Construction method of 3D model for monitoring key parameters of blast furnace

Abstract

The invention discloses a construction method of a 3D model for monitoring key parameters of a blast furnace, and belongs to the technical field of industrial network platforms, and the construction method comprises the following steps: S1, collecting the key parameters of the blast furnace and storing the key parameters in an upper computer; s2, storing the key parameters of the blast furnace in the upper computer into a network server; s3, constructing a 3D model for monitoring the key parameters of the blast furnace based on a B / S architecture according to the key parameters of the blast furnace in the network server; and S4, relevant characteristic data are selected and input into the XGBoost model for parameter prediction, and theoretical reference is provided for blast furnace operation. The furnace condition of the blast furnace can be more intuitively reflected, and production can be better assisted.

Description

Technical Field

[0001] The invention relates to the technical field of industrial network platforms, and in particular to a method for constructing a 3D model for monitoring key parameters of a blast furnace. Background Art

[0002] In the blast furnace metallurgical production process, various data indicators are read by the staff through the host computer control program (such as WINCC, IFIX) in the main control room. If the management staff want to obtain the data, they need to check it by phone, email and on-site. Data transmission has a certain delay and abstractness. In addition, the data in the host computer are all raw sensor data, and no data screening, analysis, and governance have been done. The reference remains at the original stage, and the foreman needs to conduct manual analysis based on past experience. At the same time, the traditional host computer control program is mainly based on 2D screens, and the auxiliary functions for production operations of the staff are limited. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a method for constructing a 3D model for monitoring key parameters of a blast furnace, which can more intuitively reflect the blast furnace conditions and better assist production.

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is: A method for constructing a 3D model for monitoring key parameters of a blast furnace comprises the following steps: S1, collect key parameters of blast furnace and store them in the host computer; S2, storing the key parameters of the blast furnace in the host computer to the network server; S3. According to the key parameters of the blast furnace in the network server, a 3D model for monitoring the key parameters of the blast furnace based on the B / S architecture is constructed; S4. Select relevant feature data and input them into the XGBoost model for parameter prediction to provide a theoretical reference for operating the blast furnace.

[0005] A further improvement of the technical solution of the present invention is that the key parameters of the blast furnace include: blast furnace cold air flow, hot air pressure, permeability index, pressure difference, hot air temperature, top pressure, oxygen enrichment flow, gas utilization rate, hourly coal volume and tuyere small jacket water temperature.

[0006] A further improvement of the technical solution of the present invention is that in S2, the key parameters of the blast furnace stored in the host computer are read to the network server in real time by means of kepserver.

[0007] A further improvement of the technical solution of the present invention is that in S3, .NET, highcharts, Ajax, SQLserver web page programming, and database programming technology are combined to construct a 3D model for monitoring key parameters of a blast furnace based on a B / S architecture. The 3D model for monitoring key parameters of a blast furnace based on a B / S architecture includes a rotatable blast furnace model, a blast furnace key parameter table, and a real-time updated blast furnace key parameter curve chart.

[0008] A further improvement of the technical solution of the present invention is that the rotatable blast furnace model is divided into several grids, each grid corresponds to a position in the actual application of the blast furnace, and integrates the data collected by each sensor at the actual application position to display the furnace wall temperature; the rotatable blast furnace model is used to rotate and display the data of each grid point in real time.

[0009] A further improvement of the technical solution of the present invention is that the key parameter table of the blast furnace displays the blast furnace cold air flow, hot air pressure, permeability index, pressure difference, hot air temperature, top pressure, oxygen enrichment flow, gas utilization rate, hourly coal volume and tuyere jacket water temperature.

[0010] The further improvement of the technical solution of the present invention lies in that: the key parameter curve diagram of the blast furnace vertically compares the various related indicators; the first group of related indicators includes: penetration index, pressure difference, upper pressure difference, furnace top pressure and CO utilization rate; the second group of related indicators includes: cold air flow, hot air pressure, hot air temperature, oxygen-enriched flow before the machine and oxygen-enriched flow after the machine; the third group of related indicators includes: changes in the readings of the radar probe, and the readings of the radar probe reflect the monitoring of the material height in the blast furnace.

