CN116083676A - Method, device, equipment and system for monitoring converter steelmaking process - Google Patents
Method, device, equipment and system for monitoring converter steelmaking process Download PDFInfo
- Publication number
- CN116083676A CN116083676A CN202111314402.5A CN202111314402A CN116083676A CN 116083676 A CN116083676 A CN 116083676A CN 202111314402 A CN202111314402 A CN 202111314402A CN 116083676 A CN116083676 A CN 116083676A
- Authority
- CN
- China
- Prior art keywords
- neural network
- artificial neural
- converter
- optical
- prediction result
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 66
- 238000009628 steelmaking Methods 0.000 title claims abstract description 50
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000013528 artificial neural network Methods 0.000 claims abstract description 133
- 230000003287 optical effect Effects 0.000 claims abstract description 108
- 230000002159 abnormal effect Effects 0.000 claims abstract description 64
- 238000006243 chemical reaction Methods 0.000 claims abstract description 57
- 230000005540 biological transmission Effects 0.000 claims abstract description 56
- 230000003595 spectral effect Effects 0.000 claims abstract description 50
- 238000004364 calculation method Methods 0.000 claims abstract description 42
- 238000012545 processing Methods 0.000 claims abstract description 24
- 238000004519 manufacturing process Methods 0.000 claims abstract description 16
- 230000005856 abnormality Effects 0.000 claims description 38
- 238000002834 transmittance Methods 0.000 claims description 16
- 230000004913 activation Effects 0.000 claims description 14
- 238000013041 optical simulation Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 7
- 238000000701 chemical imaging Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 229910000831 Steel Inorganic materials 0.000 description 21
- 239000010959 steel Substances 0.000 description 21
- 230000008569 process Effects 0.000 description 11
- 238000007664 blowing Methods 0.000 description 8
- 238000001228 spectrum Methods 0.000 description 7
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 6
- 238000004891 communication Methods 0.000 description 6
- 229910052760 oxygen Inorganic materials 0.000 description 6
- 239000001301 oxygen Substances 0.000 description 6
- 238000003860 storage Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000012806 monitoring device Methods 0.000 description 5
- 239000000126 substance Substances 0.000 description 5
- 230000004044 response Effects 0.000 description 4
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 2
- 238000003723 Smelting Methods 0.000 description 2
- 230000000295 complement effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 229910044991 metal oxide Inorganic materials 0.000 description 2
- 150000004706 metal oxides Chemical class 0.000 description 2
- 239000002994 raw material Substances 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 238000005094 computer simulation Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005502 peroxidation Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21C—PROCESSING OF PIG-IRON, e.g. REFINING, MANUFACTURE OF WROUGHT-IRON OR STEEL; TREATMENT IN MOLTEN STATE OF FERROUS ALLOYS
- C21C5/00—Manufacture of carbon-steel, e.g. plain mild steel, medium carbon steel or cast steel or stainless steel
- C21C5/28—Manufacture of steel in the converter
- C21C5/30—Regulating or controlling the blowing
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21C—PROCESSING OF PIG-IRON, e.g. REFINING, MANUFACTURE OF WROUGHT-IRON OR STEEL; TREATMENT IN MOLTEN STATE OF FERROUS ALLOYS
- C21C2300/00—Process aspects
- C21C2300/06—Modeling of the process, e.g. for control purposes; CII
-
- 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
- Y02P10/25—Process efficiency
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Materials Engineering (AREA)
- Metallurgy (AREA)
- Organic Chemistry (AREA)
- Carbon Steel Or Casting Steel Manufacturing (AREA)
Abstract
The embodiment of the invention provides a method, a device, equipment and a system for monitoring a converter steelmaking process, wherein the method comprises the following steps: in the converter production process, receiving transmission information from an optical artificial neural network chip, wherein the optical artificial neural network chip is used for realizing a calculation strategy of a part of an artificial neural network model trained in advance, the transmission information is obtained by processing an acquired spectral image of the flame of the converter mouth based on the calculation strategy by the optical artificial neural network chip, and the spectral image comprises spectral information; inputting the transmission information received at the current moment into the other part of the artificial neural network model to obtain an abnormal prediction result of the reaction in the converter; and generating monitoring information corresponding to the abnormal prediction result. Thus, accurate, safe and reliable monitoring of the converter steelmaking process is realized.
Description
Technical Field
The invention relates to the technical field of smelting, in particular to a method, a device, equipment and a system for monitoring a converter steelmaking process.
