CN115187102A - Model evaluation improvement method based on sinter quality prediction - Google Patents
Model evaluation improvement method based on sinter quality prediction Download PDFInfo
- Publication number
- CN115187102A CN115187102A CN202210858429.9A CN202210858429A CN115187102A CN 115187102 A CN115187102 A CN 115187102A CN 202210858429 A CN202210858429 A CN 202210858429A CN 115187102 A CN115187102 A CN 115187102A
- Authority
- CN
- China
- Prior art keywords
- quality
- prediction
- model
- algorithm
- tolerance
- 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.)
- Withdrawn
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 31
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 50
- 238000005245 sintering Methods 0.000 claims abstract description 30
- 229910052500 inorganic mineral Inorganic materials 0.000 claims abstract description 6
- 239000011707 mineral Substances 0.000 claims abstract description 6
- 238000005516 engineering process Methods 0.000 claims abstract description 4
- 238000004458 analytical method Methods 0.000 claims abstract description 3
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000012854 evaluation process Methods 0.000 claims description 2
- 239000000463 material Substances 0.000 claims 1
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 abstract description 4
- 229910052742 iron Inorganic materials 0.000 abstract description 2
- 238000012360 testing method Methods 0.000 abstract description 2
- 238000004519 manufacturing process Methods 0.000 description 5
- 229910000831 Steel Inorganic materials 0.000 description 4
- 239000010959 steel Substances 0.000 description 4
- 238000013210 evaluation model Methods 0.000 description 3
- 238000003723 Smelting Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 229910018072 Al 2 O 3 Inorganic materials 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000007637 random forest analysis Methods 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06395—Quality analysis or management
-
- C—CHEMISTRY; METALLURGY
- C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
- C22B—PRODUCTION AND REFINING OF METALS; PRETREATMENT OF RAW MATERIALS
- C22B1/00—Preliminary treatment of ores or scrap
- C22B1/14—Agglomerating; Briquetting; Binding; Granulating
- C22B1/16—Sintering; Agglomerating
-
- 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"
-
- 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
Abstract
The invention discloses a method for improving model evaluation based on sinter quality prediction, and belongs to the technical field of metallurgical iron making. Predicting the drum index, the yield, the return fines rate, the shrinkage rate and the vertical sintering speed of the sintering ore based on a sintering ore quality prediction technology, constructing a prediction model, evaluating the model, and firstly calculating the absolute value of an error; secondly, confirming a tolerance range; determining the marking value of the quality of each group of sintering ores according to the tolerance algorithm kernel function; and finally calculating the evaluation accuracy of the model. Used as a test at the same time
Interpretable variance algorithm doingAnd performing algorithm performance analysis by comparison. The method can objectively evaluate the prediction model of the sintered mineral quality and shows better performance. Compared with the traditional prediction model evaluation method, the method can finish the model evaluation more quickly.
Description
Technical Field
The invention belongs to the technical field of metallurgical iron making, and particularly relates to a model evaluation improvement method based on sinter quality prediction.
Background
The steel industry is the basic guarantee of national economy, and promotes the development of national economy. The sintering process is an extremely important link in the steel smelting process, and the quality of the sinter directly influences the blast furnace condition and the quality of steel products. The quality detection technology of the sinter has the problem of large lag, and the quality of the sinter cannot be accurately predicted and timely adjusted in the actual production process. Therefore, the establishment of the sinter quality prediction model and the reasonable evaluation thereof have very important significance on the steel smelting process.
At present, an evaluation algorithm applied to a sinter quality prediction model mainly evaluates the prediction accuracy and the convergence rate of the model, and different evaluation methods have different evaluation standards. Mean Square Error (MSE) is commonly used in predictive model evaluation as the most commonly used loss function in linear regression.
In a sintered mineral quality prediction model, the quality of a mean square error, a root mean square error, a mean absolute error and a fitting degree evaluation model is generally used, and the evaluation indexes are optimal under the condition that a predicted value and a true value are equal. However, in the actual prediction of the amount of sintered ore, the prediction result satisfies the quality requirement of the sintered ore as long as the magnitude of the predicted value is within a certain range of the standard value of quality. It can be seen that the conventional evaluation method of the prediction model cannot meet the requirement. Therefore, the sintered ore quality prediction evaluation model based on the tolerance algorithm is provided, and the evaluation method of the sintered ore quality prediction model is improved by using the tolerance algorithm, so that the sintered ore quality prediction process is more fit with the actual sintering process, and the accuracy of the prediction model is improved.
Disclosure of Invention
The invention aims to provide a model evaluation improvement method based on sinter quality prediction, so that a sinter quality prediction model is reasonably evaluated, and the accuracy of the prediction model is improved to a certain extent.
