CN111258984B - End-edge-cloud collaborative forecasting method for product quality in industrial big data environment - Google Patents
End-edge-cloud collaborative forecasting method for product quality in industrial big data environment Download PDFInfo
- Publication number
- CN111258984B CN111258984B CN202010051048.0A CN202010051048A CN111258984B CN 111258984 B CN111258984 B CN 111258984B CN 202010051048 A CN202010051048 A CN 202010051048A CN 111258984 B CN111258984 B CN 111258984B
- Authority
- CN
- China
- Prior art keywords
- data
- model
- forecast
- edge
- database
- 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
- 238000013277 forecasting method Methods 0.000 title claims abstract description 12
- 238000004519 manufacturing process Methods 0.000 claims abstract description 58
- 238000000034 method Methods 0.000 claims abstract description 26
- 238000012549 training Methods 0.000 claims description 27
- 238000004422 calculation algorithm Methods 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 8
- 238000012937 correction Methods 0.000 claims description 7
- 238000009776 industrial production Methods 0.000 claims description 7
- 238000004140 cleaning Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000010923 batch production Methods 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims 1
- 238000012545 processing Methods 0.000 abstract description 3
- 230000008901 benefit Effects 0.000 abstract description 2
- 229910000831 Steel Inorganic materials 0.000 description 15
- 239000010959 steel Substances 0.000 description 15
- 230000008859 change Effects 0.000 description 13
- XEEYBQQBJWHFJM-UHFFFAOYSA-N iron Substances [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 4
- 238000010438 heat treatment Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 239000000498 cooling water Substances 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/211—Schema design and management
- G06F16/212—Schema design and management with details for data modelling support
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- General Factory Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
本发明提供一种工业大数据环境下的产品质量端‑边‑云协同预报方法，涉及工业大数据处理与复杂工业智能建模技术领域。本方法是在云端服务器上利用工业大数据训练预报模型，同时在边缘端服务器以及终端服务器上不断地校正预报模型中的相关参数，使预报模型的结果更加精确，同时在终端服务器上对产品质量进行实时预报。本发明能够有效的利用生产过程中产生的实时数据，不断的修正预报模型中的参数，使得预报模型能够适应产品的实时变化，进而不断提高模型的预报精度，提高生产效益。
The invention provides an end-edge-cloud collaborative forecasting method for product quality in an industrial big data environment, and relates to the technical fields of industrial big data processing and complex industrial intelligent modeling. This method uses industrial big data to train the forecast model on the cloud server, and at the same time continuously corrects the relevant parameters in the forecast model on the edge server and the terminal server, so that the results of the forecast model are more accurate, and at the same time, the product quality is checked on the terminal server. Make real-time forecasts. The present invention can effectively utilize the real-time data generated in the production process to continuously correct the parameters in the forecasting model, so that the forecasting model can adapt to the real-time changes of the product, thereby continuously improving the forecasting accuracy of the model and improving the production benefit.
Description
Technical Field
The invention relates to the technical field of industrial big data processing and complex industrial intelligent modeling, in particular to a product quality end-edge-cloud collaborative forecasting method in an industrial big data environment.
Background
In recent years, as artificial intelligence develops more and more mature in theory and technology, application of large data is more and more extensive, and relatively mature results are obtained in the fields of medicine, electronic information, image recognition and the like. In the field of complex industrial intelligent modeling, the application of industrial big data is very important, and the quality of a product is an important index for describing whether an industrial production process is qualified or not.
Although the existing intelligent modeling algorithm can effectively process high-dimensional data in industrial big data and automatically mine potential characteristics hidden in production process data, most of the traditional intelligent modeling algorithm is mainly used for processing static data sets and is difficult to apply to a real-time system, and the established intelligent forecasting model can only reflect rules hidden in historical data and cannot be corrected along with tiny changes in the production process.
In an actual industrial field, production data of products can be continuously increased along with the production, and if sample data generated in real time can be effectively utilized along with the production process, and tiny changes generated by the data in the production process are excavated, a forecasting model can be continuously improved, and further the model precision is improved. However, the traditional intelligent modeling method needs a large amount of training sample data in each model training process, and the training speed is slow, so that the model cannot be updated in real time. Therefore, with the advance of the production process, how to effectively apply the data samples generated in real time in the production process, discover the slight change generated by the samples in the production process, and simultaneously save the computing resources and time is the problem to be solved at present.