[0011] A further improvement of the technical solution of the present invention is that S4 specifically includes the following steps: S41, selecting a time base point, and inputting relevant feature data within a period of time before the time base point into the XGBoost model; The characteristic data include the key parameters of the blast furnace in the 3D model of the key parameters of the blast furnace during the corresponding smelting time and the component analysis of the weighing data of the sintered ore, pellets and coke raw fuel in the batch collection record; Perform Spearman correlation analysis on the feature data and prediction parameters, and select the top N feature data as relevant feature data; S42, the historical values ​​within this period of time in the 3D model of the key parameters of the blast furnace are transferred to the performance indicators of the XGBoost model output for fitting and scoring, and the optimization is stopped when the threshold score is reached; when the threshold score is not reached, the number of related feature data is increased to N+m, that is, m new related feature data are added on the basis of the original related feature data, until the historical values ​​within this period of time in the 3D model of the key parameters of the blast furnace are fitted and scored with the performance indicators of the XGBoost model output, reaching the threshold score; S43. Input the adjusted relevant feature data into the XGBoost model to obtain the predicted parameters, which are displayed in the 3D model for monitoring the key parameters of the blast furnace.

[0012] A further improvement of the technical solution of the present invention is that the prediction parameters include molten iron temperature and Si content.

[0013] Due to the adoption of the above technical solution, the technical progress achieved by the present invention is: 1. In the present invention, data is no longer simply stored in a host computer, but is stored in a network server, which increases the sharing and security of data, and can be used for secondary development and backup of key data based on the data.

[0014] 2. The present invention constructs a 3D blast furnace data model, and the model rotates in real time to display the data of each grid point. The grasp of the blast furnace condition is more intuitive. The addition of multiple real-time curves can vertically compare various related indicators to better assist production.

[0015] 3. The system design of the present invention is based on B / S architecture, which can help managers to remotely view blast furnace index data anytime and anywhere, which is more convenient. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is the first batch of relevant characteristic parameter analysis in this embodiment; Figure 2 This is a diagram showing the fitting results of the predicted values ​​and actual values ​​under the first batch of relevant characteristic parameters in this embodiment; Figure 3 This is the second batch of relevant characteristic parameter analysis in this embodiment; Figure 4 This is the second batch of relevant characteristic parameter analysis in this embodiment; Figure 5 This is the second batch of relevant characteristic parameter analysis in this embodiment; Figure 6 This is a fitting result diagram of the predicted values ​​and actual values ​​under the second batch of relevant characteristic parameters in this embodiment. DETAILED DESCRIPTION

[0017] The present invention is further described in detail below with reference to the accompanying drawings and embodiments: In the description of the present invention, it should be understood that the terms "first", "second", etc. are only used for descriptive purposes and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first", "second", etc. may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "several" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0018] A method for constructing a 3D model for monitoring key parameters of a blast furnace comprises the following steps: S1, collect key parameters of blast furnace and store them in the host computer; The key parameters of the blast furnace include: blast furnace cold air flow, hot air pressure, permeability index, pressure difference, hot air temperature, top pressure, oxygen enrichment flow, gas utilization rate, hourly coal volume and tuyere small jacket water temperature; S2, storing the key parameters of the blast furnace in the host computer to the network server; By using KepServer, the key parameters of the blast furnace stored in the host computer are read to the network server in real time; S3. According to the key parameters of the blast furnace in the network server, a 3D model for monitoring the key parameters of the blast furnace based on the B / S architecture is constructed; Combining .NET, highcharts, Ajax, SQLserver web programming, and database programming technologies, a 3D model for monitoring key parameters of a blast furnace based on a B / S architecture is constructed. The 3D model for monitoring key parameters of a blast furnace based on a B / S architecture includes a rotatable blast furnace model, a blast furnace key parameter table, and a real-time updated blast furnace key parameter curve chart; The rotatable blast furnace model is divided into several grids, each grid corresponds to a position in the actual application of the blast furnace, and integrates the data collected by each sensor at the actual application position to display the furnace wall temperature; the rotatable blast furnace model is used to rotate and display the data of each grid point in real time; The table of key parameters of the blast furnace shows the blast furnace cold air flow, hot air pressure, permeability index, pressure difference, hot air temperature, top pressure, oxygen enrichment flow, gas utilization rate, hourly coal volume and tuyere small jacket water temperature.

