CN114971064A - Hot-rolled strip steel surface defect prediction method based on NGboost algorithm - Google Patents
Hot-rolled strip steel surface defect prediction method based on NGboost algorithm Download PDFInfo
- Publication number
- CN114971064A CN114971064A CN202210669138.5A CN202210669138A CN114971064A CN 114971064 A CN114971064 A CN 114971064A CN 202210669138 A CN202210669138 A CN 202210669138A CN 114971064 A CN114971064 A CN 114971064A
- Authority
- CN
- China
- Prior art keywords
- hot
- rolled strip
- strip steel
- data
- surface defect
- 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
- 229910000831 Steel Inorganic materials 0.000 title claims abstract description 78
- 239000010959 steel Substances 0.000 title claims abstract description 78
- 230000007547 defect Effects 0.000 title claims abstract description 68
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000004519 manufacturing process Methods 0.000 claims abstract description 42
- 238000012549 training Methods 0.000 claims abstract description 12
- 238000012360 testing method Methods 0.000 claims abstract description 10
- 238000009628 steelmaking Methods 0.000 claims abstract description 4
- 230000002159 abnormal effect Effects 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 7
- 238000010079 rubber tapping Methods 0.000 claims description 7
- 238000009499 grossing Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000003723 Smelting Methods 0.000 claims description 5
- 239000000463 material Substances 0.000 claims description 5
- 238000009749 continuous casting Methods 0.000 claims description 3
- 238000005520 cutting process Methods 0.000 claims description 3
- 239000006185 dispersion Substances 0.000 claims description 3
- 230000008030 elimination Effects 0.000 claims description 3
- 238000003379 elimination reaction Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 239000007788 liquid Substances 0.000 claims description 3
- 238000010801 machine learning Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 239000002893 slag Substances 0.000 claims description 3
- 238000010561 standard procedure Methods 0.000 claims description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 230000008901 benefit Effects 0.000 abstract description 2
- 238000005096 rolling process Methods 0.000 abstract description 2
- 238000001514 detection method Methods 0.000 description 15
- 238000010586 diagram Methods 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000002203 pretreatment Methods 0.000 description 2
- 238000005097 cold rolling Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005098 hot rolling Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- 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
- G06Q50/04—Manufacturing
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Economics (AREA)
- Mathematical Physics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Mathematical Optimization (AREA)
- Operations Research (AREA)
- Pure & Applied Mathematics (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Algebra (AREA)
- Manufacturing & Machinery (AREA)
- Quality & Reliability (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Databases & Information Systems (AREA)
- Development Economics (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
Abstract
A hot-rolled strip steel surface defect prediction method based on NGboost algorithm belongs to the technical field of automatic control of steel-making and steel-rolling processes. The method comprises the following steps: collecting the surface defect data of the hot-rolled strip steel, calculating the average value and the standard deviation of the data of each production parameter, dividing the surface defect data set of the hot-rolled strip steel into a training set and a test set by adopting a random division mode, inputting the data of the test set, verifying the prediction accuracy of the model, and comparing the accuracy of the prediction result of the model. The method has the advantages of overcoming the defects of comprehensiveness, timeliness and rapidity in the existing method for detecting the surface defects of the hot-rolled strip steel.
Description
Technical Field
The invention belongs to the technical field of automatic control of steel making and rolling processes, and particularly relates to a hot-rolled strip steel surface defect prediction method based on an NGboost algorithm.
Background
Currently, intelligent manufacturing becomes the core power of a new scientific and technical revolution and industry transformation, and based on the outstanding characteristics and existing problems of the steel industry in China, the assistance of artificial intelligence and machine learning technology is needed to improve the product quality, reduce the production cost and realize accurate service.
Along with the increasingly thinner thickness of the hot-rolled strip steel, the hot-rolled strip steel products are widely applied in the fields of automobile manufacturing, household appliances, building industry and the like, the application industries and the product types of the hot-rolled strip steel products are increasingly increased, the quality of the hot-rolled strip steel products needs to be continuously improved in order to meet the requirement of multi-scene application, and parameters in the production process need to be continuously adjusted so as to achieve the best product quality and the maximum production benefit.
