CN114838830A - Method, device and system for detecting temperature of iron-donating process section - Google Patents
Method, device and system for detecting temperature of iron-donating process section Download PDFInfo
- Publication number
- CN114838830A CN114838830A CN202210434429.6A CN202210434429A CN114838830A CN 114838830 A CN114838830 A CN 114838830A CN 202210434429 A CN202210434429 A CN 202210434429A CN 114838830 A CN114838830 A CN 114838830A
- Authority
- CN
- China
- Prior art keywords
- temperature
- temperature detection
- detection point
- emissivity
- determining
- 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
- 238000000034 method Methods 0.000 title claims abstract description 67
- 238000001514 detection method Methods 0.000 claims abstract description 184
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 claims abstract description 180
- 229910052742 iron Inorganic materials 0.000 claims abstract description 90
- 238000012937 correction Methods 0.000 claims abstract description 86
- 239000013598 vector Substances 0.000 claims description 27
- 238000003062 neural network model Methods 0.000 claims description 11
- 238000013527 convolutional neural network Methods 0.000 claims description 7
- 230000000903 blocking effect Effects 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 claims description 3
- 125000004122 cyclic group Chemical group 0.000 claims description 2
- 229910000519 Ferrosilicon Inorganic materials 0.000 claims 1
- 238000009529 body temperature measurement Methods 0.000 description 9
- 239000002893 slag Substances 0.000 description 9
- 230000003595 spectral effect Effects 0.000 description 8
- 230000009286 beneficial effect Effects 0.000 description 6
- 239000000428 dust Substances 0.000 description 6
- 238000010079 rubber tapping Methods 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 5
- 238000013500 data storage Methods 0.000 description 5
- 229910000831 Steel Inorganic materials 0.000 description 4
- 230000017525 heat dissipation Effects 0.000 description 4
- 238000013178 mathematical model Methods 0.000 description 4
- 239000010959 steel Substances 0.000 description 4
- 238000012935 Averaging Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 238000001228 spectrum Methods 0.000 description 3
- 238000003723 Smelting Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000009628 steelmaking Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 238000003556 assay Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000011112 process operation Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000003908 quality control method Methods 0.000 description 1
- 238000010791 quenching Methods 0.000 description 1
- 230000000171 quenching effect Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000000087 stabilizing effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J5/00—Radiation pyrometry, e.g. infrared or optical thermometry
- G01J5/48—Thermography; Techniques using wholly visual means
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J5/00—Radiation pyrometry, e.g. infrared or optical thermometry
- G01J5/0037—Radiation pyrometry, e.g. infrared or optical thermometry for sensing the heat emitted by liquids
- G01J5/004—Radiation pyrometry, e.g. infrared or optical thermometry for sensing the heat emitted by liquids by molten metals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J5/00—Radiation pyrometry, e.g. infrared or optical thermometry
- G01J2005/0077—Imaging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
-
- 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
- Y02P10/00—Technologies related to metal processing
- Y02P10/20—Recycling
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Radiation Pyrometers (AREA)
Abstract
The invention discloses a method, a device and a system for detecting the temperature of an iron awarding process section. The method for detecting the temperature of the iron teaching process section comprises the following steps: determining the temperature of a skimming device temperature detection point, the temperature of a dragon channel temperature detection point, the temperature of a swinging nozzle temperature detection point and the temperature of an iron ladle temperature detection point; determining the temperature of any one of a skimmer temperature detection point, a dragon channel temperature detection point, a swinging nozzle temperature detection point or an iron ladle temperature detection point comprises: acquiring a thermal image and emissivity of a temperature detection point; determining an emissivity correction factor from the thermal image; correcting the emissivity by adopting an emissivity correction coefficient to obtain corrected emissivity; and determining the temperature of the temperature detection point by correcting the emissivity.
Description
Technical Field
The embodiment of the invention relates to an iron teaching process technology, in particular to a method, a device and a system for detecting the temperature of an iron teaching process section.
Background
Blast furnace ironmaking is a key process in the steel flow, and high-quality molten iron is a precondition for smelting high-purity high-quality steel and a basis for improving the quality of steel products.
The temperature monitoring of molten iron is a key step of molten iron quality control, and generally, the temperature of molten iron during tapping is influenced by the following factors: heat dissipation during tapping (the heat dissipation is affected by the size of the iron flow, the length of the iron runner, the iron supply time, the temperature of the iron ladle before the iron and the like); heat dissipation during the process of adding scrap steel into the iron ladle (from the time of running steel making after the iron ladle is discharged); and (4) heat dissipation of the iron ladle in the waiting furnace entering time of steel making.