Disclosure of Invention
The invention aims to solve the technical problem of providing a product quality end-edge-cloud collaborative forecasting method under an industrial big data environment aiming at the defects of the prior art.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows:
the invention provides a product quality end-edge-cloud collaborative forecasting method under an industrial big data environment, which comprises the following steps:
step 1: acquiring actual production process data of a product in an actual industrial field by using a sensor in the actual industrial field;
step 2: removing abnormal data samples and data samples containing missing values in the collected production process data by using a data cleaning algorithm to form an initial sample data set; data preprocessing is carried out on data in the initial sample data set by using a data complementing algorithm, all data dimensions are the same, and the preprocessed sample data is stored in an edge end database; establishing a cloud database on a cloud server, synchronizing sample data in the edge database into the cloud database when the number of samples in the edge database is more than n, and emptying data samples in the edge database;
and step 3: judging whether the total number of data in the cloud database is more than H, if not, executing the step 1, if so, selecting an intelligent modeling method aiming at the characteristics of the production process and the production process data of the product on a cloud server, and establishing a forecasting model of the product quality;
respectively establishing w forecasting models on a cloud server according to w quality indexes of the product to form a model library; the method comprises the following steps of establishing a forecasting model aiming at the ith quality index as follows:
wherein I represents the preprocessed sample data input by the forecasting model,
a prediction value f representing the i-th quality indexi(. The) represents the structure of the established prediction model, θiA set of parameters representing the established prediction model;
according to the industrial production process, the data characteristics of the model input data and the analysis of the correlation between the input data of the forecasting model and the quality index, the theta is further determinediDivided into three sets of parameters, i.e.
And 4, step 4: according to the actual production sequence of the product, extracting the latest K sample data from the cloud database to form a training set D, and simultaneously recording that the total number of the data samples in the cloud database is S; respectively training all parameters in each forecast model in the model library by using sample data in the training set D, and recording the trained forecast model library as
Wherein Fi cA forecasting model representing the ith quality index;
taking the production process data of the sample as input data, taking the ith quality index data of the sample as tag data, and wearing the sample in a cloudOn the server, training the prediction model of the ith quality index to obtain Fi c(ii) a I.e. training the parameter set in step 3
All of the parameters in (1);
and 5: forecast model library FcThe method comprises the steps that the cloud server transmits the prediction model to an edge server, the edge server puts different prediction models to different terminal servers for operation, and a user predicts different quality indexes of a product through the prediction models in the terminal servers;
step 6: acquiring input data of a forecasting model after data cleaning and data preprocessing are carried out on actual production process data acquired from a sensor in an industrial field, transmitting the input data to all terminal servers, forecasting each quality index of a product on each terminal server by using the forecasting model respectively, and transmitting a forecasting result to a user;
and 7: when the production process of the product is finished, maintaining the corresponding parameter set of the forecasting model on each terminal server
Wherein i ∈ {1,2, …, w }, and using actual production process data of the product to perform parameter set on each forecasting model
The parameters in the model are corrected in real time to obtain a new forecasting model Fi tAnd replacing the original forecast model in the terminal server at the moment, using forecast model Fi tForecasting the subsequent products; the real-time correction is to adopt different correction methods to the parameter set according to different industrial fields and modeling algorithms
Correcting the parameters in (1);
and 8: storing the actual production data and the quality index data of the product into a historical database of the edge terminal; judging the number of all samples in the historical database of the edge end at the moment, if the number of the samples at the moment is less than n, turning to the step 6, and continuing to forecast the quality index of the subsequent product; if the number of the samples is more than n, turning to step 9;
and step 9: extracting the production data of n products from the historical database of the edge end as a new training set d, and aiming at each forecasting model in the forecasting model base, on the edge end server, utilizing the sample data in the training set d to carry out parameter set treatment on the model
Correcting in real time to obtain a prediction model Fi eAll the corrected forecasting models are combined into a new forecasting model library
Step 10: using edge server, FeThe forecasting models in the system are respectively downloaded to corresponding terminal servers and replace the original forecasting models; calling the retrained forecasting model by a user through different terminal servers, and forecasting the product data in a new round; synchronizing the data samples in the edge database to the cloud database, emptying data information in the edge database, and storing product data of a new round of production into the edge database;
step 11: judging the number of samples in the cloud database at the moment, judging whether the total number of the samples in the cloud database is increased by N samples compared with S or not, wherein N is larger than N, if yes, returning to the step 4, counting the total number S of the samples in the cloud database again to be S + N, and retraining FcThe predictive model of (1); if not, returning to the step 6, and forecasting the product quality by using the forecasting model on the terminal server.
Said step 3
Parameter set
The method is used for describing the change rule of the data samples produced in a large batch; parameter set
The data characteristic of the data sample is changed in different small batch production processes; parameter set
Is used for describing the specific data characteristics included in each data sample; wherein the number of the products produced in the industrial production process of the large batch is M, each large batch is divided into r small batches, and the number of the products produced in each small batch is M.
