Method for analyzing maximum influencing factor of blast furnace lower part heat system based on tuyere monitoring

Abstract

The application is particularly a blast furnace lower part heat system maximum influencing factor analysis method based on tuyere monitoring, a blast furnace tuyere monitoring system is built, intelligent collection of tuyere images based on different brightness is realized, and automatic analysis of tuyere brightness is realized; the automatic adjustment of tuyere monitoring system collection parameters is realized through a pre-calibration method, so as to effectively collect tuyere brightness images; the blast furnace air volume, blast temperature, charge descending speed, blast furnace tapping volume, smelting cycle and other related process parameters and tuyere brightness change trend are analyzed by using a partial least square method to realize automatic analysis of the blast furnace lower part heat system maximum influencing factor. The application can effectively solve the brightness overexposure problem in the blast furnace tuyere image monitoring, realize automatic adjustment of the tuyere image, and combine the tuyere brightness monitoring result with the blast furnace process parameters to effectively perform automatic analysis of the blast furnace lower part heat system maximum influencing factor, optimize the blast furnace lower part regulation and improve the blast furnace control level.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of maximum influencing factor analysis method of blast furnace lower thermal regime, in particular to a maximum influencing factor analysis method of blast furnace lower thermal regime based on tuyere monitoring. BACKGROUND

[0002] The steel industry is the foundation industry of national economy and national defense construction. The current mainstream process of steel in China is mainly the long process of ironmaking-steelmaking model. As the largest single smelting equipment in the world, blast furnace is the core of ironmaking. In order to improve energy utilization efficiency and reduce smelting cost, the current blast furnace generally uses techniques such as pre-tuyere injection of pulverized coal and high blast temperature to realize the optimization of blast furnace smelting efficiency and the regulation of blast furnace lower part. The blast furnace thermal regime is the key to the quality of blast furnace pig iron, smelting efficiency and operation level, and the regulation of blast furnace lower part has a greater impact on it. However, the current blast furnace thermal regime is mainly judged by artificial experience or through tuyere brightness, process parameters, etc., without determining an effective maximum influencing factor analysis method according to the actual situation.

[0003] With the development of artificial intelligence technology and CCD imaging technology, blast furnace tuyere monitoring technology based on machine vision has been widely applied. The patents "A blast furnace tuyere online monitoring imaging device" (CN201520770267.9), "A blast furnace tuyere camera device" (CN201820629808.X), and "A method for supervising and predicting blast furnace abnormal conditions based on tuyere information deep learning" (CN202110187003.0) propose blast furnace tuyere imaging and tuyere state monitoring methods based on machine vision, which can realize automatic monitoring of blast furnace tuyere coal injection and judgment of tuyere state abnormality. It has important guiding significance for improving blast furnace production efficiency, reducing safety accidents, and improving blast furnace quantitative operation level. However, the above inventions are limited to monitoring of tuyere coal injection and tuyere damage, which leads to problems such as overexposure or low brightness of tuyere images under some working conditions, and cannot be combined with blast furnace process parameters to analyze the maximum influencing factor of blast furnace thermal regime. SUMMARY

[0004] The purpose of the present application is to provide a maximum influencing factor analysis method of blast furnace lower thermal regime based on tuyere monitoring to solve the problem of brightness overexposure in blast furnace tuyere image monitoring and realize automatic adjustment of tuyere images. At the same time, the tuyere monitoring results are combined with blast furnace process parameters to effectively analyze the maximum influencing factor of blast furnace lower thermal regime, laying a good foundation for optimizing blast furnace lower regulation and improving blast furnace control level.

