CN108520311B - Haze prediction model establishing method and system combining SOFM (software on a programmable) network and BP (back propagation) neural network - Google Patents
Haze prediction model establishing method and system combining SOFM (software on a programmable) network and BP (back propagation) neural network Download PDFInfo
- Publication number
- CN108520311B CN108520311B CN201810185534.4A CN201810185534A CN108520311B CN 108520311 B CN108520311 B CN 108520311B CN 201810185534 A CN201810185534 A CN 201810185534A CN 108520311 B CN108520311 B CN 108520311B
- Authority
- CN
- China
- Prior art keywords
- sofm
- data
- neural network
- training
- aqi
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
The invention discloses a haze prediction model establishing method and system combining a SOFM (software on a screen) network and a BP (back propagation) neural network, wherein SOFM training data are obtained firstly, and then input into the SOFM network for training until the preset learning efficiency is reached; then, respectively acquiring an AQI value of the next time in each mode in each data group; and finally, inputting the data after the SOFM net training and the AQI value of the next time into a BP neural network for training, taking the value of each influence factor corresponding to one mode after the SOFM net training as input during training, and taking the AQI value of the next time in the mode as corresponding output. The haze prediction model established by the invention can accurately predict haze, and the prediction precision is higher compared with the prediction model of the BP neural network.
Description
Technical Field
The invention relates to the field of environment, in particular to a haze prediction model establishing method and system combining a SOFM (software on a screen) network and a BP (back propagation) neural network.
Background
Haze is mainly composed of three terms of sulfur dioxide, nitrogen oxides and inhalable particulate matter, whereas AQI (air quality index) is a dimensionless index that quantitatively describes the air quality condition. Nowadays, AQI data of a meteorological department are released in real time and are not beneficial to arrangement of subsequent early warning measures, so that haze forecasting becomes a problem to be solved urgently in environmental monitoring. However, the difficulty of prediction is not only that the haze prediction involves many factors, and many influencing factors change and spread in real time, but also that huge computing power and accurate computing models are required.
Disclosure of Invention
The invention aims to solve the technical problem that the haze prediction is difficult to complete in the prior art, and provides a haze prediction model establishing method and system combining a SOFM (software defined frequency modulation) network and a BP (back propagation) neural network.
According to one aspect of the present invention, the technical solution adopted by the present invention to solve the technical problem is: a haze prediction model building method combining a SOFM network and a BP neural network is constructed, and the haze prediction model building method comprises the following steps:
s1, obtaining SOFM training data, wherein the SOFM training data comprises a plurality of data subsamples, each data subsample comprises a plurality of patterns, and each pattern comprises values of a plurality of influence factors and AQI values corresponding to the influence factors of the pattern; wherein the modes are divided according to time periods of a day, each mode representing a time period;
s2, inputting data for SOFM training into a SOFM net for training until the preset learning efficiency is reached;
s3, respectively acquiring AQI values of the next time in each mode in each data group;
s4, inputting the data after the SOFM net training and the AQI value of the next time into a BP neural network for training, taking the value of each influence factor corresponding to a mode after the SOFM net training as input during training, and taking the AQI value of the next time in the mode as corresponding output;
the trained model is used for predicting haze; prediction of trained model for each outcome: the actually obtained values of the plurality of influence factors and the corresponding AQI values in one mode are used as input, and the AQI value at the next moment is used as a prediction result.
Further, in the method for establishing the haze prediction model by combining the SOFM network and the BP neural network, the plurality of influencing factors specifically refer to humidity, temperature, wind speed, rainfall, PM10, PM2.5 and SO2、NO2、CO、O3Ten classes of influencing factors.
Further, in the method for establishing the haze prediction model by combining the SOFM network and the BP neural network, the number of nodes in the hidden layer in the BP neural network is determined by the following formula:
wherein m is the number of nodes, n is the number of the plurality of influencing factors, l is the number of output nodes of the BP neural network, and [ x ] represents the maximum positive integer not greater than x.