In the traditional defect detection of the hot-rolled strip steel, the first method is manual naked eye detection, and the detection mode of carrying out surface quality spot check on a section of short tail part of the strip steel cannot reflect the full appearance of the surface quality of the strip steel in time, so that difficulty is brought to the production of the next procedure, and the quality objection of users is caused. The second method is the traditional machine vision detection, the quality of hot-rolled strip steel products is detected by utilizing the technologies such as machine vision, the surface quality of the steel plate is monitored on line in real time by utilizing a non-contact detection method, and the surface quality data of the moving steel plate is input into an image processing system through a surface scanning CCD high-speed camera system arranged above and below a hot-rolling roller way. The third method is detection based on deep learning, the method generally uses a convolutional neural network to extract features, and then locates and classifies defects, compared with manual naked eye detection and traditional machine vision detection, the defect detection based on deep learning saves manpower under the condition of ensuring the recognition rate and accuracy, however, the existing defect detection method based on deep learning has large network parameters when performing feature extraction, so that the calculation amount is large, the detection speed is slow, and the requirement of real-time property of hot-rolled steel plate detection cannot be met.
Disclosure of Invention
The invention aims to provide a hot-rolled strip steel surface defect prediction method based on an NGboost algorithm, which solves the defects of comprehensiveness, timeliness and rapidity in the existing hot-rolled strip steel surface defect detection method.
The invention comprises the following steps:
(1) collecting surface defect data of hot-rolled strip steel: the method comprises the production parameters in the steelmaking-continuous casting process, such as: smelting number, liquidus temperature, tundish molten steel quantity, slab cutting length, average temperature in the tundish, liquid level height, material length and tapping temperature; and the type of defect, such as: slag inclusion, edge crack warping, surface crack warping and the like, collecting and summarizing data, unifying formats, making an Excel table, and generating an initial hot-rolled strip steel surface defect data set;
(2) calculating the average value and standard deviation of each production parameter data;
removing abnormal values of the production data by adopting Lauda criterion in a specific mode of meeting the requirement
And (3) carrying out abnormal value elimination on data of the formula of the condition: wherein x is i Is the data of the ith production parameter,
is the mean value of the production data, S x Is the standard deviation of production data;
and smoothing the data after the abnormal values are removed, and adopting a 5-point 3-time method, wherein a specific formula is as follows:
wherein, Y i Is the production data after the ith abnormal value is removed,
is Y i A smoothed value;
the production data after the smoothing treatment is normalized by a linear normalization method, also called a dispersion standard method,min-max normalized with the transfer function:
wherein X is pre-processing data, X * Processed data, X min Minimum of pre-treatment data, X max Maximum value of data before processing;
(3) dividing a hot-rolled strip steel surface defect data set into a training set and a testing set by adopting a random division mode;
establishing a hot-rolled strip steel surface defect prediction model based on an NGboost algorithm;
inputting training set data and carrying out model training;
selecting proper model parameters
The n _ estimators parameter is the number of the integrated medium-weak estimators, namely the number of the trees, and is set to be 1000;
the learning _ rate parameter, namely the learning rate, is used for controlling the learning progress of the model and is set to be 0.1;
the minipatch _ frac parameter is a small batch fraction and is used for sampling samples and obtaining a small sample set;
the col _ sample parameter is column subsampling, and a part of features are randomly selected before each node in the same layer is split;
selecting an accuracy parameter as an evaluation index of a hot-rolled strip steel surface defect multi-classification prediction model,
wherein, TP i To correctly predict the true class i as class i; FP i To incorrectly predict the true class i as class i.