At present, temperature detection during tapping is mainly realized by the following modes: taking a molten iron sample in the main ditch through manpower or a mechanical arm for assay, and then obtaining a molten iron component index to obtain the temperature of the molten iron; or measuring the temperature of the molten iron at the main channel position by using a portable thermocouple by a human or a mechanical hand. The intermittent sampling operation is adopted when the manual sampling temperature measurement is adopted, 2-3 parts of iron samples are generally randomly taken after the blast furnace is tapped, and the sampling interval time is generally 30-50 minutes.
Based on the above, in the prior art, it is difficult to continuously detect the temperature of the molten iron during the process of receiving iron, and the trend of the change of the molten iron energy during the tapping process of the blast furnace cannot be accurately mastered, which is not favorable for better guiding the process operation of the blast furnace.
Disclosure of Invention
The invention provides a method, a device and a system for detecting the temperature of an iron-donating process section, which aim to achieve the purposes of continuously detecting the temperature of molten iron and improving the temperature measurement precision.
In a first aspect, an embodiment of the present invention provides a method for detecting a temperature of an iron-donating process segment, including:
determining the temperature of a skimming device temperature detection point, the temperature of a dragon channel temperature detection point, the temperature of a swinging nozzle temperature detection point and the temperature of an iron ladle temperature detection point;
determining the temperature of any one of a skimmer temperature detection point, a dragon channel temperature detection point, a swinging nozzle temperature detection point or an iron ladle temperature detection point comprises:
acquiring a thermal image and emissivity of a temperature detection point;
determining an emissivity correction factor from the thermal image;
correcting the emissivity by adopting the emissivity correction coefficient to obtain corrected emissivity;
and determining the temperature of the temperature detection point through the corrected emissivity.
Optionally, acquiring the thermal image of the temperature detection point comprises:
acquiring an original thermal image of a temperature detection point, and eliminating blocking interference elements in the original thermal image to obtain the thermal image.
Optionally, determining an emissivity correction factor according to the thermal image comprises:
taking the thermal image as input, and extracting image features of the thermal image by adopting a first model to obtain an image feature vector;
and inputting the image characteristic vector, and determining the emissivity correction coefficient by adopting a second model.
Optionally, inputting the thermal image, and extracting image features of the thermal image by using the first model to obtain an image feature vector includes:
acquiring a thermal image sequence, and sequentially extracting the image characteristics of each thermal image by adopting the first model to generate an image characteristic sequence;
inputting the image feature vector, and determining the emissivity correction coefficient by using a second model comprises:
sequentially determining a correction coefficient of each image feature through the second model by adopting the image feature sequence to generate a correction coefficient sequence;
further comprising:
and determining the emissivity correction coefficient by adopting a third model according to the correction coefficient sequence.
Optionally, the first model comprises a convolutional neural network model, and the second model comprises a cyclic neural network model.
Optionally, the third model comprises a support vector regression model.
Optionally, determining the temperature of any one of the temperature detection points includes:
acquiring thermal images of at least two wave bands of a temperature detection point, and acquiring emissivity of the temperature detection point;
respectively determining single-band correction coefficients corresponding to the thermal images of each band, and determining emissivity correction coefficients by using all the single-band correction coefficients;
correcting the emissivity by adopting the emissivity correction coefficient to obtain corrected emissivity;
and determining the temperature of the temperature detection point through the corrected emissivity.
Optionally, the temperature of the skimmer temperature detection point, the temperature of the dragon channel temperature detection point, the temperature of the swinging nozzle temperature detection point and the temperature of the iron ladle temperature detection point are used as a basis for adjusting the flow of molten iron in the iron supply process section.
In a second aspect, an embodiment of the present invention further provides an apparatus for detecting a temperature of an iron-feeding process section, including a temperature detecting unit, where the temperature detecting unit is configured to:
determining the temperature of a skimming device temperature detection point, the temperature of a dragon channel temperature detection point, the temperature of a swinging nozzle temperature detection point and the temperature of an iron ladle temperature detection point;
determining the temperature of any one of a skimmer temperature detection point, a dragon channel temperature detection point, a swinging nozzle temperature detection point or an iron ladle temperature detection point comprises:
acquiring a thermal image and emissivity of a temperature detection point;
determining an emissivity correction factor from the thermal image;
correcting the emissivity by adopting the emissivity correction coefficient to obtain a corrected emission coefficient;
and determining the temperature of the temperature detection point through the corrected emission coefficient.
In a third aspect, an embodiment of the present invention further provides a system for detecting a temperature of an iron teaching process segment, including a controller, where the controller is configured with an executable program, and the executable program is used to implement the method for detecting a temperature of an iron teaching process segment described in the embodiment of the present invention when the executable program runs.