Adopt the produced beneficial effect of above-mentioned technical scheme to lie in: according to the product quality end-edge-cloud collaborative forecasting method under the industrial big data environment, the parameters of the established forecasting model can be divided into parameters sensitive to the change of a large number of data samples; parameters sensitive to changes in a small number of data samples; and parameters sensitive to the change of a single data sample are classified into three types, and the parameters in the forecasting model are continuously trained and updated by an end-edge-cloud collaborative forecasting method. Meanwhile, the invention can effectively utilize real-time data generated in the production process and continuously correct parameters in the forecasting model, so that the forecasting model can adapt to the real-time change of the product, thereby continuously improving the forecasting precision of the model and improving the production benefit.
Drawings
Fig. 1 is a structural block diagram of a product quality end-edge-cloud collaborative prediction method in an industrial big data environment according to an embodiment of the present invention;
fig. 2 is a flowchart of a product quality end-edge-cloud collaborative forecasting method in an industrial big data environment according to an embodiment of the present invention.
Detailed Description
The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
As shown in fig. 2, the method of the present embodiment is as follows.
The invention provides a product quality end-edge-cloud collaborative forecasting method under an industrial big data environment, which is characterized in that a forecasting model is trained on a cloud server by utilizing industrial big data, and related parameters in the forecasting model are continuously corrected on an edge end server and a terminal server at the same time, as shown in figure 1, so that the result of the forecasting model is more accurate. The method for intelligently forecasting the quality of the steel plate product in the embodiment comprises the following steps:
step 1: acquiring actual production process data of a steel plate product in an industrial field by using a sensor in the industrial field of the steel plate;
step 2: removing abnormal data samples and data samples containing missing values in the collected production process data by using a data cleaning algorithm to form an initial sample data set; then, data preprocessing is carried out on data in the initial sample data set by using a data complementing algorithm according to the difference of the sample data dimensions caused by different types of the steel plates to obtain sample data with the same data dimensions, and the preprocessed sample data is stored in an edge end database; the preprocessed sample data is used as input data of a product quality forecasting model, the final quality label data of the product is used as output data of the product quality forecasting model, the model is used for constructing the product quality forecasting model, useless data are added around the input data, the final forecasting result is not affected, and meanwhile the input data dimensions of all samples are unified. And a cloud database is established on the cloud server and used for storing all the preprocessed sample data, so that the data in the industrial production process can be analyzed and modeled conveniently. When the number of the samples in the edge end database is more than n, synchronizing the sample data in the edge end database to the cloud end database, and emptying the data samples in the edge end database; and the edge end database on the edge end server is used for storing sample data generated in the small-batch production process.
And step 3: judging whether the total number of data in the cloud database is greater than H, if not, executing the step 1, and if so, selecting a proper intelligent modeling method, such as a convolutional neural network, a graph neural network, a random forest and other intelligent modeling methods, aiming at the characteristics of the steel plate production process and the steel plate production process data on a cloud server; in the embodiment, a model building method of mechanism and data is selected to build a forecasting model of product quality;
as 5 quality evaluation indexes are provided for describing whether the steel plate product is qualified or not, the indexes are respectively the size, the surface, the shape, the internal quality and the performance of the steel plate. Therefore, on the cloud server, a forecasting model needs to be established respectively for each quality index to form a model library.
The method comprises the following steps of establishing a forecasting model aiming at the ith quality index as follows:
wherein I represents the preprocessed sample data input by the forecasting model,
a prediction value representing the i-th quality index given by the prediction model, fi(. The) represents the structure of the established prediction model, θiA set of parameters representing the established prediction model;
according to the production process of the steel plate, the data characteristics of the model input data and the analysis of the correlation between the input data of the forecasting model and the quality index, the theta is further determinediDivided into three sets of parameters, i.e.
In the industrial production process, the number of the products produced in a large batch is M,dividing each large batch into r small batches, wherein the number of the products produced in each small batch is m; the parameter set
The method is used for describing the change rule of the data samples produced in a large batch; parameter set
The data characteristic of the data sample is changed in different small batch production processes; parameter set
Is used for describing the specific data characteristics included in each data sample;
the analysis method in the embodiment may be that a typical correlation analysis algorithm has mic, pearson correlation coefficient, and the like;
parameter set
The parameters in (1) are mainly used for describing the change rule of input data of products produced in a large batch, and are used for mining the characteristics implied in a large data sample, and are the main parameters in the forecasting model. The input data corresponding to the partial parameters are often too complex, more data samples and complex model structures are needed as supports to dig out the implicit characteristics, and the partial parameters are not obviously changed along with the passage of time in the production process for a long time, such as the temperature change rule of the steel plate in the production process, the equipment information of the steel plate in the production process, and the like.