(4) Inputting test set data, and verifying the prediction accuracy of the model;
establishing a hot-rolled strip steel surface defect prediction model based on other algorithms, wherein the hot-rolled strip steel surface defect prediction model comprises a hot-rolled strip steel surface defect prediction model based on an XGboost algorithm, a hot-rolled strip steel surface defect prediction model based on a LightGBM algorithm and a hot-rolled strip steel surface defect prediction model based on a Catboost algorithm;
(5) comparing the accuracy of the model prediction results;
analyzing the characteristic importance of the model, and analyzing production parameters which have great influence on the result, such as: actual tapping temperature, actual length of a plate blank, maximum pressure of a water feeding port, smelting number, average heat exchange, minimum molten steel amount of a tundish and the like;
analyzing the correlation among the input features of the model, and exploring the correlation among the production parameters;
the invention provides a hot-rolled strip steel surface defect detection method, which is based on a new machine learning algorithm, namely an NGboost algorithm, proposed in 2019 and 10 months, is combined with production parameters in a steelmaking process to predict the hot-rolled strip steel surface defects, and the prediction result is compared with a hot-rolled strip steel surface defect prediction model based on an XGboost algorithm, a hot-rolled strip steel surface defect prediction model based on a LightGBM algorithm and a hot-rolled strip steel surface defect prediction model based on a CatBOost algorithm to verify the accuracy of the models. The method can overcome the defects of comprehensiveness, timeliness and rapidity in the existing hot-rolled strip steel surface defect detection method to a certain extent, and has certain theoretical significance and practical significance on how to conform to the development trend of the times, combine the artificial intelligence technology with the steel industry and solve the actual problem in the steel production process.
Drawings
FIG. 1 is a schematic diagram of a hot-rolled strip steel surface defect model established based on an NGboost algorithm.
FIG. 2 is a prediction result scatter diagram of a prediction model established based on the XGboost algorithm.
Fig. 3 is a prediction result scatter diagram of a prediction model established based on the LightGBM algorithm.
FIG. 4 is a prediction result scatter plot of a prediction model established based on the Catboost algorithm.
Fig. 5 is a prediction result scatter diagram of a prediction model established based on the NGBoost algorithm.
FIG. 6 is an input feature importance result diagram of a hot-rolled strip steel surface defect prediction model established based on an NGboost algorithm.
FIG. 7 is a graph showing the results of correlation between input characteristics of surface defects of a hot rolled steel strip.
Detailed Description
The invention provides a method for predicting the surface defects of hot-rolled strip steel, which is used for further explaining the application by combining the attached drawings and the specific implementation process so as to help technicians in the industry to more completely and clearly understand the contents of the invention in order to make the model construction and calculation processes of the invention clearer.
Collecting surface defect data of the hot-rolled strip steel, wherein the data comprise production parameters in the steelmaking-continuous casting process, such as: smelting number, liquidus temperature, tundish molten steel quantity, slab cutting length, average temperature in the tundish, liquid level height, material length and tapping temperature. And the type of defect, such as: slag inclusion, edge crack warping, surface crack warping and the like, collecting and summarizing data, unifying formats, making an Excel table, and generating an initial hot-rolled strip steel surface defect data set;
calculating the average value and standard deviation of each production parameter data;
removing abnormal values of the production data by adopting Lauda criterion in a specific mode of meeting the requirement
And (3) carrying out abnormal value elimination on data of the formula of the condition: wherein x is i Is the data of the ith production parameter,
is the mean value of the production data, S x Is the standard deviation of production data;
and smoothing the data after the abnormal values are removed, and adopting a 5-point 3-time method, wherein a specific formula is as follows:
wherein, Y i Is the production data after the ith abnormal value is removed，
Is Y i A smoothed value;
the production data after the smoothing processing is normalized by a linear normalization method, also called a dispersion standard method and min-max standardization, and the conversion function is as follows:
wherein X is pre-processing data, X * Processed data, X min Minimum of pre-treatment data, X max Maximum value of data before processing;
dividing a hot-rolled strip steel surface defect data set into a training set and a testing set in a random division mode, wherein the training set accounts for 90% of the total data set, and the testing set accounts for 10% of the total data set;
establishing a hot-rolled strip steel surface defect prediction model based on an NGboost algorithm;
inputting training set data to perform model training, wherein FIG. 1 is a training and verifying process of a hot-rolled strip steel surface defect model established based on an NGboost algorithm;
selecting proper model parameters to obtain the optimal test precision, wherein n _ estimators parameters are the number of the integrated medium-weak estimators, namely the number of the trees, and are set to be 1000; the learning _ rate parameter, namely the learning rate, is used for controlling the learning progress of the model and is set to be 0.1;
the minipatch _ frac parameter is a small batch fraction and is used for sampling samples and obtaining a small sample set; the col _ sample parameter, i.e., the column sub-sample, is a part of the features randomly selected before splitting each node in the same layer. According to research, the prediction precision effect of the model is better when n _ estimators is 1000, learning _ rate is 0.01, minibratch _ frac is 0.5 and col _ sample is 0.5;
accuracy is selected as an evaluation index of a hot-rolled strip steel surface defect multi-classification prediction model,
can be understood as the ratio of the number of positive examples of correct prediction to the total number of positive examples of prediction;
inputting test set data, and verifying the prediction accuracy of the model;
establishing a hot-rolled strip steel surface defect prediction model based on other algorithms, wherein the hot-rolled strip steel surface defect prediction model comprises a hot-rolled strip steel surface defect prediction model based on an XGboost algorithm, a hot-rolled strip steel surface defect prediction model based on a LightGBM algorithm and a hot-rolled strip steel surface defect prediction model based on a Catboost algorithm;
comparing the model prediction results, wherein the prediction results are shown in FIG. 2, and the prediction accuracy of the hot-rolled strip steel surface defect prediction model based on the NGboost algorithm is obviously superior to that of the other three prediction models;
analyzing the characteristic importance of the model, analyzing the production parameters which have great influence on the result, and taking the characteristic importance result as a figure 3, wherein the production parameters which have great influence on the types of the surface defects of the hot-rolled strip steel are as follows: material length, offline time, average heat exchange (north), average heat exchange (south), average heat exchange (east), average heat exchange (west), actual tapping temperature, minimum molten steel amount of a tundish, actual length of a plate blank, maximum pressure of a water feeding port and the like;
the correlation analysis among the input characteristics of the model is carried out to explore the correlation among the production parameters, fig. 4 is a correlation result graph among the important characteristics, and can be obtained that the correlation between the two characteristics of tundish molten steel temperature and liquidus temperature is strong, the correlation between the four characteristics of heat exchange average (north), heat exchange average (south), heat exchange average (east) and heat exchange average (west) is strong, the correlation between the three characteristics of upper sliding plate pressure average, material length and actual tapping temperature and other characteristics is low, and the correlation between the two characteristics of ELM _ NB _ ACT and the cold rolling mill train and other characteristics is high.
Finally, it should be noted that the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to examples, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.