Compared with the prior art, the invention has the beneficial effects that: the invention provides a temperature detection method for an iron awarding process section, which comprises the steps of acquiring a thermal image and emissivity of a selected temperature detection point, determining an emissivity correction coefficient through the thermal image based on a neural network model, correcting the emissivity through the correction coefficient, and determining the temperature of the temperature detection point through the emissivity, so that the automatic detection of the temperature can be realized, and the influence of factors such as environmental influence on the temperature measurement precision can be reduced and the temperature measurement precision can be improved based on the emissivity correction coefficient determined through the neural network model; in addition, the selected temperature detection points in the method comprise a skimmer temperature detection point, a dragon channel temperature detection point, a swinging nozzle temperature detection point and an iron ladle temperature detection point, the temperature detection points cover main nodes of an iron teaching process section, and the temperature continuous detection aiming at the whole iron tapping process can be realized.
Drawings
FIG. 1 is a schematic view of an iron-donating process segment in an embodiment;
FIG. 2 is a flowchart of a temperature detection method in the embodiment;
FIG. 3 is a flow chart of another temperature sensing method in an embodiment;
FIG. 4 is a flow chart of a further method for temperature detection in an embodiment;
FIG. 5 is a flow chart of a further method for temperature detection in an embodiment;
FIG. 6 is a schematic diagram of a temperature detection system of the iron feeding process section in the embodiment.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
Example one
FIG. 1 is a schematic diagram of an example of an iron-donating process segment, and referring to FIG. 1, in this example, the iron-donating process segment mainly includes a main channel segment and a dragon channel segment. A skimmer is arranged in the main channel section, a swing nozzle is arranged in the dragon channel section, and iron ladles are arranged on two sides of the swing nozzle.
During smelting operation, high-temperature slag iron in the blast furnace flows out from an iron notch, firstly passes through the main channel, slag iron mixture completes slag iron separation in the main channel according to the characteristic of large density difference, and a slag discharging port on the upper layer of the slag surface flows into the slag channel under the obstruction of the slag skimmer and forms water slag through water quenching;
the high-density molten iron flows into the dragon channel along the lower channel of the lower slag skimmer, then flows into the iron ladle through the swinging nozzle (the swinging nozzle can obliquely swing left and right according to the position condition of the iron ladle), and then flows into the iron ladle through the swinging nozzle, and after the iron ladle on one side is filled with the expected weight of the molten iron, the swinging nozzle is moved to the iron ladle on the other side, so that the continuous iron supply process is completed.
In this embodiment, a temperature detection point is respectively set at the skimmer, the dragon channel, the swinging nozzle and the ladle, and the temperature of each temperature detection point (i.e. the temperature of the skimmer temperature detection point, the temperature of the dragon channel temperature detection point, the temperature of the swinging nozzle temperature detection point and the temperature of the ladle temperature detection point) is determined to realize continuous detection of the temperature of the molten iron flowing through the iron feeding process section.
For example, in the present embodiment, the temperature detection points of the skimmer, the dragon groove, the swinging nozzle and the ladle position can be determined as follows:
and respectively determining the first point with the lowest dust concentration in the area where the skimmer is located, the area where the dragon ditch is located, the area where the swinging nozzle is located and the area where the iron ladle is located, and taking the points as the temperature detection point of the skimmer, the temperature detection point of the dragon ditch, the temperature detection point of the swinging nozzle and the temperature detection point of the iron ladle.
Illustratively, in a preferred scheme, the temperature detection points of the skimmer, the dragon ditch, the swinging nozzle and the ladle position are determined in the following way:
taking the rear through hole of the skimmer as a starting point, taking the maximum liquid level allowed by the rear through hole as an end point in the ladle, and setting N temporary temperature detection points at uniform intervals on a path through which molten iron flows, wherein the N temporary temperature detection points are marked as N 1 ～N n ；
Manually measuring the temperature of each temporary temperature detection point for multiple times by using a portable thermometer to obtain the average temperature of each temporary temperature detection point, and recording the average temperature as T 1 ～T n ；
Using the mean temperature T 1 And the average temperature T n Determining the maximum loss amount of the temperature when the molten iron flows, and recording the maximum loss amount of the temperature as T s Then T is s ＝T n -T 1 ；
Taking the position of the rear through hole of the skimmer as a skimmer temperature detection point;
self-average temperature T 1 Determining and averaging the temperature T 1 Is equal to T s Average temperature T of/4 x1 Will be related to the average temperature T x1 The corresponding temporary temperature detection point is used as a dragon ditch temperature detection point;
determining and averaging the temperature T x1 Is equal to T s Average temperature T of/4 x2 Will be related to the average temperature T x2 The corresponding temporary temperature detection point is used as a swinging nozzle temperature detection point;
determining and averaging the temperature T 2 Is equal to T s Average temperature T of/4 x3 Will be related to the average temperature T x3 And the corresponding temporary temperature detection point is used as an iron ladle temperature detection point.