Parameter set
The parameters in (1) are mainly used for describing slight changes of data characteristics of input data in different small batch production processes. The part of the parameters are not changed in the production process of a large batch, but are changed along with the part of the parametersThe partial parameters are usually suitable for the data samples in a small batch, but not necessarily suitable for the data samples in other batches. For example, the type of steel plate, the production mode, the heating temperature of the heating furnace, the gas supply amount, etc. in a small lot.
Parameter set
The parameters in (1) are mainly used for describing the specific data characteristics contained in each data sample, the partial parameters are sensitive to the change of the data sample and can change along with the change of the data sample, and the corresponding input data is often specific to one data sample. For example, information on the amount of wear of rolls during production of a steel sheet, the temperature of cooling water during cooling, and the like.
And 4, step 4: according to the actual production sequence of the product, extracting the nearest K (15000) sample data from the cloud database to form a training set D, and simultaneously recording the total number of the data samples in the cloud database as S (28700); the first training is based on the model base which is established in the step 3 and comprises 5 forecasting models, each forecasting model in the model base is respectively trained by utilizing sample data in the training set D aiming at each forecasting model, and then the forecasting model base aiming at each quality index is obtained and recorded as
Wherein Fi cA forecasting model representing the ith quality index;
taking the production process data of the sample as input data, taking the ith quality index data of the sample as tag data, and training a forecasting model of the ith quality index on a cloud server to obtain Fi c(ii) a I.e. training the parameter set in step 3
The parameter (1) of (1);
and 5: forecast model library FcFrom cloud server to edgeThe edge server downloads different forecasting models to different terminal servers for operation, and the user forecasts different quality indexes of the product through the forecasting models in the terminal servers;
step 6: acquiring input data of a forecasting model after data cleaning and data preprocessing are carried out on actual production process data acquired from a sensor in an iron plate production industrial field, transmitting the input data to all terminal servers, forecasting each quality index of a product on each terminal server by using the forecasting model corresponding to the product, transmitting the forecasting result to a user, and making a decision on the production process by the user according to the forecasting result so as to improve the qualification rate of the product;
and 7: when the production process of the steel plate is finished, maintaining the corresponding parameter set of the forecasting model on each terminal server
Wherein i ∈ {1,2, …, w }, and using actual production process data of the product to perform parameter set on each forecasting model
The parameters in the model are corrected in real time to obtain a new forecasting model Fi tAnd replacing the original forecast model in the terminal server at the moment, using forecast model Fi tForecasting the subsequent products; the real-time correction is to adopt different correction methods to the parameter set according to different industrial fields and modeling algorithms
Correcting the parameters in (1);
and 8: storing the actual production data and the quality index data of the steel plate into a historical database of the edge end; judging the number of all samples in the historical database of the edge end at the moment, if the number of the samples at the moment is less than n and is 1000, turning to the step 6, and continuing to forecast the quality index of the subsequent product; if the number of the samples is larger than n, which is 1000, the step is switched to step 9;
and step 9: extracting the production data of 1000 products from the historical database of the edge end as a new training set d, and aiming at each forecast model in the forecast model base, on the edge end server, utilizing the sample data in the training set d to carry out the parameter set in the model
Correcting in real time to obtain a prediction model Fi eAll the corrected forecasting models are combined into a new forecasting model library
Due to the parameter set
The parameters in (1) do not change due to the change of the small batch data samples in the production process, so the parameter set in the fixed prediction model
Correcting parameter set in each prediction model on edge end server by using data sample in training set d
Thereby obtaining a new forecasting model Fi eAnd then a new forecasting model library is formed. Fi eCompared with Fi cOnly the parameter set is changed
And the number of samples in the training set D is much smaller than the number of samples in the training set D, so the model F is predictedi eThe training process can save more computing resources, the training speed is faster, and the time is saved.
Step 10: using edge server, FeThe forecasting models in the system are respectively downloaded to corresponding terminal servers and replace the original forecasting models; calling the retrained forecasting model by a user through different terminal servers, and forecasting the product data in a new round; synchronizing the data samples in the edge database to the cloud database, emptying data information in the edge database, and storing product data of a new round of production into the edge database;
step 11: judging the number of samples in the cloud database at the moment, judging whether the total number of the samples in the cloud database is increased by N to 10000 samples (N is far more than N) compared with S, if so, returning to the step 4, counting the total number of the samples in the cloud database again, namely S + N, and retraining FcThe predictive model of (1); if not, returning to the step 6, and forecasting the product quality by using a forecasting model on the terminal server;
finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit of the corresponding technical solutions and scope of the present invention as defined in the appended claims.