Background
Converter steelmaking is a widely used steelmaking mode, and during steelmaking, raw materials such as molten iron, scrap steel and the like are firstly added into a converter, and then oxygen blowing and smelting are carried out to form molten steel, namely blowing is carried out. The steel-making process is complicated and various due to the different components and the different amounts of raw materials added into the converter and the different reaction degrees in the blowing process. The reaction condition in the steelmaking process in the converter is accurately known, and the method has important effects on producing high-quality steel and ensuring the safety of the steelmaking process. For example, the reaction condition in the converter cannot be known accurately, difficulty is brought to a worker in determining a proper oxygen blowing amount, if the oxygen blowing amount is improper, steel quality may be affected, abnormal conditions such as molten steel splashing and the like may be caused, safety accidents and the like are caused, wherein if the oxygen blowing amount is low, insufficient reaction may be caused, and if the oxygen blowing amount is high, problems such as peroxidation or molten steel splashing may be caused. As another example, if the equipment in the furnace is damaged, abnormal reaction in the converter may occur, and the molten steel may splash, and in severe cases, even the converter explodes, causing safety problems. Therefore, it is necessary to accurately monitor the reaction conditions in the converter during the steelmaking process of the converter so as to make adjustments in time.
At present, the reaction condition in the converter is generally judged according to manual experience and adjusted. However, on the one hand, different operators have different experiences, the standardization is difficult, the quality of the produced steel is uneven, and on the other hand, the manual experience is difficult to accurately reflect the reaction condition in the furnace, so that the safety of the steelmaking process cannot be ensured.
Disclosure of Invention
The embodiment of the invention provides a method, a device, equipment and a system for monitoring a converter steelmaking process, which are used for solving the problems that the safety cannot be ensured due to uneven steel quality caused by different manual experience and inaccurate manual experience in the conventional mode of judging the reaction condition in a converter by manual experience.
In a first aspect, an embodiment of the present invention provides a method for monitoring a steelmaking process of a converter, including:
in the converter production process, receiving transmission information from an optical artificial neural network chip, wherein the optical artificial neural network chip is used for realizing a calculation strategy of a part of an artificial neural network model trained in advance, the transmission information is obtained by processing an acquired spectral image of the flame of the converter mouth based on the calculation strategy by the optical artificial neural network chip, and the spectral image comprises spectral information;
inputting the transmission information received at the current moment into the other part of the artificial neural network model to obtain an abnormal prediction result of the reaction in the converter;
and generating monitoring information corresponding to the abnormal prediction result.
Optionally, a portion of the artificial neural network model includes an input layer and a first fully connected layer;
the optical artificial neural network chip comprises an optical modulation layer and an image sensor;
the light modulation layer is used for modulating incident light from the fire hole flame based on a calculation strategy of the input layer;
the image sensor is used for sensing the optical signal modulated by the optical modulation layer based on the calculation strategy of the first full-connection layer, and processing the sensed optical signal into the transmission information in the form of an electric signal.
Optionally, the method further comprises:
inputting a spectral image sample of the fire hole flame acquired by a spectral imaging chip in advance into a preset artificial neural network model for training to obtain the artificial neural network model, wherein the spectral image sample is marked with an actual result of the reaction in the converter;
performing optical simulation on the light modulation layer, and adjusting parameters of the structure of the simulated light modulation layer to enable the transmittance of the simulated light modulation layer to different wavelength components in incident light to be target transmittance;
wherein the target transmittance is determined based on a connection weight of the input layer to the first fully connected layer;
and the parameters of the structure of the simulated light modulation layer are used for manufacturing the light modulation layer on the image sensor so as to obtain the optical artificial neural network chip.
Optionally, another part of the artificial neural network model includes a second full connection layer and a nonlinear activation layer;
inputting the transmission information received at the current moment into another part of the artificial neural network model to obtain an abnormal prediction result of the reaction in the converter, wherein the method comprises the following steps:
inputting the transmission information received at the current moment into the second full-connection layer to obtain a full-connection result;
and inputting the full connection result to the nonlinear activation layer to obtain the abnormality prediction result.
Optionally, the abnormal prediction result includes an abnormal prediction result corresponding to the current time and/or an abnormal prediction result corresponding to a preset time period after the current time.
Optionally, the abnormal prediction result includes normal reaction or occurrence of a preset type of abnormality.
Optionally, the generating the monitoring information corresponding to the abnormal prediction result includes:
and responding to the abnormality prediction result to generate early warning information corresponding to the abnormality of the preset type, wherein the abnormality of the preset type occurs.