Exemplary embodiments of the inventionIn the above solution, if the average temperature and the average temperature T of the last temporary temperature detection point located in the dragon ditch section are determined 1 Is greater than T s And/4, adjusting the structure of the dragon ditch, comprising:
the method comprises the steps of increasing the inclination angle of the end of the dragon ditch close to the skimmer by utilizing the time of stopping the ditch for repairing, increasing the horizontal groove at the tail end of the dragon ditch by adopting the method of adjusting the thickness of a pouring material layer (the purpose of increasing the horizontal groove is to reduce the overhigh impact force on the swinging nozzle caused by the overhigh flow speed of molten iron and ensure the constant angle between the starting point and the terminal point of the dragon ditch), reducing the length of the inclined section of the end of the dragon ditch close to the skimmer until the average temperature and the average temperature T of the last temporary temperature detection point of the dragon ditch section 1 Is approximately equal to T s /4；
Wherein, if the inclination angle increases, the longer the horizontal groove, and the inclination angle increases by 1 degree, the length of the horizontal groove increases by 100mm on the basis of the original length.
In the above scheme, for example, if the dragon ditch structure is adjusted, the dragon ditch temperature detection point, the swing mouth temperature detection point, and the ladle temperature detection point need to be determined again.
For example, in this embodiment, the method adopted when determining the temperature of each temperature detection point is the same, fig. 2 is a flowchart of the temperature detection method in the embodiment, and referring to fig. 2, determining the temperature of any one of the skimmer temperature detection point, the dragon ditch temperature detection point, the swinging nozzle temperature detection point, or the ladle temperature detection point includes:
s101, acquiring thermal images and emissivity of temperature detection points.
For example, in this embodiment, a spectral thermometer is disposed at the temperature detection point, and an image generated by the spectral thermometer is used as a thermal image of the temperature detection point.
For example, in this embodiment, a manner of obtaining the emissivity is not specifically limited, and for example, the emissivity of the position of the temperature detection point is determined by using a reflectivity method, a multi-wavelength method, and the like.
And S102, determining an emissivity correction coefficient according to the thermal image.
In this embodiment, the neural network model may be used to determine the emissivity correction coefficient, and when the neural network model is used, the thermal image is used as an input and the emissivity correction coefficient is used as an output (or one of the output vectors of the neural network model).
For example, in this embodiment, the type and structure of the neural network model used are not specifically limited, and for example, neural network models such as BP, CNN, RNN, and the like may be used.
And S103, correcting the emissivity by adopting an emissivity correction coefficient to obtain the corrected emissivity.
For example, in this embodiment, the product of the emissivity correction coefficient and the emissivity may be used as the corrected emissivity.
And S104, determining the temperature of the temperature detection point by correcting the emissivity.
For example, in this embodiment, a mathematical model representing a relationship between emissivity and temperature may be obtained, and the temperature of the temperature detection point may be determined according to the mathematical model.
For example, the mathematical model may be determined by function fitting using experimentally measured data or data disclosed in the literature, in this embodiment, the manner of the function fitting is not limited, and the mathematical model after fitting may be a linear model, an exponential model, or a function model in other forms.
For example, in this embodiment, in addition to the continuous detection of the temperature of the molten iron flowing through the iron teaching process segment, the temperature of the skimmer temperature detection point, the temperature of the dragon channel temperature detection point, the temperature of the swinging nozzle temperature detection point, and the temperature of the iron ladle temperature detection point may also be used as a basis for adjusting the flow rate of the molten iron in the iron teaching process segment.
For example, in one possible embodiment, the flow rate of molten iron may be adjusted as follows:
determining the temperature difference between two adjacent temperature detection points in the skimming device temperature detection point temperature, the dragon channel temperature detection point temperature, the swinging nozzle temperature detection point temperature and the iron ladle temperature detection point temperature;
if any temperature difference is larger than a set value, the aperture of the tap hole is increased to improve the flow of the molten iron, and the air suction capacity (power) of a dust removal system is reduced to realize stable energy of the molten iron;
if any temperature difference is smaller than a set value, the aperture of the iron opening is reduced to reduce the flow of the molten iron, and the air suction capacity (power) of the dust removal system is improved to realize the energy stability of the molten iron.
For example, the set value may be set empirically, and may be T s /4。
Illustratively, in the scheme, the control of the molten iron flow is carried out based on the temperature of the temperature detection point, so that the blast furnace operation can be accurately guided, and a certain promoting effect can be realized on stabilizing the molten iron quality and reducing the production quality accidents.