[0005] The maximum influencing factor analysis method of blast furnace lower thermal regime based on tuyere monitoring provided by the present application comprises the following steps in sequence:

[0006] S1, collecting real-time blast furnace tuyere images, judging whether the blast furnace tuyere images are in overexposure or overdark state, adjusting the exposure time and gain of the blast furnace tuyere images in overexposure or overdark state by using pre-calibration method until the brightness of the blast furnace tuyere images meets the standard;

[0007] S2, taking a period of time S of blast furnace tuyere images, and taking the average brightness value of a group of the blast furnace tuyere images as the tuyere brightness of the group every fixed time interval s, taking the average value of each blast furnace smelting process parameter at the same fixed time interval s, and the sample data is composed of the tuyere brightness and the average value of all blast furnace smelting process parameters, and the sample number is S / s;

[0008] S3, the blast furnace smelting process parameters include blast furnace blast volume XB, blast temperature XBT, blast furnace burden descending speed XBD, blast furnace tapping volume XI, smelting period XS, coal injection volume XPCI, and oxygen enrichment rate XBO, the tuyere brightness is recorded as L, the average value of the blast furnace smelting process parameters is recorded as X={XB, XBT, XBD, XI, XS, XPCI, XBO} 7×S / s , the dependent variable matrix composed of the tuyere brightness is recorded as Y={L} 1×S / s , the two matrices X and Y are standardized to obtain new data matrices E0 and F0, then a partial least squares regression equation is constructed, and the variable projection importance index is used to measure the importance of the influence of the blast furnace smelting process parameters on the tuyere brightness, the greater the variable projection importance index, the higher the importance of the influence of the blast furnace smelting process parameters on the tuyere brightness, that is, the blast furnace smelting process parameter that has the greatest influence on the blast furnace thermal regime in this period.

[0009] Preferably, the blast furnace tuyere images are extracted by a blast furnace tuyere monitoring system, the blast furnace tuyere monitoring system comprises a high-frame-rate camera based on FPGA, a telecentric lens, a power supply system, a data transmission system, an image storage and processing system, and a high-temperature-resistant dustproof protective kit, the telecentric lens is connected to the high-frame-rate camera based on FPGA, the high-frame-rate camera based on FPGA is connected to the data transmission system, the data transmission system is connected to the image storage and processing system, the power supply system is used to supply power to the blast furnace tuyere monitoring system, the high-temperature-resistant dustproof protective kit is used for high-temperature-resistant and dustproof protection of the blast furnace tuyere monitoring system, and the high-frame-rate camera based on FPGA is used to shoot the blast furnace tuyere images and upload them to the image storage and processing system through the data transmission system.

[0010] Preferably, the high-frame-rate camera based on FPGA judges that the collected real-time blast furnace tuyere images are in overexposure or overdark state when the gray value of the images is in the range of >200 or <100.

[0011] Preferably, the step 1, the blast furnace tuyere image in overexposure or overdark state adopts the pre-calibration method to adjust the exposure time and gain until the average brightness of the blast furnace tuyere image in 5 seconds is in the range of 150 ± 50, then the brightness of the blast furnace tuyere image meets the standard.

[0012] The beneficial effects of the present application are: the traditional blast furnace tuyere state monitoring cannot realize the automatic adjustment function of the camera parameters based on the brightness change of the tuyere image, resulting in the low quality of the photographed tuyere image, and the problem of over-brightness or over-darkness is easy to occur, and the maximum influencing factor of the blast furnace heat system cannot be quantitatively judged, and the blast furnace lower part cannot be directly guided and the quality of molten iron cannot be directly optimized, the method can effectively solve the brightness overexposure problem in the blast furnace tuyere image monitoring, and realize the automatic adjustment of the tuyere image; at the same time, the tuyere monitoring result is combined with the blast furnace smelting process parameters to effectively analyze the maximum influencing factor of the blast furnace lower part heat system, optimize the blast furnace lower part regulation, and improve the blast furnace control level. DETAILED DESCRIPTION

[0013] The blast furnace lower part heat system maximum influencing factor analysis method based on the tuyere monitoring provided by the specific embodiment comprises the following steps in sequence:

[0014] S1, collecting real-time blast furnace tuyere image, judging whether the blast furnace tuyere image is in overexposure or overdark state, and adopting the pre-calibration method to adjust the exposure time and gain for the blast furnace tuyere image in overexposure or overdark state until the brightness of the blast furnace tuyere image meets the standard.