Further, in the method for establishing the haze prediction model by combining the SOFM network and the BP neural network, the transformation functions adopted in the hidden layer and the output layer in the BP neural network are unipolar Sigmoid functions.
Further, in the method for establishing the haze prediction model by combining the SOFM network and the BP neural network, the method further comprises the following steps:
and acquiring test data, processing the test data by utilizing the steps S1-S4 to obtain a test result of the AQI, comparing the test result with an actual numerical value, and judging the prediction accuracy of the model.
According to another aspect of the present invention, to solve the technical problem, the present invention further provides a haze prediction model establishing system combining a SOFM network and a BP neural network, including the following modules:
the system comprises a SOFM data acquisition module, a data analysis module and a data analysis module, wherein the SOFM data acquisition module is used for acquiring SOFM training data, the SOFM training data comprises a plurality of data subsamples, each data subsample comprises a plurality of patterns, and each pattern comprises values of a plurality of influence factors and AQI values corresponding to the influence factors of the pattern; wherein the modes are divided according to time periods of a day, each mode representing a time period;
the SOFM training module is used for inputting SOFM training data into a SOFM net for training until preset learning efficiency is reached;
a BP data acquisition module, configured to acquire an AQI value at the next time in each mode in each data group;
the model establishing module is used for inputting the data after the SOFM net training and the AQI value at the next time into a BP neural network for training, taking the value of each influence factor corresponding to a mode after the SOFM net training as input during the training, and taking the AQI value at the next time in the mode as corresponding output;
the trained model is used for predicting haze; prediction of trained model for each outcome: the actually obtained values of the plurality of influence factors and the corresponding AQI values in one mode are used as input, and the AQI value at the next moment is used as a prediction result.
Further, in the haze prediction model establishing system combining the SOFM network and the BP neural network, the plurality of influencing factors specifically refer to humidity, temperature, wind speed, rainfall, PM10, PM2.5, and SO2、NO2、CO、O3Ten classes of influencing factors.
Further, in the haze prediction model establishing system combining the SOFM network and the BP neural network of the present invention, the number of nodes in the hidden layer in the BP neural network is determined by the following formula:
wherein m is the number of nodes, n is the number of the plurality of influencing factors, l is the number of output nodes of the BP neural network, and [ x ] represents the maximum positive integer not greater than x.
Further, in the haze prediction model establishing system combining the SOFM network and the BP neural network, the transformation functions adopted in the hidden layer and the output layer in the BP neural network are unipolar Sigmoid functions.
Further, in the haze prediction model establishing system combining the SOFM network and the BP neural network of the present invention, the present invention further comprises the following modules:
and the model verification module is used for acquiring the test data, processing the test data by utilizing the steps S1-S4 to obtain a test result of the AQI, comparing the test result with an actual numerical value and judging the prediction accuracy of the model.
By implementing the method and the system for establishing the haze prediction model by combining the SOFM and the BP neural network, the haze can be accurately predicted by the established haze prediction model, and the prediction precision is higher compared with that of the BP neural network.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a diagram of the combined model topology of the present invention;
FIG. 2 is a flow chart of an embodiment of a haze prediction model building method that combines a SOFM network and a BP neural network;
FIG. 3 is a graph comparing predicted results with true values using a combined model, BP model;
FIG. 4 is a flow diagram of an embodiment of a haze prediction model building system that combines a SOFM net and a BP neural network.
Detailed Description
For a more clear understanding of the technical features, objects and effects of the present invention, embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
The haze prediction model not only needs to consider multiple influence factors and meet the change of data regularity, but also needs to have strong generalization capability. Based on the special attribute of the influence factor of haze weather, the invention provides a self-organizing feature mapping neural network and BP neural network integrated combined prediction model; the self-organizing feature mapping neural network (SOFM network is used below) aims at searching and summarizing potential attributes of haze data and mastering internal rules in historical sample data; the BP neural network forms certain generalization ability by self-learning the attributes and rules of various historical sample data, thereby realizing more accurate prediction ability.