[0019] The key parameter curve of the blast furnace compares the related indicators vertically to better assist production; the first group of related indicators include: penetration index, pressure difference, upper pressure difference, furnace top pressure, CO utilization rate; the second group of related indicators include: cold air flow, hot air pressure, hot air temperature, oxygen-enriched flow before the machine, and oxygen-enriched flow after the machine; the third group of related indicators include: changes in radar probe readings, and the radar probe readings reflect the monitoring of material height in the blast furnace.

[0020] S4, taking the raw material data record and the iron tapping data record as input, and predicting the molten iron temperature prediction value according to the 3D model of the key parameters of the blast furnace, to provide a theoretical reference for the blast furnace foreman to operate the blast furnace, specifically including the following steps: S41, select a time base point, and input relevant feature data in a period of time before the time base point into the XGBoost (eXtreme Gradient Boosting extreme gradient boosting tree) model; The characteristic data include the key parameters of the blast furnace in the 3D model of the key parameters of the blast furnace during the corresponding smelting time and the component analysis of the weighing data of the sintered ore, pellets and coke raw fuel in the batch collection record; Perform Spearman correlation analysis on the feature data and prediction parameters, and select the top N feature data as relevant feature data; S42, the historical values ​​within this period of time in the 3D model of the key parameters of the blast furnace are transferred to the performance indicators of the XGBoost model output for fitting and scoring, and the optimization is stopped when the threshold score is reached; when the threshold score is not reached, the number of related feature data is increased to N+m, that is, m new related feature data are added on the basis of the original related feature data, until the historical values ​​within this period of time in the 3D model of the key parameters of the blast furnace are fitted and scored with the performance indicators of the XGBoost model output, reaching the threshold score; S43, inputting the adjusted relevant feature data into the XGBoost model to obtain prediction parameters, which are displayed in the 3D model for monitoring key parameters of the blast furnace; The prediction parameters include molten iron temperature and Si content. Specific implementation method: Create a new database server, connect to the on-site PLC through kepserver, read relevant data and write it into the database, use the Python program to call the historical data in the database for model training, input the current real-time data into the model, obtain the predicted value of the molten iron temperature, build a 3D model through the CSS style sheet, read the database data with js+asp.Net, and finally realize the 3D blast furnace data model display.

[0022] The prediction method of molten iron temperature (Si content) is: Taking a certain time point as the base point, the characteristic data of a certain time in the past is input into the model, and the molten iron temperature (Si content) at a certain time point in the future is calculated through the model.

[0023] A1. Based on the process data analysis, some characteristic data were selected for Spearman correlation analysis, such as Figure 1 As shown; 15 minutes after the opening time of this iron tapping, coke-1 is the total amount of coke within 1 hour, coke-2 is the total amount within 1-2 hours, and so on. Environmental variables such as oxygen-enriched flow rate and hot air pressure are: oxygen-enriched flow rate-1 is the average value from the integration base point to the previous 50 minutes, and oxygen-enriched flow rate-2 is the average value of minute data from 50 minutes before the integration base point to the previous 100 minutes. Composition-related data such as silicon, FeO and iron tapping are distinguished by furnace, and silicon-1 is the silicon content in the test data of the previous furnace.

[0024] A2. Input data into the XGBoost model to generate results such as Figure 2As shown, the blue line is the actual value and the orange line is the predicted value.

[0025] The average prediction score is 47.79, which is not very high, but the trend fit is good from the results.

[0026] The feature data set was re-analyzed and screened, and the number of process parameters increased to 84, such as Figure 3-5 As shown, the data selection rules are the same as before.

[0027] After cleaning and screening these feature data, they are input into the XGBoost model, and the calculation results are as follows Figure 6 The average score increased to 59.72, which is in the excellent range. From the trend, the fitting degree is high and has certain reference value. A3. Use the adjusted relevant feature parameters for prediction.

[0028] In summary, the present invention can more intuitively reflect the blast furnace conditions and better assist production.