[0015] The blast furnace tuyere monitoring system comprises a high-frame-rate camera based on FPGA, a telecentric lens, a power supply system, a data transmission system, an image storage and processing system, and a high-temperature-resistant dustproof protective kit. The telecentric lens is connected to the high-frame-rate camera based on FPGA. The high-frame-rate camera based on FPGA is connected to the data transmission system. The data transmission system is connected to the image storage and processing system. The power supply system is used to supply power to the blast furnace tuyere monitoring system. The high-temperature-resistant dustproof protective kit is used for high-temperature-resistant and dustproof protection of the blast furnace tuyere monitoring system. The high-frame-rate camera based on FPGA can realize the functions of image edge extraction and brightness average value analysis in the camera body and make corresponding judgments to realize intelligent collection of tuyere images and tuyere coal injection amounts based on different brightness. The high-frame-rate camera based on FPGA shoots the blast furnace tuyere image and uploads it to the image storage and processing system through the data transmission system. The image storage and processing system is used for processing and analyzing the tuyere image. The detailed processing and analyzing steps are in steps S2-S3. The high-temperature-resistant dustproof protective kit is wrapped outside the electronic components of the entire blast furnace tuyere monitoring system to protect the internal electronic components from the influence of high temperature and coal powder of the blast furnace tuyere, so as to avoid inaccurate analysis results.

[0016] The standard for judging whether the blast furnace tuyere image is in overexposure or overdark state in this embodiment is that when the FPGA-based high-frame-rate camera judges that the gray value of the real-time collected blast furnace tuyere image is >200, it is considered that the blast furnace tuyere image is in overexposure state, and when the FPGA-based high-frame-rate camera judges that the gray value of the real-time collected blast furnace tuyere image is <100, it is considered that the blast furnace tuyere image is in overdark state.

[0017] The pre-calibration method of the blast furnace tuyere brightness in this embodiment is specifically that during the blast furnace tuyere stop-blast process, the blast furnace smelting process parameters such as different coal injection amount, blast volume, blast temperature, burden descending speed, blast furnace tapping amount, smelting period, and oxygen enrichment rate are changed to obtain the change rule of the blast furnace tuyere image brightness, and the exposure time and gain of the FPGA-based high-frame-rate camera are adjusted to obtain the brightness of the best blast furnace tuyere image effective area under different blast furnace smelting process parameters.

[0018] The judgment method of whether the blast furnace tuyere image brightness meets the standard is that the average brightness of the blast furnace tuyere image in the last 5 seconds is within the range of 150±50, that is, it is considered that the camera parameter adjustment is in place, and the blast furnace tuyere image brightness meets the standard.

[0019] S2, take the blast furnace tuyere image with a length of S=600 seconds, and take the average brightness value of each group of the blast furnace tuyere image as the tuyere brightness of the group every s=5 seconds, and take the average value of each blast furnace smelting process parameter at the same fixed time s=5 seconds, and the sample data is composed of the tuyere brightness and the average value of all blast furnace smelting process parameters, and the sample number is S / s=120;

[0020] S3, the blast furnace smelting process parameters include blast furnace blast volume XB, blast temperature XBT, burden descending speed XBD, blast furnace tapping amount XI, smelting period XS, coal injection amount XPCI, and oxygen enrichment rate XB0, the tuyere brightness is recorded as L, the average value of the blast furnace smelting process parameters is recorded as X={XB, XBT, XBD, XI, XS, XPCI, XBO} 7×120 , and the tuyere brightness is recorded as Y={L} 1×120 , and then a partial least squares regression equation is constructed, the process is as follows:

[0021] Firstly, the base w k and c k of the new space of the kth step are calculated, (k=1, …120), wherein w k is the residual matrix E k-1 after the principal component extraction of the (k-1)th step T F k-1 F k-1T E k-1 corresponding to the maximum eigenvalue of the matrix E k is the residual matrix F k-1 T E k-1 E k-1 T F k-1 corresponding to the maximum eigenvalue of the matrix E k-1 extracts components t k extracts components u k-1 in the residual matrix F k Under the constraint condition of (w k ) 2 = 1 and (c k ) 2 = 1, the covariance cov(t k , u k ) is solved → max, and then the eigenvector corresponding to the eigenvalue is solved to obtain the eigenvector w k and c k corresponding to the principal axis.