In a second aspect, an embodiment of the present invention provides a monitoring device for a steelmaking process of a converter, including:
the information receiving module is used for receiving transmission information from an optical artificial neural network chip in the converter production process, wherein the optical artificial neural network chip is used for realizing a calculation strategy of a part of an artificial neural network model trained in advance, the transmission information is obtained by processing an acquired spectral image of the flame of the converter mouth based on the calculation strategy by the optical artificial neural network chip, and the spectral image comprises spectral information;
the abnormality prediction module is used for inputting the transmission information received at the current moment into the other part of the artificial neural network model to obtain an abnormality prediction result of the reaction in the converter;
and the information generation module is used for generating monitoring information corresponding to the abnormal prediction result.
In a third aspect, an embodiment of the present invention provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the converter steelmaking process monitoring method as provided in the first aspect when executing the program.
In a fourth aspect, an embodiment of the present invention provides a monitoring system for a steelmaking process of a converter, including:
an optical artificial neural network chip and an electronic device as provided in the third aspect.
In a fifth aspect, embodiments of the present invention provide a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the method as provided by the first aspect.
According to the converter steelmaking process monitoring method, in the converter production process, transmission information from the optical artificial neural network chip is received, the optical artificial neural network chip is used for realizing a calculation strategy of a part of an artificial neural network model trained in advance, the transmission information is obtained by processing a collected spectral image of the converter mouth flame of the converter by the optical artificial neural network chip based on the calculation strategy, the spectral image contains the spectral information, the transmission information received at the current moment is input into the other part of the artificial neural network model, an abnormal prediction result of the reaction in the converter is obtained, monitoring information corresponding to the abnormal prediction result is generated, accurate prediction of the abnormal condition in the converter is automatically achieved, so that adjustment is performed in time, the problems that the quality of steel is uneven due to the difference of the existing artificial experience and the safety cannot be guaranteed due to the inaccuracy of the artificial experience are avoided, and therefore the quality of the steel is improved and the safety is improved. And the calculation strategy of one part of the artificial neural network model is transferred to the optical artificial neural network chip to perform hardware calculation, so that the processing speed is greatly improved, the real-time requirement on monitoring of the converter steelmaking process can be met, the calculation strategy of the other part of the artificial neural network model is calculated through software, and the abnormal prediction result is more accurate. Thus, accurate, safe and reliable monitoring of the converter steelmaking process is realized.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a method for monitoring a steelmaking process of a converter according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an optical artificial neural network chip according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of an application scenario including a monitoring system for a steelmaking process of a converter according to an embodiment of the present invention;
FIG. 4 is a schematic structural view of a monitoring device for a steelmaking process of a converter according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Fig. 1 is a schematic flow chart of a method for monitoring a steelmaking process of a converter according to an embodiment of the present invention, as shown in fig. 1, where the method includes:
The optical artificial neural network chip, that is, the optical neural network chip, can utilize the optical component to realize the calculation of the neural network, and the processing speed of the calculation based on the hardware of the optical component is far greater than the processing speed of computer software. In the step, a calculation strategy of a part of the pre-trained artificial neural network model is arranged in the optical artificial neural network chip, so that the processing speed is greatly improved, and the real-time requirement on monitoring of the converter steelmaking process can be met.
The pixel information of each pixel of the spectral image includes, in addition to spatial information (i.e., coordinate information of a two-dimensional plane), spectral information including intensity information of each wavelength component. The spectrum is a fingerprint of the substance, and therefore, the spectral information of the furnace mouth flame of the converter can reflect the substance in the converter. Since substances in different reaction conditions in the converter are different, spectral information of the flame at the mouth of the converter can further reflect the reaction conditions in the converter.
The method for monitoring the converter steelmaking process provided by the embodiment can be executed by electronic equipment such as a computer or the like or a combination of software and/or hardware in the electronic equipment.
In practical application, when the converter starts to produce, the optical artificial neural network chip can start to process the acquired spectral image of the converter mouth flame of the converter in real time to obtain transmission information, the transmission information is transmitted to the electronic equipment, and the electronic equipment can receive the transmission information from the optical artificial neural network chip.
And 120, inputting the transmission information received at the current moment into the other part of the artificial neural network model to obtain an abnormal prediction result of the reaction in the converter.
The transmission information received at the current moment can reflect the condition of the reaction in the current converter, how the trend of the reaction is, whether the reaction is abnormal or not, so that based on the transmission information received at the current moment, an abnormal prediction result corresponding to the current moment can be predicted and obtained, and also an abnormal prediction result corresponding to a preset time period after the current moment can be predicted and obtained, and the preset time period after the current moment can comprise the next moment of the current moment by way of example. Based on the above, the abnormal prediction result output by the artificial neural network model in the step includes an abnormal prediction result corresponding to the current time and/or an abnormal prediction result corresponding to a preset time period after the current time.