Claims (2)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010051048.0A CN111258984B (en) | 2020-01-17 | 2020-01-17 | End-edge-cloud collaborative forecasting method for product quality in industrial big data environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010051048.0A CN111258984B (en) | 2020-01-17 | 2020-01-17 | End-edge-cloud collaborative forecasting method for product quality in industrial big data environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111258984A CN111258984A (en) | 2020-06-09 |
| CN111258984B true CN111258984B (en) | 2021-06-22 |
Family
ID=70954189
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010051048.0A Active CN111258984B (en) | 2020-01-17 | 2020-01-17 | End-edge-cloud collaborative forecasting method for product quality in industrial big data environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111258984B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111741133B (en) * | 2020-08-17 | 2020-11-24 | 成都信息工程大学 | A cloud-side-terminal collaborative meteorological intelligent early warning system |
| CN113110311A (en) * | 2021-03-05 | 2021-07-13 | 中海创科技(福建)集团有限公司 | Industrial big data online modeling and model sharing method, device, equipment and medium |
| CN113220455B (en) * | 2021-05-21 | 2024-08-13 | 润联智能科技股份有限公司 | Cloud edge cooperation-based data preprocessing method and system |
| CN113379135A (en) * | 2021-06-18 | 2021-09-10 | 浙江大学 | Intelligent production line product quality low-delay integrated prediction method and system based on cloud edge collaborative computing |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101728183B1 (en) * | 2012-07-06 | 2017-04-18 | 한국전자통신연구원 | Real time water quality prediction apparatus and method using hydrodynamic model |
| CN108414682A (en) * | 2018-01-29 | 2018-08-17 | 东莞理工学院 | A fast calibration method for air quality monitoring data based on wireless sensor network |
| CN110363355B (en) * | 2019-07-16 | 2022-11-29 | 东北大学 | A cloud-edge collaborative forecasting system and method for alumina production indicators |
-
2020
- 2020-01-17 CN CN202010051048.0A patent/CN111258984B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111258984A (en) | 2020-06-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111258984B (en) | End-edge-cloud collaborative forecasting method for product quality in industrial big data environment | |
| CN115578015A (en) | The whole process supervision method, system and storage medium of sewage treatment based on Internet of Things | |
| CN111814956B (en) | A multi-task learning air quality prediction method based on multi-dimensional quadratic feature extraction | |
| CN118469313B (en) | Road bridge construction monitoring method and system based on deep learning and BIM | |
| Wan et al. | Network traffic prediction based on LSTM and transfer learning | |
| CN117252851B (en) | Standard quality detection management platform based on image detection and identification | |
| CN115860211A (en) | Casting blank quality prediction method based on local online modeling | |
| CN117689256A (en) | A quality traceability method for aluminum alloy melting and casting products | |
| CN114943328A (en) | SARIMA-GRU time sequence prediction model based on BP neural network nonlinear combination | |
| CN118628097B (en) | Method and system for managing full life cycle cost of equipment | |
| CN110674883A (en) | Active learning method based on k nearest neighbor and probability selection | |
| CN111651890B (en) | Data-driven aluminum electrolysis digital twin factory, control method and system | |
| CN111258996B (en) | A product quality multi-source deep fusion forecasting method for industrial big data | |
| CN116957367B (en) | Parameter multi-time scale prediction method and system for integrated energy system operation strategy | |
| CN118607722A (en) | Grain delivery farmer identity verification and grain yield prediction system based on image recognition | |
| CN116960962A (en) | Mid-long term area load prediction method for cross-area data fusion | |
| CN113300884B (en) | A step-by-step network traffic prediction method based on GWO-SVR | |
| CN116629918A (en) | User consumption prediction method and system based on cross-border electronic commerce | |
| CN115730197A (en) | Converter steelmaking end point carbon temperature prediction method based on DRSuPAE and application | |
| CN116304691A (en) | A Network Analysis Method for Multi-view Contrastive Learning Based on Average Teacher | |
| CN115374848A (en) | Product quality prediction method based on gated multi-expert hybrid stacking self-encoder | |
| JPH11126102A (en) | How to forecast gas demand | |
| CN118608526B (en) | Industrial Internet detection methods, systems, media and equipment based on digital twins | |
| CN118427562B (en) | Multi-device federal dynamic graph-oriented multi-element time sequence prediction method and system | |
| CN117174197B (en) | Ozone pollution risk forecasting method based on dynamic supervised machine 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 |