Many production fields are concerned with products, and the quality of the products reaches a certain error range, and the products can be regarded as qualified products. Model evaluation algorithm-tolerance algorithm designed and developed based on concept and improved
To solve the above problem, embodiments of the present invention provide the following solutions: a model evaluation improvement method based on sinter quality prediction is characterized in that a sinter drum index, a finished product rate, a return fines rate, a shrinkage rate and a vertical sintering speed are predicted based on a sinter quality prediction technology, a prediction model is built, a model is evaluated, algorithm comparison analysis is performed, and the evaluation process comprises the following steps:
firstly, calculating absolute values of errors, inputting the real values of the quality of each group of sintering ores and the predicted values of the models, and calculating the absolute values of the errors of the real values and the predicted values of the quality of each group, wherein the calculation formula of the absolute values of the errors is as follows:
Secondly, confirming a tolerance range, and determining a qualified quality range, namely a permissible deviation range according to each group of quality true values aiming at the permissible deviation ranges of different sintered mineral contents, wherein a calculation formula for predicting the permissible deviation range of the quality index is as follows:
And determining the marking value of the quality of each group of sintering ores according to the tolerance algorithm kernel function, wherein the tolerance algorithm kernel function calculation formula is as follows:
And finally, calculating the model evaluation accuracy, wherein the model evaluation accuracy calculation formula is as follows:
Further the algorithm contrasts: at the same time use
And comparing the interpretable variance algorithm to analyze the performance of the algorithm.
The invention has the following advantages:
1. in actual production work of the sintered ore, the quality of the sintered ore can meet the production requirement within a standard interval range. The invention can better reflect the real performance of the sintering prediction model by providing a tolerance algorithm and considering the actual sintering prediction requirement.
2. By using a tolerance algorithm to evaluate the sintering prediction model, the accuracy of the sintering prediction model can be improved to a certain extent, and the reliability of the prediction model is improved.
3. Compared with the traditional model evaluation algorithm, the tolerance algorithm has higher operation speed to a certain extent by using the tolerance algorithm to evaluate the sinter quality prediction model. This characteristic is more pronounced in the case of large amounts of data.
4. By applying the tolerance algorithm concept, the tolerance algorithm concept is applicable to, but not limited to, the field of sintering prediction. The tolerance algorithm is used as a core algorithm of a prediction evaluation model and is suitable for all prediction models with definite definition on prediction indexes.
Drawings
FIG. 1: a flow chart of a tolerance algorithm;
FIG. 2: and (4) a model evaluation flow chart.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.
The embodiment of the invention provides an improved model evaluation method based on sinter quality prediction. A model evaluation flow chart as shown in fig. 2, the method comprises the following steps:
step one, training a prediction model
The variables used in the existing studies are different for the prediction of sinter quality. In the prediction experiment, the test data of the sintering cup is selected for regression prediction, 269 sintering cup data of a certain sintering plant are collected, and part of sintering data is shown in table 1. Sintering the TFe, feO, siO of the mixed ore 2 ，CaO，MgO，Al 2 O 3 The content is used as an input variable, a gradient boosting regression prediction algorithm, a random forest regression prediction algorithm, a gradient boosting regression prediction algorithm and a nearest neighbor regression prediction algorithm are respectively applied to 5 sinter quality indexes of a rotary drum index, a finished product rate, a return mine rate, a shrinkage rate and a vertical sintering speed, a sinter quality prediction model is established by the gradient boosting regression prediction algorithm and the nearest neighbor regression prediction algorithm, and the 5 sinter quality prediction models are respectively trained.
Table 1 example of experimental data for predicting quality of sintered ore
Step two, model evaluation
And after the training of the prediction model is finished, respectively evaluating the 5 sintered mineral quality prediction models by using a tolerance algorithm. When the quality of the sintering ore is predicted, the range of the tolerance deviation of each index is used as a tolerance value according to actual sintering production. The prediction indexes of the quality of the sinter comprise the drum index, the yield, the return rate, the shrinkage rate and the vertical sintering speed of the sinter. The corresponding tolerance values of the predicted quality indices of the sintered ores are shown in table 2.
TABLE 2 tolerance values for prediction of sinter quality
Step three, algorithm comparative analysis
Evaluation improvement of prediction model by applying tolerance algorithm and simultaneously using
And comparing the interpretable variance algorithm. In order to ensure the confidence of the comparison result of the algorithms, the initialization states of the algorithms are the same and are iterated for 300 times, and the average value is used as the model evaluation result. As can be seen from the comparison of the algorithms applied by the predictive models in Table 3, compared to
The accuracy of the tolerance algorithm on 5 sintering prediction indexes is optimal, and the estimator of the sintering ore quality prediction model is improved by using the tolerance algorithm, so that the accuracy of the model is improved to a certain extent while the prediction effect is ensured. In addition, the tolerance algorithm is better than the tolerance algorithm in the running time
And can solveAnd (4) solving a variance algorithm.
TABLE 3 comparison of the algorithms applied to the prediction model
According to the tolerance algorithm obtained by the invention, the estimator of the sintering ore quality prediction model is improved, the process requirement of sintering quality prediction can be met, the accuracy of the prediction model can be improved, and the estimation time of the model can be saved.
The invention is suitable for all sinter quality prediction models. And after predicting the quality of the sinter, obtaining a predicted value and a true value of the quality of the sinter according to the prediction model, and evaluating the prediction model by using a tolerance algorithm.