The embodiment provides a temperature detection method for an iron teaching process section, wherein a thermal image and emissivity of a selected temperature detection point are obtained, an emissivity correction coefficient is determined through the thermal image based on a neural network model, the emissivity is corrected through the correction coefficient, and then the temperature of the temperature detection point is determined through the emissivity, so that the automatic detection of the temperature can be realized, and the influence of factors such as environmental influence on the temperature measurement precision can be reduced and the temperature measurement precision can be improved based on the emissivity correction coefficient determined through the neural network model; in addition, the selected temperature detection points in the method comprise a skimmer temperature detection point, a dragon channel temperature detection point, a swinging nozzle temperature detection point and an iron ladle temperature detection point, the temperature detection points cover main nodes of an iron teaching process section, and the temperature continuous detection aiming at the whole iron tapping process can be realized.
Fig. 3 is a flow chart of another temperature detection method in the example, and referring to fig. 3, as an implementation, the method may further include:
s201, acquiring the emissivity of the temperature detection point.
S202, acquiring an original thermal image of the temperature detection point, and eliminating blocking interference elements in the original thermal image to obtain a thermal image.
In the present embodiment, the spectral tester generates an image as an original thermal image of the temperature detection point.
In an exemplary embodiment, the method for eliminating the shielding interference element in the original thermal image mainly solves the problem that the gray level of some pixel points is abnormal due to dust shielding in the original thermal image.
For example, in the present solution, the occlusion interference element in the original thermal image can be eliminated to obtain the thermal image as follows:
generating an original image with the same size (the same number of pixels and the same arrangement mode of the pixels) as the original thermal image;
acquiring continuous multi-frame original thermal images, and determining whether the gray level of the same pixel point in an original thermal image sequence is abnormally changed or not aiming at each pixel point in the original thermal images (for example, in a certain frame of original thermal image, the gray level of the same pixel point is suddenly and obviously increased or suddenly and obviously decreased);
if the gray level of a pixel point is abnormally changed, judging that shielding interference elements exist in the original thermal image, fixing the gray level of the pixel point in the original image at the moment, and in a frame of original thermal image, determining the gray level of the pixel point when the gray level is normal;
if the gray level of a pixel point does not change abnormally, fixing the gray level of the pixel point in the original image, and in a frame of original thermal image, fixing the gray level of the pixel point;
and after the gray levels of all the pixel points in the original image are determined, the original image is the thermal image.
For example, if there is dust interference during measurement by the spectrum tester, the gray level of a corresponding pixel point in the original thermal image is usually within a fixed gray level interval, and when the gray level of a certain pixel point changes abnormally, the gray level of the pixel point in the original thermal image sequence that is normal can be determined based on the gray level interval.
For example, the manner of determining the thermal image in this step may also be applied to the scheme shown in fig. 1, instead of the corresponding content in step S101.
And S203, taking the thermal image as input, and extracting the image characteristics of the thermal image by adopting the first model to obtain an image characteristic vector.
And S204, inputting the image characteristic vector, and determining an emissivity correction coefficient by adopting a second model.
With reference to step S203 and step S204, in this embodiment, the emissivity correction coefficient is determined by using the first model and the second model.
In this embodiment, the first model is a Convolutional Neural Network (CNN) model, and the second model is a Neural Network (NN) model.
For example, in the present solution, the model structure of the CNN is set according to requirements, for example, the CNN may be specifically ResNet-18, ResNet-50, and the like.
In an exemplary embodiment, the model structure of the NN is related to the image feature vector and the set model output, where the set model output of the NN is at least a one-dimensional vector, and the one-dimensional vector represents the emissivity correction coefficient.
S205, the emissivity is corrected by adopting the emissivity correction coefficient, and the correction rate emissivity is obtained.
S206, determining the temperature of the temperature detection point by correcting the emissivity.
In the scheme, on the basis of the beneficial effects of the scheme shown in fig. 1, the image generated by the spectrum thermometer is preprocessed, abnormal pixel points caused by dust shielding in the image are eliminated, and the calculation precision of the emissivity correction coefficient can be improved.
Fig. 4 is a flow chart of another temperature detection method in the embodiment, and referring to fig. 4, on the basis of the scheme shown in fig. 3, the method may further include:
s301, acquiring the emissivity of the temperature detection point.
S302, acquiring an original thermal image of the temperature detection point, and eliminating blocking interference elements in the original thermal image to obtain a thermal image.
And S303, acquiring a thermal image sequence, and sequentially extracting the image characteristics of each thermal image by adopting a first model to generate an image characteristic sequence.
Illustratively, in the scheme, the first model adopts a ResNet-18 model, and the input of the ResNet-18 is set as a thermal image, and the output is a 512-dimensional image feature vector.
In the scheme, a continuous specified number (for example, 100 frames) of thermal images are acquired, the image feature vectors of each frame of thermal image are respectively determined through a ResNet-18 model, and a set of all the image feature vectors is used as an image feature sequence.