[0022] Then the first component t1 and the residual matrix E1,

[0023] t1 = E0·w1 (1)

[0024]

[0025] In formula (2)

[0026] Then the second component t2 to the seventh component t7 and the residual matrix E2 to the residual matrix E7 are solved in turn until the marginal contribution of the extracted principal component to the model prediction effect is not significant.

[0027] The partial least squares regression equation is constructed: such as formula (3)

[0028]

[0029] h takes a value of 1-7.

[0030] The variable projection importance index I j is used to measure the importance of the influence of the blast furnace smelting process parameters on the tuyere brightness, and the larger the variable projection importance index I j , the higher the importance of the influence of the blast furnace smelting process parameters on the tuyere brightness, that is, the blast furnace smelting process parameter that has the greatest impact on the blast furnace thermal system in this period.

[0031]

[0032] While embodiments of the application have been shown and described, it is to be understood that the application is not limited to the details of the embodiments described, since the scope of the expenditure will be defined with respect to the appended claims and equivalents thereof.

Claims

1. A blast furnace lower part thermal system maximum influencing factor analysis method based on tuyere monitoring, characterized in that, Comprise the following steps in turn: S1, collect real-time blast furnace tuyere image, judge whether the blast furnace tuyere image is in overexposure or overdark state, adopt pre-calibration method to adjust exposure time and gain for the blast furnace tuyere image in overexposure or overdark state until the blast furnace tuyere image brightness meets the standard; S2, take a period of time S blast furnace tuyere image, and take a group of average brightness value of the blast furnace tuyere image as the tuyere brightness of the group every fixed time s, take the average value of each blast furnace smelting process parameter at the same fixed time s, and the sample data is composed of the tuyere brightness and the average value of all blast furnace smelting process parameters, and the sample number is S / s; S3, the blast furnace smelting process parameters include blast furnace blast volume XB, blast temperature XBT, burden descending speed XBD, blast furnace iron output XI, smelting period XS, coal injection volume XPCI, oxygen enrichment rate XBO, the wind port brightness is recorded as L, the average value of the blast furnace smelting process parameters is recorded as X={XB, XBT, XBD, XI, XS, XPCI, XBO} 7×S / s , the wind port brightness is recorded as Y={L} 1×S / s , after the standardization of X and Y two matrixes, new data matrixes E0, F0 are obtained, then a partial least squares regression equation is constructed, a variable projection importance index is used to measure the importance of the influence of the blast furnace smelting process parameters on the wind port brightness, the greater the variable projection importance index, the higher the importance of the influence of the blast furnace smelting process parameters on the wind port brightness, namely the blast furnace smelting process parameter that has the greatest influence on the blast furnace heat system in the period; The blast furnace tuyere image is extracted by the blast furnace tuyere monitoring system, the blast furnace tuyere monitoring system comprises an FPGA-based high frame rate camera, a telecentric lens, a power supply system, a data transmission system, an image storage and processing system, a high-temperature-resistant dustproof protective kit, the telecentric lens is connected with the FPGA-based high frame rate camera, the FPGA-based high frame rate camera is connected with the data transmission system, the data transmission system is connected with the image storage and processing system, the power supply system is used for power supply of the blast furnace tuyere monitoring system, the high-temperature-resistant dustproof protective kit is used for high-temperature-resistant and dustproof protection of the blast furnace tuyere monitoring system, the FPGA-based high frame rate camera shoots the blast furnace tuyere image and uploads it to the image storage and processing system through the data transmission system.

2. The blast furnace lower part thermal system maximum influencing factor analysis method based on tuyere monitoring according to claim 1, characterized in that: When the gray value of the collected real-time blast furnace tuyere image is in the range of >200 or <100, the FPGA-based high frame rate camera considers that the blast furnace image is in overexposure or overdark state.

3. The blast furnace lower part thermal system maximum influencing factor analysis method based on tuyere monitoring according to claim 2, characterized in that: In step 1, for the blast furnace tuyere image in overexposure or overdark state, pre-calibration method is adopted to adjust exposure time and gain, until the average brightness of the blast furnace tuyere image in 5 seconds is in the range of 150±50, then the blast furnace tuyere image brightness meets the standard.

Images