And 130, generating monitoring information corresponding to the abnormal prediction result.
In this embodiment, on the one hand, because the substances in the converter under different reaction conditions are different, and the spectral information contained in the spectral image of the flame at the furnace mouth can reflect the substances in the converter and further can reflect the reaction conditions in the converter, therefore, abnormal prediction can be performed through the artificial neural network model trained in advance based on the spectral image of the flame at the furnace mouth, and monitoring information corresponding to the abnormal prediction result is generated, so that a worker can know whether the abnormal condition occurs in the reaction in the converter through the monitoring information, and the accurate prediction of the abnormal condition in the converter is automatically realized, so that adjustment can be performed in time, the problems that the quality of steel is uneven due to the difference of the existing artificial experience and the safety cannot be ensured due to the inaccuracy of the artificial experience are avoided, and the quality of the steel is improved and the safety is improved. On the other hand, the calculation strategy of a part of the pre-trained artificial neural network model is realized through the optical artificial neural network chip, the optical artificial neural network chip processes the acquired spectral image of the converter mouth flame of the converter based on the calculation strategy to obtain transmission information, namely, the calculation strategy of the part of the artificial neural network model is transferred to the optical artificial neural network chip to perform hardware calculation, the processing speed is greatly improved, the real-time requirement on monitoring of the converter steelmaking process can be met, after the transmission information is received, the transmission information received at the current moment is continuously input into the other part of the artificial neural network model, so that the abnormal prediction result of the reaction in the converter is obtained, namely, the calculation strategy of the other part of the artificial neural network model is calculated through software, and the abnormal prediction result is more accurate. Thus, accurate, safe and reliable monitoring of the converter steelmaking process is realized.
In practical applications, after the monitoring information corresponding to the abnormality prediction result is generated, the monitoring information may be displayed for the operator to view. Specifically, the monitoring information may be displayed on a screen of the electronic device.
For example, the exception prediction result may include a normal reaction, or an occurrence of a preset type of exception. In practical application, the preset type of abnormality can be set according to practical requirements. By way of example, the predetermined type of anomaly may be molten steel splash, or the like. If the abnormal prediction result comprises normal reaction, the adjustment processing is not needed, and if the abnormal prediction result comprises the abnormal of the preset type, the adjustment can be carried out in time. For example, if molten steel splashing occurs, the amount of oxygen blowing or the like may be adjusted.
Based on the above embodiment, the generation of the monitoring information corresponding to the abnormality prediction result may include: and generating early warning information corresponding to the abnormality of the preset type in response to the abnormality prediction result comprising the occurrence of the abnormality of the preset type.
For example, when the abnormality prediction result includes the occurrence of molten steel splash, the early warning information "may be about to occur molten steel splash, please note-! By means of the early warning information displayed, workers can know that molten steel splashing possibly occurs, and accordingly adjustment is timely made.
If the abnormal prediction result includes normal reaction, and correspondingly, generating monitoring information corresponding to the abnormal prediction result, the specific implementation manner may include: and generating prompt information for prompting the normal reaction in the converter in response to the abnormal prediction result including the normal reaction. The prompt information can be, for example, "the reaction in the converter is normal-! ". At this time, the worker knows that the reaction in the converter is normal by the displayed prompt information, and does not need to perform adjustment processing.
If the abnormal prediction result output by the artificial neural network model comprises an abnormal prediction result corresponding to the current moment and/or an abnormal prediction result corresponding to a preset time period after the current moment. Then, when generating the monitoring information corresponding to the abnormality prediction result, specifically may include: and generating monitoring information corresponding to the abnormal prediction result corresponding to the current moment, and/or generating monitoring information corresponding to the abnormal prediction result corresponding to a preset time period after the current moment.
The generating the monitoring information corresponding to the abnormal prediction result corresponding to the current time may specifically include: responding to an abnormality prediction result corresponding to the current time, wherein the abnormality prediction result comprises abnormality of a preset type, and generating early warning information corresponding to the abnormality of the preset type aiming at the current time; responding to the abnormal prediction result corresponding to the current moment to comprise normal reaction, and generating prompt information for prompting normal reaction in the converter aiming at the current moment.