According to the conclusion obtained by analysis of causes and research topics of the dust-haze phenomenon of cities in Wuhan, the invention divides one day into four modes, and performs cluster analysis on other sample data by respectively taking 0:00, 6:00, 12:00 and 18:00 as central data every day, so that the SOFM network selects one-dimensional linear array with definite meaning and simple structure; both the training function and the prediction function are implemented using the overall topology, as shown in FIG. 1.
The embodiment mainly considers humidity, temperature, wind speed, rainfall, PM10, PM2.5 and SO2、NO2、CO、O3The ten types of influence factors also need to use the AQI value at the current time for reference in order to accurately predict the AQI value at the next time, so that the data input number n is equal to 11. The initial weight of the SOFM network determines the clustering accuracy of the samples and the training of the samples, and provides a simple and easy method: t input samples are randomly drawn from the training set as initial values. Since four patterns have been determined, i.e., t is 4, the initial value is extracted from the training samples at times around the four key time points (± 2 hours, i.e., four time periods), so that the initialized weight vectors are close to the pattern classes in the input space from the beginning of training.
The data of the input layer of the BP neural network is various sample data classified by the SOFM network, so n is 11; taking an AQI value of the next time as instructor data, and setting the number of nodes of an output layer to be 1; a method for determining the number of hidden layer nodes comprises the following steps:
(if not an integer, it is not less than
The smallest positive integer of). The transformation functions adopted in the hidden layer and the output layer are unipolar Sigmoid functions
The specific prediction method and steps of the combination method of this embodiment are shown in fig. 2. The haze prediction model building method combining the SOFM network and the BP neural network specifically comprises the following steps:
s1, obtaining SOFM training data, wherein the SOFM training data comprises a plurality of data subsamples, each data subsample comprises a plurality of patterns, and each pattern comprises values of a plurality of influence factors and AQI values corresponding to the influence factors of the pattern; wherein the patterns are divided by time periods of a day, each pattern representing a time period. And when the number of the modes is 4, the influence factors are 10, each mode corresponds to 11 data, and the SOFM trains by taking the 11 data as a data unit to obtain the submodel of the mode.
S2, inputting data for SOFM training into a SOFM net for training until the preset learning efficiency is reached; the initial value α of the learning efficiency in the SOFM net is 0.8, and the end condition is that α is 0.
S3, respectively acquiring AQI values of the next time in each mode in each data group;
and S4, inputting the data after the SOFM net training and the AQI value of the next time into a BP neural network for training, taking the value of each influence factor corresponding to the mode after the SOFM net training as input during training, and taking the AQI value of the next time in the mode as corresponding output. During training, the BP neural network takes 12 data as one data unit for training to obtain the sub-model corresponding to the model.
The trained combined model is used for predicting haze; prediction of each outcome for the trained combined model: the actually obtained values of the plurality of influence factors and the corresponding AQI values in one mode are used as input, and the AQI value at the next moment is used as a prediction result.
The node number of the hidden layer in the BP neural network is determined according to the following formula:
wherein m is the number of nodes, n is the number of the plurality of influencing factors, l is the number of output nodes of the BP neural network, and [ x ] represents the maximum positive integer not greater than x.
After the training is completed, the implementation can also acquire test data, process the test data by using the steps S1-S4 to obtain the test result of the AQI, compare the test result with the actual numerical value, and judge the prediction accuracy of the model.
1 in 2015, 9, 12 in the morning: the prediction results of fine granularity of the test data are shown in the table 1 in 00-7: 00.
TABLE 1 comparison of predicted one hour results
The model built by the method can predict the AQI value of one hour later, and can effectively predict the AQI value of 7 continuous hours later by using historical data of the last 7 hours; the data from 2015/9/120: 00-2015/9/126: 00 is used as historical data to predict the AQI values from 2015/9127: 00-13:00, with the specific prediction results shown in Table 2.