Claims

1. A method for constructing a 3D model for monitoring key parameters of a blast furnace, characterized in that: The following steps are involved: S1, collect key parameters of blast furnace and store them in the host computer; S2, storing the key parameters of the blast furnace in the host computer to the network server; S3. According to the key parameters of the blast furnace in the network server, a 3D model for monitoring the key parameters of the blast furnace based on the B / S architecture is constructed; S4. Select relevant feature data and input them into the XGBoost model for parameter prediction to provide a theoretical reference for operating the blast furnace.

2. The method for constructing a 3D model for monitoring key parameters of a blast furnace according to claim 1, characterized in that: The key parameters of the blast furnace include: blast furnace cold air flow, hot air pressure, permeability index, pressure difference, hot air temperature, top pressure, oxygen enrichment flow, gas utilization rate, hourly coal volume and tuyere jacket water temperature.

3. The method for constructing a 3D model for monitoring key parameters of a blast furnace according to claim 1, characterized in that: In S2, the key parameters of the blast furnace stored in the host computer are read to the network server in real time with the help of kepserver.

4. The method for constructing a 3D model for monitoring key parameters of a blast furnace according to claim 1, characterized in that: In S3, .NET, highcharts, Ajax, SQLserver web programming, and database programming technologies are combined to build a 3D model for monitoring the key parameters of a blast furnace based on the B / S architecture. The 3D model for monitoring the key parameters of a blast furnace based on the B / S architecture includes a rotatable blast furnace model, a blast furnace key parameter table, and a real-time updated blast furnace key parameter curve chart.

5. The method for constructing a 3D model for monitoring key parameters of a blast furnace according to claim 4, characterized in that: The rotatable blast furnace model is divided into several grids, each grid corresponds to a position in the actual application of the blast furnace, and integrates the data collected by various sensors at the actual application position to display the furnace wall temperature; the rotatable blast furnace model is used to rotate and display the data of each grid point in real time.

6. The method for constructing a 3D model for monitoring key parameters of a blast furnace according to claim 4, characterized in that: The key parameters table of the blast furnace includes the blast furnace cold air flow, hot air pressure, permeability index, pressure difference, hot air temperature, top pressure, oxygen enrichment flow, gas utilization rate, hourly coal volume and tuyere small jacket water temperature.

7. The method for constructing a 3D model for monitoring key parameters of a blast furnace according to claim 4, characterized in that: The blast furnace key parameter curve diagram vertically compares various related indicators; The first group of related indicators include: penetration index, pressure difference, upper pressure difference, furnace top pressure and CO utilization rate; the second group of related indicators include: cold air flow, hot air pressure, hot air temperature, oxygen-enriched flow before the machine and oxygen-enriched flow after the machine; the third group of related indicators include: changes in radar probe readings, and the radar probe readings reflect the monitoring of material height in the blast furnace.

8. The method for constructing a 3D model for monitoring key parameters of a blast furnace according to claim 1, characterized in that: S4 specifically includes the following steps: S41, selecting a time base point, and inputting relevant feature data within a period of time before the time base point into the XGBoost model; The characteristic data include the key parameters of the blast furnace in the 3D model of the key parameters of the blast furnace during the corresponding smelting time and the component analysis of the weighing data of the sintered ore, pellets and coke raw fuel in the batch collection record; Perform Spearman correlation analysis on the feature data and prediction parameters, and select the top N feature data as relevant feature data; S42, the historical values ​​within this period of time in the 3D model of the key parameters of the blast furnace are transferred to the performance indicators of the XGBoost model output for fitting and scoring, and the optimization is stopped when the threshold score is reached; when the threshold score is not reached, the number of related feature data is increased to N+m, that is, m new related feature data are added on the basis of the original related feature data, until the historical values ​​within this period of time in the 3D model of the key parameters of the blast furnace are fitted and scored with the performance indicators of the XGBoost model output, reaching the threshold score; S43. Input the adjusted relevant feature data into the XGBoost model to obtain the predicted parameters, which are displayed in the 3D model for monitoring the key parameters of the blast furnace.

9. The method for constructing a 3D model for monitoring key parameters of a blast furnace according to claim 8, characterized in that: The prediction parameters include molten iron temperature and Si content.

Images