The generating of the monitoring information corresponding to the abnormal prediction result corresponding to the preset time period after the current time may specifically include: responding to an abnormality prediction result corresponding to a preset time period after the current moment, wherein the abnormality prediction result comprises occurrence of an abnormality of a preset type, and generating early warning information corresponding to the abnormality of the preset type aiming at the preset time period after the current moment; and generating prompt information for prompting the normal reaction in the converter according to the preset time period after the current time in response to the abnormal prediction result corresponding to the preset time period after the current time including normal reaction.
Therefore, the working personnel can conveniently know the reaction conditions in the converter at the current moment and in the preset time period after the current moment, and the reaction conditions are more comprehensive.
Based on the above embodiments, a portion of the artificial neural network model may include an input layer and a first fully connected layer. Accordingly, as shown in fig. 2, the optical artificial neural network chip 210 may include an optical modulation layer 211 and an image sensor 212. Wherein the light modulation layer 211 is used for modulating the incident light from the burner flame based on the calculation strategy of the input layer. The image sensor 212 is configured to sense an optical signal modulated by the optical modulation layer 211 based on a calculation policy of the first fully-connected layer, and process the sensed optical signal into transmission information in the form of an electrical signal.
The light modulation layer 211 has different modulation effects on light of different wavelengths, so that the transmittance of light of different wavelengths is different. The light modulation layer 211 may include a plurality of modulation units 2111, and the modulation units 2111 may include an array of modulation holes, and the shape of the modulation holes may be, but is not limited to, circular, cross, triangular, star-shaped, rectangular, or the like.
The image sensor 212 has a photosensitive region, and the light modulation layer 211 is disposed on the photosensitive region of the image sensor 212. The image sensor 212 can sense the optical signal modulated by the optical modulation layer 211 through the photosensitive area, and process the sensed optical signal to obtain transmission information in the form of an electrical signal, and transmit the transmission information to the electronic device in the form of an electrical signal.
Because the optical artificial neural network chip can process optical signals, the input layer of the artificial neural network model can be directly arranged on the optical artificial neural network chip to process incident light from fire hole flames, namely, the collected spectral images of the fire hole flames are processed, in addition, the first full-connection layer of the artificial neural network model can be arranged on the optical artificial neural network chip to realize dimension reduction processing through preliminary full connection, and therefore, the calculation of follow-up software is reduced through the pretreatment of the optical artificial neural network chip, and the processing speed is improved.
Based on any one of the above embodiments, the method for monitoring a steelmaking process of a converter may further include:
firstly, a spectral image sample of the fire hole flame acquired by a spectral imaging chip in advance is input into a preset artificial neural network model for training, so that an artificial neural network model is obtained, and an actual result of the reaction in the converter is marked in the spectral image sample.
And then, carrying out optical simulation on the optical modulation layer, and adjusting parameters of the structure of the simulated optical modulation layer to enable the transmittance of the simulated optical modulation layer to different wavelength components in incident light to be the target transmittance. Wherein the target transmittance is determined based on a connection weight of the input layer to the first fully connected layer. And the parameters of the structure of the simulated light modulation layer are used for manufacturing the light modulation layer on the image sensor so as to obtain the optical artificial neural network chip.
The spectrum imaging chip can collect and output an original spectrum image for incident light. In practical application, an existing spectral imaging chip can be directly utilized to collect spectral image samples of the flame at the furnace mouth, then, an actual result of the reaction in the converter corresponding to each spectral image sample is collected, the actual result of the reaction in the converter is marked in the spectral image samples, for example, the actual result comprises an actual result corresponding to the current moment, and an actual result corresponding to a preset time period (for example, a moment next to the current moment) after the current moment, and the actual result can comprise normal reaction or abnormal reaction of a preset type.
An artificial neural network model can be built in advance, namely a preset artificial neural network model is obtained, spectrum image samples at all moments in the production process of the converter are sequentially input into the preset artificial neural network model for training until convergence, and the trained artificial neural network model is obtained. The artificial neural network model may include, for example, an input layer, a first fully-connected layer, a second fully-connected layer, and a nonlinear activation layer. And transferring the calculation strategies of the input layer and the first full-connection layer of the artificial neural network model into an optical artificial neural network chip, wherein the optical modulation layer of the optical artificial neural network chip corresponds to the input layer, and the image sensor corresponds to the first full-connection layer. After the image sensor is selected, optical simulation can be performed on the structure of the optical modulation layer through a computer simulation technology, and parameters of the structure of the simulated optical modulation layer are adjusted, so that the transmittance of the optical modulation layer to different wavelength components in incident light is a target transmittance, and the target transmittance can be the connection weight of the input layer to the first full-connection layer. And finally, adopting the parameters of the structure of the regulated simulated light modulation layer to manufacture the light modulation layer on the selected image sensor.
