps:
step 1, collecting load data;
step 2, preprocessing the load data;
step 3, a variational self-encoder is used for carrying out balancing processing on the load data of a few classes;
step 4, converting the load data into a two-dimensional threshold-free recursive graph by using a recursive graph algorithm;
and 5, building a two-dimensional convolutional neural network, training and optimizing the two-dimensional convolutional neural network to obtain a load data type identification model, and inputting the recursive graph into the load data type identification model to obtain a classification result of the load data.
2. The method for identifying unbalanced load data types based on VAE preprocessing and RP-2DCNN as claimed in claim 1, wherein:
wherein, step 2 includes the following substeps:
step 2-1, filling missing values in the load data by using a multi-order Lagrange interpolation method, wherein the formula is as follows:
step 2-2, carrying out normalization processing on the load data by adopting the maximum and minimum values to eliminate the influence of the load magnitude under each category on the classification result, wherein the formula is as follows:
in the formula (1), the first and second groups,
correction value, x, representing abnormal load data collected at time t for the ith sample pointi,t-a、xi,t+bRepresenting the sample points taken forward and backward, a, respectively1、b1The number of samples taken for forward and backward is selected to be 4-6,
in the formula (2), x'iRepresenting normalized load data, xmaxRepresenting the maximum value, x, in the load dataminRepresenting the minimum value in the load data.
3. The method for identifying unbalanced load data types based on VAE preprocessing and RP-2DCNN as claimed in claim 1, wherein:
wherein, the step 3 comprises the following substeps:
step 3-1, building a network structure of the variational self-encoder;
step 3-2, the load data of the minority class is used as the input of the variation self-encoder, and the distribution characteristics of the load data of the minority class are learned by training the variation self-encoder;
step 3-3, sampling hidden variables from a standard normal distribution N (0,1) and inputting the hidden variables into a generator of the variational self-encoder to generate samples with specified number;
and 3-4, repeating the steps 3-2 to 3-3, merging the generated data of the variational self-encoder and the original data to obtain the expanded load data, and finishing the balance processing of the few types of load data.
4. The method for identifying unbalanced load data types based on VAE preprocessing and RP-2DCNN as claimed in claim 1, wherein:
wherein, step 4 comprises the following substeps:
step 4-1, the load data xiReconstructing to a two-dimensional space, and the formula is as follows:
Xi＝(xi,xi+τ,…,xi+(m-1)τ) (3)
in equation (3), τ is the delay time, m represents the embedding dimension,
step 4-2, drawing the recursion graph according to the recursion matrix, wherein each element R in the recursion graphijCalculated by the following formula:
Rij＝θ(ε-Eij) (4)
in the formula (4), EijFor reconstructing vector XiAnd XjThe Euclidean distance between the two, epsilon is a threshold function, and theta (x) is a Heaviside function, and the expression is as follows:
5. the method for identifying unbalanced load data types based on VAE preprocessing and RP-2DCNN as claimed in claim 1, wherein:
wherein, the step 5 comprises the following substeps:
step 5-1, dividing the recursion diagram into a training set and a test set according to a ratio of 1.5:1, and designing an input layer, a convolutional layer, a pooling layer, a flattening layer and a classification layer of the two-dimensional convolutional neural network;
step 5-2, inputting the training set into the two-dimensional convolutional neural network, performing iterative learning, judging whether the network is converged by calculating cross entropy loss between an output result and a true value of the two-dimensional convolutional neural network, and saving the two-dimensional convolutional neural network as the load data type identification model after optimizing convolutional neural network parameters;
and 5-3, inputting the test set into the load data type identification model, and obtaining the classification result of the load data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210188027.2A CN114511747A (en) | 2022-02-28 | 2022-02-28 | Unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210188027.2A CN114511747A (en) | 2022-02-28 | 2022-02-28 | Unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114511747A true CN114511747A (en) | 2022-05-17 |
Family
ID=81554294
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210188027.2A Pending CN114511747A (en) | 2022-02-28 | 2022-02-28 | Unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114511747A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115171888A (en) * | 2022-08-04 | 2022-10-11 | 云南师范大学 | Method, system, electronic device and storage medium for classifying SMPLC and IM |
-
2022
- 2022-02-28 CN CN202210188027.2A patent/CN114511747A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115171888A (en) * | 2022-08-04 | 2022-10-11 | 云南师范大学 | Method, system, electronic device and storage medium for classifying SMPLC and IM |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111314353B (en) | Network intrusion detection method and system based on hybrid sampling | |
| CN109492026B (en) | Telecommunication fraud classification detection method based on improved active learning technology | |
| CN112037009A (en) | Risk assessment method for consumption credit scene based on random forest algorithm | |
| CN113869208B (en) | Rolling bearing fault diagnosis method based on SA-ACWGAN-GP | |
| CN110147760B (en) | Novel efficient electric energy quality disturbance image feature extraction and identification method | |
| CN110826618A (en) | Personal credit risk assessment method based on random forest | |
| CN111681132B (en) | Typical power consumption mode extraction method suitable for massive class unbalanced load data | |
| CN114019370B (en) | Motor fault detection method based on gray level image and lightweight CNN-SVM model | |
| CN112529638B (en) | Service demand dynamic prediction method and system based on user classification and deep learning | |
| CN112418476A (en) | Ultra-short-term power load prediction method | |
| CN111832839B (en) | Energy consumption prediction method based on sufficient incremental learning | |
| CN110987436A (en) | Bearing fault diagnosis method based on excitation mechanism | |
| CN115204035A (en) | Generator set operation parameter prediction method and device based on multi-scale time sequence data fusion model and storage medium | |
| CN112784031A (en) | Method and system for classifying customer service conversation texts based on small sample learning | |
| CN113361803A (en) | Ultra-short-term photovoltaic power prediction method based on generation countermeasure network | |
| CN114511747A (en) | Unbalanced load data type identification method based on VAE preprocessing and RP-2DCNN | |
| CN112990371B (en) | Unsupervised night image classification method based on feature amplification | |
| CN113239199B (en) | Credit classification method based on multi-party data set | |
| CN117035837B (en) | Method for predicting electricity purchasing demand of power consumer and customizing retail contract | |
| CN117786441A (en) | Multi-scene photovoltaic user electricity consumption behavior analysis method based on improved K-means clustering algorithm | |
| CN116910506A (en) | Load dimension reduction clustering method based on space-time network variation self-encoder algorithm | |
| CN116663745A (en) | LSTM drainage basin water flow prediction method based on PCA_DWT | |
| CN111275109A (en) | Power equipment state data characteristic optimization method and system based on self-encoder | |
| CN111476402A (en) | Wind power generation capacity prediction method coupling meteorological information and EMD technology | |
| CN116303386A (en) | Intelligent interpolation method and system for missing data based on relational graph |
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 |