Claims (2)
1. A model evaluation improvement method based on sinter quality prediction is characterized in that a sinter drum index, a finished product rate, a return fines rate, a shrinkage rate and a vertical sintering speed are predicted based on a sinter quality prediction technology, a prediction model is built, a model is evaluated, and algorithm comparison analysis is performed, wherein the model evaluation process comprises the following steps:
firstly, calculating absolute values of errors, inputting the real values of the quality of each group of sintering ores and the predicted values of the models, and calculating the absolute values of the errors of the real values and the predicted values of the quality of each group, wherein the calculation formula of the absolute values of the errors is as follows:
secondly, confirming a tolerance range, and determining a qualified quality range, namely a permissible deviation range according to each group of quality true values aiming at the permissible deviation ranges of different sintered mineral contents, wherein a calculation formula for predicting the permissible deviation range of the quality index is as follows:
and determining the marking value of the quality of each group of sintering ores according to the tolerance algorithm kernel function, wherein the tolerance algorithm kernel function calculation formula is as follows:
and finally, calculating the model evaluation accuracy, wherein the model evaluation accuracy calculation formula is as follows:
2. A sintered mineral-based material according to claim 1Improved method for model evaluation under quantity prediction, the algorithm contrasts and analyzes for simultaneous use
And comparing the interpretable variance algorithm to analyze the performance of the algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210858429.9A CN115187102A (en) | 2022-07-21 | 2022-07-21 | Model evaluation improvement method based on sinter quality prediction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210858429.9A CN115187102A (en) | 2022-07-21 | 2022-07-21 | Model evaluation improvement method based on sinter quality prediction |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115187102A true CN115187102A (en) | 2022-10-14 |
Family
ID=83519532
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210858429.9A Withdrawn CN115187102A (en) | 2022-07-21 | 2022-07-21 | Model evaluation improvement method based on sinter quality prediction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115187102A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116124677A (en) * | 2023-04-18 | 2023-05-16 | 江苏沙钢集团有限公司 | Rapid evaluation method for air permeability of blast furnace sintering mineral aggregate layer |
-
2022
- 2022-07-21 CN CN202210858429.9A patent/CN115187102A/en not_active Withdrawn
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116124677A (en) * | 2023-04-18 | 2023-05-16 | 江苏沙钢集团有限公司 | Rapid evaluation method for air permeability of blast furnace sintering mineral aggregate layer |
| CN116124677B (en) * | 2023-04-18 | 2023-06-16 | 江苏沙钢集团有限公司 | Rapid evaluation method for air permeability of blast furnace sintering mineral aggregate layer |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110066895B (en) | Stacking-based blast furnace molten iron quality interval prediction method | |
| CN109935280B (en) | Blast furnace molten iron quality prediction system and method based on ensemble learning | |
| Li et al. | Dynamic time features expanding and extracting method for prediction model of sintering process quality index | |
| CN115187102A (en) | Model evaluation improvement method based on sinter quality prediction | |
| Yan et al. | DSTED: A denoising spatial–temporal encoder–decoder framework for multistep prediction of burn-through point in sintering process | |
| CN116307289B (en) | Textile processing procedure parameter detection and prediction method, system and storage medium | |
| CN106777684B (en) | Method for establishing comprehensive coke ratio prediction model and predicting comprehensive coke ratio | |
| CN102601881A (en) | Method for monitoring on-line quality and updating prediction model of rubber hardness | |
| Feng et al. | Endpoint temperature prediction of molten steel in RH using improved case-based reasoning | |
| CN113642666A (en) | Active enhanced soft measurement method based on sample expansion and screening | |
| CN111047103A (en) | Sintering process carbon effect online prediction method and device based on multi-working-condition mode identification | |
| CN111915080A (en) | Raw fuel cost optimal proportioning method based on molten iron quality constraint | |
| Shi et al. | Key issues and progress of industrial big data-based intelligent blast furnace ironmaking technology | |
| CN114764547A (en) | Heating furnace steel billet temperature prediction method and device integrating mechanism and deep learning | |
| CN107808221A (en) | Blast furnace material distribution Parameter Decision Making method based on case matching | |
| Kai et al. | End temperature prediction of molten steel in RH based on case-based reasoning with optimized case base | |
| CN102621953A (en) | Automatic online quality monitoring and prediction model updating method for rubber hardness | |
| CN111798023B (en) | Comprehensive coke ratio prediction method in steelmaking sintering production | |
| CN107679659B (en) | Novel industrial power consumption prediction method | |
| CN110276128B (en) | Underflow concentration prediction method of thickener based on DAJYPLS algorithm | |
| CN114525372B (en) | Multi-mode fusion based blast furnace state monitoring method and device | |
| Jiang et al. | A novel intelligent monitoring method for the closing time of the taphole of blast furnace based on two-stage classification | |
| CN113223634B (en) | Blast furnace molten iron silicon content prediction method based on two-dimensional self-attention enhanced GRU model | |
| CN112434961B (en) | Method and device for predicting temperature drop of molten iron on iron-steel interface and terminal equipment | |
| CN109840309A (en) | A kind of calculation method of iron melt desulfurizing agent dosage |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20221014 |