S304, adopting the image feature sequence, sequentially determining the correction coefficient of each image feature through a second model, and generating a correction coefficient sequence.
In an exemplary embodiment, in the second model, a Recurrent Neural Network (RNN) model is adopted, an input of the RNN is set as an image feature vector, and an output of the RNN is set as a two-dimensional vector, where two elements in the two-dimensional vector are set as a temperature measurement quality coefficient and a correction coefficient, respectively.
In the above scheme, the thermometric quality factor is used to indicate the accuracy of the temperature determination using the corresponding frame thermal image.
For example, in the present scheme, a set of all two-dimensional vectors is used as the correction coefficient sequence.
S305, determining an emissivity correction coefficient by adopting a third model according to the correction coefficient sequence.
In an exemplary embodiment, in the third model, a Support Vector Regression (SVR) model is used, the input of the SVR model is set as a high-dimensional matrix formed by a sequence of correction coefficients, and the output is an emissivity correction coefficient (or an emissivity correction coefficient type, where one emissivity correction coefficient is set to correspond to one emissivity correction coefficient).
S306, the emissivity is corrected by adopting the emissivity correction coefficient, and the corrected emissivity is obtained.
And S307, determining the temperature of the temperature detection point by correcting the emissivity.
In the scheme, the structures of the ResNet-18, RNN and SVR models are not improved, and sample data adopted in training of the models can be obtained through a calibration test.
On the basis of the beneficial effects of the scheme shown in fig. 2, in the scheme, the correction coefficient sequence is determined based on the thermal image sequence, the correction coefficient sequence is adopted, and then the emissivity correction coefficient is determined based on the third model, so that the problem that the emissivity correction coefficient is calculated inaccurately due to accidental factors when the emissivity correction coefficient is determined through a single-frame thermal image can be solved.
Fig. 5 is a flow chart of another temperature detection method in the example, and referring to fig. 5, as an implementation, the method may further include:
s401, the emissivity of the temperature detection point is obtained.
S402, acquiring thermal images of at least two wave bands of the temperature detection point.
For example, in this scheme, a multichannel spectral thermometer may be configured at the temperature detection point, and an image of one channel in the multichannel spectral thermometer is used as a thermal image of one band.
And S403, respectively determining single-band correction coefficients corresponding to the thermal images of each band, and determining emissivity correction coefficients by using all the single-band correction coefficients.
Exemplarily, in this step, it is set that N bands exist, and determining the emissivity correction coefficient specifically includes:
and respectively acquiring the thermal image sequence of each wave band, and sequentially extracting the image characteristics of each thermal image by adopting a first model aiming at the thermal image sequence of each wave band to generate an image characteristic sequence.
For example, in this embodiment, a thermal image sequence is set to include 100 thermal images, and the image feature sequence of the ith band is denoted as { x } i,1 ,…,x i,100 }。
Illustratively, in the scheme, the ResNet-18 model is adopted as the first model, and the image characteristic x is set i,j Is a 512-dimensional image feature vector.
And sequentially determining the correction coefficients of the image feature sequence by the second model aiming at the image feature sequence of each wave band, and generating the correction coefficient sequence.
Illustratively, in the present scheme, the second model adopts an RNN model, and the input of the RNN is set as 512-dimensional image feature x i,j Output as a two-dimensional vector y i,j And setting two elements in the two-dimensional vector as a temperature measurement quality coefficient and a correction coefficient respectively.
Illustratively, in this scenario, the output of the RNN model includes { y } i,1 ,…,y i,100 H, will y i,100 The correction coefficient is used as the correction coefficient of the image characteristic sequence of the ith wave band, and the correction coefficient sequence is { y } based on the correction coefficient 1,100 ,…,y N,100 }。
And determining an emissivity correction coefficient by adopting a third model according to the correction coefficient sequence.
In an exemplary embodiment, the third model adopts an SVR model, and the input of the SVR model is a high-dimensional matrix formed by a correction coefficient sequence, and the output is an emissivity correction coefficient.
S404, adopting the emissivity correction coefficient to correct the emissivity to obtain the corrected emissivity.
S405, determining the temperature of the temperature detection point by correcting the emissivity.
On the basis of the beneficial effects of the scheme shown in fig. 1, in the scheme, a multiband thermal image sequence is obtained, the emissivity correction coefficient is determined based on the multiband thermal image, and the problem that the uncertainty of the calculated emissivity correction coefficient is large due to the fact that the temperature measured setting is not an ideal black body and the surface of the black body has an unknown radiation coefficient can be solved.