Claims (1)
1. A hot-rolled strip steel surface defect prediction method based on an NGboost algorithm is characterized by comprising the following steps:
(1) collecting surface defect data of hot-rolled strip steel: the method comprises the production parameters and the defect types in the steelmaking-continuous casting process, wherein the production parameters comprise: smelting number, liquidus temperature, tundish molten steel quantity, slab cutting length, average temperature in the tundish, liquid level height, material length and tapping temperature; the types of defects include: slag inclusion, edge crack warping skin and surface crack warping skin; collecting and summarizing the data, unifying the format, making an Excel table, and generating an initial hot-rolled strip steel surface defect data set;
(2) calculating the average value and standard deviation of each production parameter data;
removing abnormal values of the production data by adopting Lauda criterion in a specific mode of meeting the requirement
And (3) carrying out abnormal value elimination on data of the formula of the condition: wherein x is i Is the data of the ith production parameter,
is the mean value of the production data, S x Is the standard deviation of production data;
and smoothing the data after the abnormal values are removed, and adopting a 5-point 3-time method, wherein a specific formula is as follows:
wherein, Y i Is the production data after the ith abnormal value is removed,
is Y i A smoothed value;
the production data after the smoothing processing is normalized by a linear normalization method, also called a dispersion standard method and min-max standardization, and the conversion function is as follows:
wherein, X: data before processing, X * : processed data, X min : minimum of data before processing, X max : maximum value of pre-processing data;
(3) dividing a hot-rolled strip steel surface defect data set into a training set and a testing set by adopting a random division mode;
establishing a hot-rolled strip steel surface defect prediction model based on an NGboost algorithm;
inputting training set data and carrying out model training;
selecting proper model parameters
The n _ estimators parameter is the number of the integrated medium-weak estimators, namely the number of the trees, and is set to be 1000;
the learning _ rate parameter, namely the learning rate, is used for controlling the learning progress of the model and is set to be 0.1;
the minipatch _ frac parameter is a small batch fraction and is used for sampling samples and obtaining a small sample set;
the col _ sample parameter is column subsampling, and a part of features are randomly selected before each node in the same layer is split;
selecting an accuracycacy accuracy parameter as an evaluation index of a hot-rolled strip steel surface defect multi-classification prediction model,
wherein, TP i To correctly predict the true class i as class i; FP i To incorrectly predict the true class i as class i;
(4) inputting test set data, and verifying the prediction accuracy of the model;
establishing hot-rolled strip steel surface defect prediction models based on other algorithms, wherein the hot-rolled strip steel surface defect prediction models comprise a hot-rolled strip steel surface defect prediction model based on an XGboost algorithm, a hot-rolled strip steel surface defect prediction model based on a LightGBM algorithm and a hot-rolled strip steel surface defect prediction model based on a Catboost algorithm;
(5) comparing the accuracy of the model prediction results;
analyzing the characteristic importance of the model, and analyzing the production parameters which have great influence on the result: the method comprises the steps of actual tapping temperature, actual length of a plate blank, maximum pressure of a water feeding port, smelting number, average heat exchange and minimum molten steel amount of a tundish;
analyzing the correlation among the input features of the model, and exploring the correlation among the production parameters;
predicting the surface defects of the hot-rolled strip steel based on a machine learning algorithm, namely an NGboost algorithm, in combination with production parameters in the steelmaking process, comparing the prediction result with a hot-rolled strip steel surface defect prediction model based on an XGboost algorithm, a hot-rolled strip steel surface defect prediction model based on a LightGBM algorithm and a hot-rolled strip steel surface defect prediction model based on a Catboost algorithm, and verifying the accuracy of the models.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210669138.5A CN114971064A (en) | 2022-06-14 | 2022-06-14 | Hot-rolled strip steel surface defect prediction method based on NGboost algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210669138.5A CN114971064A (en) | 2022-06-14 | 2022-06-14 | Hot-rolled strip steel surface defect prediction method based on NGboost algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114971064A true CN114971064A (en) | 2022-08-30 |
Family
ID=82960960