In this embodiment, after the training of the artificial neural network model is completed, parameters of the structure of the optical modulation layer are obtained in an optical simulation manner, so that the optical artificial neural network chip is obtained more rapidly and accurately.
In practice, a complementary metal oxide semiconductor (Complementary Metal Oxide Semiconductor, CMOS) flow sheet process may be used to fabricate the light modulating layer on the wafer level image sensor, thereby implementing monolithically integrated optical artificial neural network chips. The CMOS flow sheet process is a very mature process, greatly reduces the volume, the power consumption and the cost of the chip, can realize large-scale application, and has better quality of the manufactured optical artificial neural network chip, thereby ensuring the processing speed and the processing precision of the optical artificial neural network chip.
Based on any of the above embodiments, another portion of the artificial neural network model may include a second fully connected layer and a nonlinear activation layer. Correspondingly, the transmission information received at the current moment is input into the other part of the artificial neural network model to obtain an abnormal prediction result of the reaction in the converter, and the specific implementation mode of the method can comprise the following steps: the transmission information received at the current moment is input to a second full-connection layer, and a full-connection result is obtained; and inputting the full connection result into a nonlinear activation layer to obtain an abnormal prediction result.
Here, the second fully connected layer further achieves dimension reduction. The nonlinear activation layer may map the full connection result of the second full connection layer to a preset abnormal prediction result. In this embodiment, after full connection and nonlinear activation are performed on the transmission information, an anomaly prediction result can be obtained quickly.
The method for monitoring the steelmaking process of the converter provided by the embodiment of the invention is described in more detail below through a specific application scene.
The monitoring method for the steelmaking process of the converter of the embodiment is implemented based on a steelmaking process monitoring system of the converter, and in the application scenario shown in fig. 3, the steelmaking process monitoring system of the converter includes an optical artificial neural network chip 210, a transmission line 220 and a monitoring computer 230.
The optical artificial neural network chip 210 includes the optical modulation layer 211 and the image sensor 212, and the specific structure of the optical artificial neural network chip 210 may refer to the above related embodiments, which are not described herein. The optical artificial neural network chip 210 is capable of implementing a computational strategy that is part of a pre-trained artificial neural network model. Wherein the pre-trained artificial neural network model may include an input layer, a first fully-connected layer, a second fully-connected layer, and a nonlinear activation layer. The optical artificial neural network chip 210 corresponds to a portion of an input layer and a first full connection layer of the artificial neural network model.
The monitoring computer 230 is provided with a spectrum processing program and a real-time display program. The second fully-connected layer and the nonlinear activation layer of the pre-trained artificial neural network model are arranged in a spectrum processing program.
The optical artificial neural network chip 210 is disposed at a safety distance of one side of the shaft of the converter 240, which may be set according to actual requirements, and which may be at least 20m, for example. The surface of the light modulation layer of the optical artificial neural network chip 210 faces the furnace mouth flame 241 of the converter 240. Based on this, the incident light 242 from the furnace mouth flame 241 enters the optical artificial neural network chip 210, is modulated by the optical modulation layer 211 of the optical artificial neural network chip 210, and the modulated optical signal is sensed by the image sensor 212, and is processed into transmission information in the form of an electrical signal, the transmission information is transmitted to the monitoring computer 230 through the transmission line 220, the processor in the monitoring computer 230 performs full-connection and nonlinear activation on the electrical signal by executing a spectrum processing program, so as to obtain an abnormal prediction result of the reaction in the converter, and generates monitoring information corresponding to the abnormal prediction result, and the monitoring information is displayed by executing a real-time display program. The abnormal prediction result may be the occurrence of molten steel splash, etc.
According to the embodiment, the optical artificial neural network chip is utilized to acquire the spectral image, the real-time monitoring of the reaction in the converter steelmaking process is realized through rapid and accurate calculation, the abnormal conditions such as molten steel splashing and the like are prejudged, and the steelmaking operation of workers is guided, so that the quality of steel is improved, and the safe and reliable operation of the converter is ensured. The optical artificial neural network chip adopted by the scheme can be produced by utilizing the existing CMOS flow sheet process, so that the volume, the power consumption and the cost of the chip are greatly reduced, and the large-scale application can be realized.
The converter steelmaking process monitoring device provided by the invention is described below, and the converter steelmaking process monitoring device described below and the converter steelmaking process monitoring method described above can be correspondingly referred to each other.