Example two
This embodiment provides a award indisputable technology section temperature-detecting device, including the temperature-detecting element, the temperature-detecting element is used for:
determining the temperature of a skimming device temperature detection point, the temperature of a dragon channel temperature detection point, the temperature of a swinging nozzle temperature detection point and the temperature of an iron ladle temperature detection point;
determining the temperature of any one of a skimmer temperature detection point, a dragon channel temperature detection point, a swinging nozzle temperature detection point or an iron ladle temperature detection point comprises:
acquiring a thermal image and emissivity of a temperature detection point;
determining an emissivity correction factor from the thermal image;
correcting the emissivity by adopting an emissivity correction coefficient to obtain a corrected emission coefficient;
the temperature of the temperature detection point is determined by correcting the emission coefficient.
In this embodiment, the specific implementation and beneficial effects of the temperature detection unit are the same as those of the scheme shown in fig. 2, and detailed descriptions are omitted.
For example, in this embodiment, a temperature detection unit may also be configured to implement temperature detection by using the schemes shown in fig. 3, fig. 4, or fig. 5, and details of the specific implementation and beneficial effects are not described again.
EXAMPLE III
FIG. 6 is a schematic diagram of a temperature detection system of the iron feeding process section in an embodiment, and referring to FIG. 6, the system may include: a controller 100 and a plurality of multi-channel spectrum thermometers (1-n).
For example, in this embodiment, the multi-channel spectral thermometers are respectively disposed on the upper skimmer temperature detection point, the dragon groove temperature detection point, the swinging nozzle temperature detection point, and the ladle temperature detection point, specifically, referring to fig. 1, one temperature detection point may be respectively disposed at the skimmer and the dragon groove positions, two temperature detection points may be symmetrically disposed at the swinging nozzle position, and one temperature detection point may be disposed at each ladle position.
For example, in this embodiment, the controller 100 is configured with an executable program, and the executable program is used to implement any one of the methods for detecting the temperature of the iron teaching process segment described in the first embodiment.
Referring to fig. 6, in one possible embodiment, the system may further include a server 200, a data storage device 300, and a local area network device 400.
The multichannel spectral thermometers are connected to the controller 100 through the server 200, and the server 200 is further connected to the data storage device 300 and the local area network device 400, respectively.
In the present embodiment, for example, the controller 100, the server 200, the data storage device 300, the local area network device 400, and the multichannel spectral thermometer form a distributed system;
the controller 100 may be located in a central control room and the server 200, data storage device 300, and local area network device 400 may be located in an area covered by a local area network.
In the present embodiment, the data storage device 300 is configured to store a temperature measurement history; the lan device 400 may be an office computer or the like.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (10)
1. A method for detecting the temperature of an iron-donating process section is characterized by comprising the following steps:
determining the temperature of a skimming device temperature detection point, the temperature of a dragon channel temperature detection point, the temperature of a swinging nozzle temperature detection point and the temperature of an iron ladle temperature detection point;
determining the temperature of any one of a skimmer temperature detection point, a dragon channel temperature detection point, a swinging nozzle temperature detection point or an iron ladle temperature detection point comprises:
acquiring a thermal image and emissivity of a temperature detection point;
determining an emissivity correction factor from the thermal image;
correcting the emissivity by adopting the emissivity correction coefficient to obtain corrected emissivity;
and determining the temperature of the temperature detection point through the corrected emissivity.
2. The method of claim 1, wherein obtaining the thermal image of the temperature detection spot comprises:
acquiring an original thermal image of a temperature detection point, and eliminating blocking interference elements in the original thermal image to obtain the thermal image.
3. The method of claim 1, wherein determining an emissivity correction factor from the thermal image comprises:
taking the thermal image as input, and extracting image features of the thermal image by adopting a first model to obtain an image feature vector;
and inputting the image characteristic vector, and determining the emissivity correction coefficient by adopting a second model.
4. The method of claim 3, wherein inputting the thermal image and using the first model to extract image features of the thermal image to obtain an image feature vector comprises:
acquiring a thermal image sequence, and sequentially extracting the image characteristics of each thermal image by adopting the first model to generate an image characteristic sequence;
inputting the image feature vector, and determining the emissivity correction coefficient by using a second model comprises:
sequentially determining a correction coefficient of each image feature through the second model by adopting the image feature sequence to generate a correction coefficient sequence;
further comprising:
and determining the emissivity correction coefficient by adopting a third model according to the correction coefficient sequence.
5. The ferrosilicon process segment temperature sensing method of claim 4, wherein the first model comprises a convolutional neural network model and the second model comprises a cyclic neural network model.
6. The method of claim 4, wherein the third model comprises a support vector regression model.
7. The method of claim 1, wherein determining the temperature at any one of the temperature detection points comprises:
acquiring thermal images of at least two wave bands of a temperature detection point, and acquiring emissivity of the temperature detection point;
respectively determining single-band correction coefficients corresponding to the thermal images of each band, and determining emissivity correction coefficients by using all the single-band correction coefficients;
correcting the emissivity by adopting the emissivity correction coefficient to obtain corrected emissivity;
and determining the temperature of the temperature detection point through the corrected emissivity.