TABLE 2 continuous prediction of 7 hours
In order to show the difference between the combined model of the invention and the traditional BP neural network model, the same parameters are set for the BP neural network model in Matlab to train and predict the same sample data, and the specific comparative analysis result is shown in Table 3.
TABLE 3 comparison of models
Secondly, fig. 3 also highlights that the combined model has superiority in predicting trend compared with the BP model; the BP model prediction result has larger variation difference relative to the true value; and the AQI predicted by the combined model is basically consistent with the change trend of the actual haze weather.
Referring to fig. 4, the haze prediction model establishing system combining the SOFM network and the BP neural network of the present embodiment includes the following modules: SOFM data acquisition module 41, SOFM training module 42, BP data acquisition module 43, and model building module 44. The SOFM data acquiring module 41 is configured to acquire SOFM training data, where the SOFM training data includes a plurality of data subsamples, each data subsample includes a plurality of patterns, and each pattern includes values of a plurality of influencing factors and an AQI value corresponding to the influencing factor of the pattern; wherein the modes are divided according to time periods of a day, each mode representing a time period; the SOFM training module 42 is used for inputting SOFM training data into the SOFM net for training until a preset learning efficiency is reached; the BP data obtaining module 43 is configured to obtain an AQI value at the next time in each mode in each data group; the model establishing module 44 is configured to input the data after the SOFM mesh training and the AQI value at the next time into the BP neural network for training, and during training, take the values of the influencing factors corresponding to the mode after the SOFM mesh training as input, and take the AQI value at the next time in the mode as corresponding output. The trained model is used for predicting haze; prediction of trained model for each outcome: the actually obtained values of the plurality of influence factors and the corresponding AQI values in one mode are used as input, and the AQI value at the next moment is used as a prediction result.
The haze prediction model establishing system combining the SOFM network and the BP neural network corresponds to the method, and the method can be referred to specifically.
While the present invention has been described with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, which are illustrative and not restrictive, and it will be apparent to those skilled in the art that various changes and modifications can be made therein without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (10)
1. A haze prediction model building method combining a SOFM network and a BP neural network is characterized by comprising the following steps:
s1, obtaining SOFM training data, wherein the SOFM training data comprises a plurality of data subsamples, each data subsample comprises a plurality of patterns, and each pattern comprises values of a plurality of influence factors and AQI values corresponding to the influence factors of the pattern; wherein the modes are divided according to time periods of a day, each mode representing a time period;
s2, inputting data for SOFM training into a SOFM net for training until the preset learning efficiency is reached;
s3, respectively acquiring AQI values of the next time in each mode in each data group;
s4, inputting the data after the SOFM net training and the AQI value of the next time into a BP neural network for training, taking the value of each influence factor corresponding to a mode after the SOFM net training as input during training, and taking the AQI value of the next time in the mode as corresponding output;
the trained model is used for predicting haze; prediction of trained model for each outcome: the actually obtained values of the plurality of influence factors and the corresponding AQI values in one mode are used as input, and the AQI value at the next moment is used as a prediction result.
2. The method for establishing the haze prediction model by combining the SOFM network and the BP neural network as claimed in claim 1, wherein the plurality of influence factors specifically refer to humidity, temperature, wind speed, rainfall, PM10, PM2.5, SO2、NO2、CO、O3Ten classes of influencing factors.
3. The method for establishing the haze prediction model by combining the SOFM network and the BP neural network as claimed in claim 1, wherein the number of nodes of the hidden layer in the BP neural network is determined by the following formula:
wherein m is the number of nodes, n is the number of the plurality of influencing factors, l is the number of output nodes of the BP neural network, and [ x ] represents the maximum positive integer not greater than x.