Fig. 4 is a schematic structural diagram of a monitoring device for a steelmaking process of a converter according to an embodiment of the present invention, where, as shown in fig. 4, the device includes an information receiving module 410, an anomaly prediction module 420, and an information generating module 430;
the information receiving module 410 is configured to receive transmission information from an optical artificial neural network chip during a converter production process, where the optical artificial neural network chip is configured to implement a calculation strategy of a part of an artificial neural network model trained in advance, and the transmission information is obtained by processing an acquired spectral image of a converter mouth flame of the converter by the optical artificial neural network chip based on the calculation strategy;
the anomaly prediction module 420 is configured to input the transmission information received at the current moment to another part of the artificial neural network model, and obtain an anomaly prediction result of the reaction in the converter;
the information generating module 430 is configured to generate monitoring information corresponding to the abnormal prediction result.
Based on any of the above embodiments, a portion of the artificial neural network model includes an input layer and a first fully connected layer;
the optical artificial neural network chip comprises an optical modulation layer and an image sensor;
the light modulation layer is used for modulating incident light from the fire hole flame based on a calculation strategy of the input layer;
the image sensor is used for sensing the optical signal modulated by the optical modulation layer based on the calculation strategy of the first full-connection layer, and processing the sensed optical signal into transmission information in the form of an electric signal.
Based on any of the above embodiments, the converter steelmaking process monitoring apparatus may further include:
the network training module is used for inputting a spectral image sample of the fire hole flame acquired by the spectral imaging chip in advance into a preset artificial neural network model for training to obtain the artificial neural network model, and the spectral image sample is marked with an actual result of the reaction in the converter;
the optical simulation module is used for carrying out optical simulation on the optical modulation layer, and adjusting parameters of the structure of the simulated optical modulation layer so that the transmittance of the simulated optical modulation layer to different wavelength components in incident light is the target transmittance;
wherein the target transmittance is determined based on a connection weight of the input layer to the first fully connected layer;
and the parameters of the structure of the simulated light modulation layer are used for manufacturing the light modulation layer on the image sensor so as to obtain the optical artificial neural network chip.
Based on any of the above embodiments, another portion of the artificial neural network model includes a second fully connected layer and a nonlinear activation layer;
the anomaly prediction module is specifically configured to input the transmission information received at the current moment into the second full-connection layer to obtain a full-connection result; and inputting the full connection result into a nonlinear activation layer to obtain an abnormal prediction result.
Based on any of the above embodiments, the anomaly prediction result includes an anomaly prediction result corresponding to the current time and/or an anomaly prediction result corresponding to a preset time period after the current time.
Based on any of the above embodiments, the exception prediction result includes that the reaction is normal, or that a preset type of exception occurs.
Based on any of the foregoing embodiments, the information generating module is specifically configured to generate, in response to the abnormality prediction result including occurrence of an abnormality of a preset type, early warning information corresponding to the abnormality of the preset type.
Fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, as shown in fig. 5, the electronic device may include: processor 510, communication interface (Communications Interface) 520, memory 530, and communication bus 540, wherein processor 510, communication interface 520, memory 530 complete communication with each other through communication bus 540. Processor 510 may invoke logic commands in memory 530 to perform the following method: in the converter production process, receiving transmission information from an optical artificial neural network chip, wherein the optical artificial neural network chip is used for realizing a calculation strategy of a part of an artificial neural network model trained in advance, the transmission information is obtained by processing a collected spectral image of the flame at the furnace mouth of the converter based on the calculation strategy, and the spectral image comprises spectral information; the transmission information received at the current moment is input into the other part of the artificial neural network model to obtain an abnormal prediction result of the reaction in the converter; and generating monitoring information corresponding to the abnormal prediction result.
In addition, the logic commands in the memory 530 may be implemented in the form of software functional units and may be stored in a computer readable storage medium when sold or used as a separate product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in the form of a software product stored in a storage medium, comprising several commands for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method of the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The embodiment of the invention also provides a converter steelmaking process monitoring system, which comprises: optical artificial neural network chip and electronic device provided by the above embodiments.
Embodiments of the present invention also provide a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the methods provided by the above embodiments, for example, comprising: in the converter production process, receiving transmission information from an optical artificial neural network chip, wherein the optical artificial neural network chip is used for realizing a calculation strategy of a part of an artificial neural network model trained in advance, the transmission information is obtained by processing a collected spectral image of the flame at the furnace mouth of the converter based on the calculation strategy, and the spectral image comprises spectral information; the transmission information received at the current moment is input into the other part of the artificial neural network model to obtain an abnormal prediction result of the reaction in the converter; and generating monitoring information corresponding to the abnormal prediction result.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several commands for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for monitoring a steelmaking process of a converter, comprising:
in the converter production process, receiving transmission information from an optical artificial neural network chip, wherein the optical artificial neural network chip is used for realizing a calculation strategy of a part of an artificial neural network model trained in advance, the transmission information is obtained by processing an acquired spectral image of the flame of the converter mouth based on the calculation strategy by the optical artificial neural network chip, and the spectral image comprises spectral information;
inputting the transmission information received at the current moment into the other part of the artificial neural network model to obtain an abnormal prediction result of the reaction in the converter;
and generating monitoring information corresponding to the abnormal prediction result.