8. The method as claimed in claim 1, wherein the skimmer temperature detection point temperature, the dragon channel temperature detection point temperature, the swing nozzle temperature detection point temperature, and the ladle temperature detection point temperature are used as a basis for adjusting the flow rate of molten iron in the iron teaching process section.
9. The utility model provides a award indisputable technology section temperature-detecting device which characterized in that, includes the temperature-detecting element, the temperature-detecting element is used for:
determining the temperature of a skimming device temperature detection point, the temperature of a dragon channel temperature detection point, the temperature of a swinging nozzle temperature detection point and the temperature of an iron ladle temperature detection point;
determining the temperature of any one of a skimmer temperature detection point, a dragon channel temperature detection point, a swinging nozzle temperature detection point or an iron ladle temperature detection point comprises:
acquiring a thermal image and emissivity of a temperature detection point;
determining an emissivity correction factor from the thermal image;
correcting the emissivity by adopting the emissivity correction coefficient to obtain a corrected emission coefficient;
and determining the temperature of the temperature detection point through the corrected emission coefficient.
10. A system for detecting the temperature of a donor stage, comprising a controller configured with an executable program that is operable to implement the method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210434429.6A CN114838830A (en) | 2022-04-24 | 2022-04-24 | Method, device and system for detecting temperature of iron-donating process section |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210434429.6A CN114838830A (en) | 2022-04-24 | 2022-04-24 | Method, device and system for detecting temperature of iron-donating process section |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114838830A true CN114838830A (en) | 2022-08-02 |
Family
ID=82566235
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210434429.6A Pending CN114838830A (en) | 2022-04-24 | 2022-04-24 | Method, device and system for detecting temperature of iron-donating process section |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114838830A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117387778A (en) * | 2023-12-11 | 2024-01-12 | 合肥金星智控科技股份有限公司 | Temperature measurement method and device, electronic equipment and storage medium |
-
2022
- 2022-04-24 CN CN202210434429.6A patent/CN114838830A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117387778A (en) * | 2023-12-11 | 2024-01-12 | 合肥金星智控科技股份有限公司 | Temperature measurement method and device, electronic equipment and storage medium |
| CN117387778B (en) * | 2023-12-11 | 2024-04-02 | 合肥金星智控科技股份有限公司 | Temperature measurement method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4580466B2 (en) | Hot metal temperature detection method and blast furnace operation method using the same | |
| CN114838830A (en) | Method, device and system for detecting temperature of iron-donating process section | |
| JP5527180B2 (en) | Converter blowing method and converter blowing system | |
| KR20170094560A (en) | Method and device for predicting, controlling and/or regulating steelworks processes | |
| CA2235499C (en) | Method and apparatus to determine and control the carbon content of steel in a bof vessel | |
| US4133036A (en) | Method and system for monitoring a physical condition of a medium | |
| CN111518980B (en) | Correction method and system for converter end point carbon content prediction model | |
| US4229412A (en) | Apparatus for the determination of bond forms of gases | |
| JP6825711B2 (en) | Molten component estimation device, molten metal component estimation method, and molten metal manufacturing method | |
| WO2022004119A1 (en) | Converter blowing control method and converter blowing control system | |
| CN113113088A (en) | Converter carbon content index evaluation and temperature analysis method based on artificial intelligence | |
| JP6516906B1 (en) | Wind blow calculation method, blow blow calculation program | |
| KR102098457B1 (en) | Monitoring apparatus for outputting amount of molten steel | |
| CN115029491B (en) | Method for quickly measuring upward warping of middle sleeve | |
| RU2813298C1 (en) | Converter purge control method and converter purge control system | |
| JP7156550B2 (en) | Converter blowing control method and converter blowing control system | |
| KR20190076573A (en) | Monitoring apparatus for splash status of molten steel | |
| RU2811549C1 (en) | Converter purge control method and converter purge control system | |
| JPH11246907A (en) | Method for controlling blowing in converter | |
| KR102511007B1 (en) | Apparatus for measuring molten iron temperature of blast furnace | |
| KR100383277B1 (en) | Measurement method of change in blast furnace gas distribution | |
| SU992593A1 (en) | Method for interrupting blasting of oxygen converter with pre-set carbon content | |
| JPH07140007A (en) | Method and apparatus for measuring temperature of molten metal | |
| CN113504725A (en) | Real-time slag state monitoring device of converter | |
| Ren et al. | Experimental Research of Continuous Temperature Measurement for Molten Metal Bath through Bottom‐Blowing Component |
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 |