4. The method for establishing the haze prediction model by combining the SOFM network and the BP neural network as claimed in claim 1, wherein the transformation functions adopted in the hidden layer and the output layer in the BP neural network are unipolar Sigmoid functions.
5. The method for establishing the haze prediction model by combining the SOFM net and the BP neural network according to claim 1, further comprising the following steps:
and acquiring test data, processing the test data by utilizing the steps S1-S4 to obtain a test result of the AQI, comparing the test result with an actual numerical value, and judging the prediction accuracy of the model.
6. A haze prediction model establishing system combining a SOFM network and a BP neural network is characterized by comprising the following modules:
the system comprises a SOFM data acquisition module, a data analysis module and a data analysis module, wherein the SOFM data acquisition module is used for acquiring SOFM training data, the SOFM training data comprises a plurality of data subsamples, each data subsample comprises a plurality of patterns, and each pattern comprises values of a plurality of influence factors and AQI values corresponding to the influence factors of the pattern; wherein the modes are divided according to time periods of a day, each mode representing a time period;
the SOFM training module is used for inputting SOFM training data into a SOFM net for training until preset learning efficiency is reached;
a BP data acquisition module, configured to acquire an AQI value at the next time in each mode in each data group;
the model establishing module is used for inputting the data after the SOFM net training and the AQI value at the next time into a BP neural network for training, taking the value of each influence factor corresponding to a mode after the SOFM net training as input during the training, and taking the AQI value at the next time in the mode as corresponding output;
the trained model is used for predicting haze; prediction of trained model for each outcome: the actually obtained values of the plurality of influence factors and the corresponding AQI values in one mode are used as input, and the AQI value at the next moment is used as a prediction result.
7. The system for establishing the haze prediction model by combining the SOFM network and the BP neural network as claimed in claim 6, wherein the plurality of influence factors are humidity, temperature, wind speed, rainfall, PM10, PM2.5, SO2、NO2、CO、O3Ten classes of influencing factors.
8. The system of claim 6, wherein the number of nodes in the hidden layer in the BP neural network is determined by the following formula:
wherein m is the number of nodes, n is the number of the plurality of influencing factors, l is the number of output nodes of the BP neural network, and [ x ] represents the maximum positive integer not greater than x.
9. The system for establishing the haze prediction model by combining the SOFM network and the BP neural network as claimed in claim 6, wherein the transformation functions adopted in the hidden layer and the output layer in the BP neural network are unipolar Sigmoid functions.
10. The system for establishing the haze prediction model by combining the SOFM net and the BP neural network according to claim 6, further comprising the following modules:
and the model verification module is used for acquiring the test data, processing the test data by utilizing the steps S1-S4 to obtain a test result of the AQI, comparing the test result with an actual numerical value and judging the prediction accuracy of the model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810185534.4A CN108520311B (en) | 2018-03-07 | 2018-03-07 | Haze prediction model establishing method and system combining SOFM (software on a programmable) network and BP (back propagation) neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810185534.4A CN108520311B (en) | 2018-03-07 | 2018-03-07 | Haze prediction model establishing method and system combining SOFM (software on a programmable) network and BP (back propagation) neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108520311A CN108520311A (en) | 2018-09-11 |
| CN108520311B true CN108520311B (en) | 2021-05-28 |
Family
ID=63433520
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810185534.4A Active CN108520311B (en) | 2018-03-07 | 2018-03-07 | Haze prediction model establishing method and system combining SOFM (software on a programmable) network and BP (back propagation) neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108520311B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110766219A (en) * | 2019-10-21 | 2020-02-07 | 成都理工大学工程技术学院 | Haze prediction method based on BP neural network |