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210669138.5A Pending CN114971064A (en) | 2022-06-14 | 2022-06-14 | Hot-rolled strip steel surface defect prediction method based on NGboost algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114971064A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117753795A (en) * | 2024-02-07 | 2024-03-26 | 东北大学 | feedforward control method for hot rolled products with multiple steel types and specifications |
| CN117753795B (en) * | 2024-02-07 | 2024-05-31 | 东北大学 | Feedforward control method for hot rolled products with multiple steel types and specifications |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120073778A1 (en) * | 2009-06-26 | 2012-03-29 | Hyundai Steel Company | Method for predicting surface quality of thin slab hot rolled coil and method for producing thin slab hot rolled coil using the same |
| CN110363253A (en) * | 2019-07-25 | 2019-10-22 | 安徽工业大学 | A kind of Surfaces of Hot Rolled Strip defect classification method based on convolutional neural networks |
| CN111402236A (en) * | 2020-03-17 | 2020-07-10 | 北京科技大学 | Hot-rolled strip steel surface defect grading method based on image gray value |
| CN112381787A (en) * | 2020-11-12 | 2021-02-19 | 福州大学 | Steel plate surface defect classification method based on transfer learning |
-
2022
- 2022-06-14 CN CN202210669138.5A patent/CN114971064A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120073778A1 (en) * | 2009-06-26 | 2012-03-29 | Hyundai Steel Company | Method for predicting surface quality of thin slab hot rolled coil and method for producing thin slab hot rolled coil using the same |
| CN110363253A (en) * | 2019-07-25 | 2019-10-22 | 安徽工业大学 | A kind of Surfaces of Hot Rolled Strip defect classification method based on convolutional neural networks |
| CN111402236A (en) * | 2020-03-17 | 2020-07-10 | 北京科技大学 | Hot-rolled strip steel surface defect grading method based on image gray value |
| CN112381787A (en) * | 2020-11-12 | 2021-02-19 | 福州大学 | Steel plate surface defect classification method based on transfer learning |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117753795A (en) * | 2024-02-07 | 2024-03-26 | 东北大学 | feedforward control method for hot rolled products with multiple steel types and specifications |
| CN117753795B (en) * | 2024-02-07 | 2024-05-31 | 东北大学 | Feedforward control method for hot rolled products with multiple steel types and specifications |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114329940A (en) | Continuous casting billet quality prediction method based on extreme learning machine | |
| CN106126941B (en) | A kind of cold mill complex modeling power method | |
| CN115035106A (en) | Strip steel defect intelligent detection method | |
| CN110989510A (en) | Hot galvanizing product full-process quality control and grade automatic judgment system | |
| CN116758084B (en) | Intelligent detection method for welding defects of sheet metal parts based on image data | |
| CN110287456A (en) | Bulk lots volume rolled surface defect analysis method based on data mining | |
| CN111563686A (en) | Cold-rolled silicon steel quality judgment method based on full-process data | |
| CN115601313A (en) | Visual monitoring management system for tempered glass production process | |
| CN114118355A (en) | Stainless steel pickling process parameter control method based on neural network | |
| CN111914889A (en) | Rectifying tower abnormal state identification method based on brief kernel principal component analysis | |
| CN115815345A (en) | Mechanism collaborative prediction method and system for predicting mechanical property of full-flow hot-rolled strip steel | |
| CN109308707B (en) | Non-contact type online measuring method for thickness of aluminum ingot | |
| CN114896735A (en) | Partial least square improved hot-rolled strip steel head wave defect reason identification method | |
| CN114971064A (en) | Hot-rolled strip steel surface defect prediction method based on NGboost algorithm | |
| CN116944818B (en) | Intelligent processing method and system for new energy automobile rotating shaft | |
| KR20190078338A (en) | Apparatus for measuring camber in hot rolling process | |
| CN109622926B (en) | Ingot block edge detection method for ingot casting process | |
| CN111318579A (en) | Roughing camber control method based on data driving | |
| CN111563553A (en) | Method for predicting alkalinity of pellet ore phase based on PCA-CNN classification recognition | |
| CN114565563B (en) | Color steel plate surface abnormity detection method based on artificial intelligence | |
| CN115446125A (en) | Automatic steel rotating control method for medium plate based on machine vision | |
| CN115187102A (en) | Model evaluation improvement method based on sinter quality prediction | |
| CN114998254A (en) | Method and system for automatically grading center segregation of continuous casting slab | |
| CN114472542A (en) | Method, device and equipment for controlling plate shape in production process of hot-rolled strip steel | |
| Feng et al. | A method for surface detect classification of hot rolled strip steel based on Xception |
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 |