| CN116739394B (en) * | 2023-08-15 | 2023-10-27 | 中科三清科技有限公司 | Atmospheric pollution weather influence evaluation system and evaluation method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102255965A (en) * | 2011-07-06 | 2011-11-23 | 无锡泛太科技有限公司 | Test platform for multi-sensor fused Internet of things (IOT) based on rough set BP (Back Propagation) neural network |
| CN104112357A (en) * | 2014-07-04 | 2014-10-22 | 浙江大学 | City area traffic emergency plan method for severe haze weather |
| CN104965558A (en) * | 2015-05-27 | 2015-10-07 | 华北电力大学（保定） | Photovoltaic power generation system maximum power tracking method and apparatus considering the factor of haze |
| CN107590549A (en) * | 2016-07-08 | 2018-01-16 | 上海海洋大学 | A kind of Japanese squid winter life group's resource abundance Forecasting Methodology |
-
2018
- 2018-03-07 CN CN201810185534.4A patent/CN108520311B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102255965A (en) * | 2011-07-06 | 2011-11-23 | 无锡泛太科技有限公司 | Test platform for multi-sensor fused Internet of things (IOT) based on rough set BP (Back Propagation) neural network |
| CN104112357A (en) * | 2014-07-04 | 2014-10-22 | 浙江大学 | City area traffic emergency plan method for severe haze weather |
| CN104965558A (en) * | 2015-05-27 | 2015-10-07 | 华北电力大学（保定） | Photovoltaic power generation system maximum power tracking method and apparatus considering the factor of haze |
| CN107590549A (en) * | 2016-07-08 | 2018-01-16 | 上海海洋大学 | A kind of Japanese squid winter life group's resource abundance Forecasting Methodology |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108520311A (en) | 2018-09-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110674604B (en) | Transformer DGA data prediction method based on multi-dimensional time sequence frame convolution LSTM | |
| CN109508360B (en) | Geographical multivariate stream data space-time autocorrelation analysis method based on cellular automaton | |
| CN108491970A (en) | A kind of Predict Model of Air Pollutant Density based on RBF neural | |
| Li et al. | Predicting short-term electricity demand by combining the advantages of arma and xgboost in fog computing environment | |
| CN110381523B (en) | Cellular base station network traffic prediction method based on TVF-EMD-LSTM model | |
| CN113157800B (en) | Identification method for discovering dynamic target in air in real time | |
| Zhang et al. | Short-term power load forecasting using integrated methods based on long short-term memory | |
| CN113205203A (en) | CNN-LSTM-based building energy consumption prediction method and system | |
| CN110956312A (en) | Photovoltaic power distribution network voltage prediction method based on EMD-CNN deep neural network | |
| Tian et al. | A network traffic prediction method based on IFS algorithm optimised LSSVM | |
| CN108520311B (en) | Haze prediction model establishing method and system combining SOFM (software on a programmable) network and BP (back propagation) neural network | |
| CN116681176B (en) | Traffic flow prediction method based on clustering and heterogeneous graph neural network | |
| CN112419711B (en) | Closed parking lot parking demand prediction method based on improved GMDH algorithm | |
| CN114119273A (en) | Park comprehensive energy system non-invasive load decomposition method and system | |
| CN116227716A (en) | Multi-factor energy demand prediction method and system based on Stacking | |
| CN115327041A (en) | Air pollutant concentration prediction method based on correlation analysis | |
| CN106570104A (en) | Multi-partition clustering preprocessing method of stream data | |
| CN111723257A (en) | User portrait drawing method and system based on water usage law | |
| CN111260082A (en) | Spatial object motion trajectory curve prediction method based on neural network | |
| CN104933471A (en) | Phase space reconstruction method for use in short-term wind speed prediction method | |
| Jiang et al. | Short-term pm2. 5 forecasting with a hybrid model based on ensemble gru neural network | |
| CN110866840A (en) | Database modeling method for power load characteristic quantity training based on knowledge graph | |
| CN114881338A (en) | Power distribution network line loss prediction method based on maximum mutual information coefficient and deep learning | |
| CN115310355A (en) | Multi-energy coupling-considered multi-load prediction method and system for comprehensive energy system | |
| Xu et al. | Self-similarity analysis and application of network traffic |
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 |