2. The converter steelmaking process monitoring method as defined in claim 1 wherein a portion of said artificial neural network model includes an input layer and a first fully connected layer;
the optical artificial neural network chip comprises an optical modulation layer and an image sensor;
the light modulation layer is used for modulating incident light from the fire hole flame based on a calculation strategy of the input layer;
the image sensor is used for sensing the optical signal modulated by the optical modulation layer based on the calculation strategy of the first full-connection layer, and processing the sensed optical signal into the transmission information in the form of an electric signal.
3. The method of monitoring a converter steelmaking process according to claim 2, further comprising:
inputting a spectral image sample of the fire hole flame acquired by a spectral imaging chip in advance into a preset artificial neural network model for training to obtain the artificial neural network model, wherein the spectral image sample is marked with an actual result of the reaction in the converter;
performing optical simulation on the light modulation layer, and adjusting parameters of the structure of the simulated light modulation layer to enable the transmittance of the simulated light modulation layer to different wavelength components in incident light to be target transmittance;
wherein the target transmittance is determined based on a connection weight of the input layer to the first fully connected layer;
and the parameters of the structure of the simulated light modulation layer are used for manufacturing the light modulation layer on the image sensor so as to obtain the optical artificial neural network chip.
4. The converter steelmaking process monitoring method as defined in claim 1 wherein another portion of said artificial neural network model includes a second fully connected layer and a nonlinear activation layer;
inputting the transmission information received at the current moment into another part of the artificial neural network model to obtain an abnormal prediction result of the reaction in the converter, wherein the method comprises the following steps:
inputting the transmission information received at the current moment into the second full-connection layer to obtain a full-connection result;
and inputting the full connection result to the nonlinear activation layer to obtain the abnormality prediction result.
5. The converter steelmaking process monitoring method according to claim 1, wherein the abnormality prediction result includes an abnormality prediction result corresponding to the current time and/or an abnormality prediction result corresponding to a preset period of time after the current time.
6. The method according to claim 1, wherein the abnormality prediction result includes a normal reaction or occurrence of a preset type of abnormality.
7. The method of claim 6, wherein generating the monitoring information corresponding to the abnormality prediction result comprises:
and responding to the abnormality prediction result to generate early warning information corresponding to the abnormality of the preset type, wherein the abnormality of the preset type occurs.
8. A converter steelmaking process monitoring apparatus, comprising:
the information receiving module is used for receiving transmission information from an optical artificial neural network chip in the converter production process, wherein the optical artificial neural network chip is used for realizing a calculation strategy of a part of an artificial neural network model trained in advance, the transmission information is obtained by processing an acquired spectral image of the flame of the converter mouth based on the calculation strategy by the optical artificial neural network chip, and the spectral image comprises spectral information;
the abnormality prediction module is used for inputting the transmission information received at the current moment into the other part of the artificial neural network model to obtain an abnormality prediction result of the reaction in the converter;
and the information generation module is used for generating monitoring information corresponding to the abnormal prediction result.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the converter steelmaking process monitoring method according to any one of claims 1 to 7 when the program is executed.
10. A converter steelmaking process monitoring system, comprising:
an optical artificial neural network chip and an electronic device as claimed in claim 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111314402.5A CN116083676A (en) | 2021-11-08 | 2021-11-08 | Method, device, equipment and system for monitoring converter steelmaking process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111314402.5A CN116083676A (en) | 2021-11-08 | 2021-11-08 | Method, device, equipment and system for monitoring converter steelmaking process |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116083676A true CN116083676A (en) | 2023-05-09 |
Family
ID=86197816
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111314402.5A Pending CN116083676A (en) | 2021-11-08 | 2021-11-08 | Method, device, equipment and system for monitoring converter steelmaking process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116083676A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0598335A (en) * | 1991-10-09 | 1993-04-20 | Nkk Corp | Method for predicting slopping in converter |
| CN102206727A (en) * | 2011-05-31 | 2011-10-05 | 湖南镭目科技有限公司 | Converter steelmaking endpoint determination method and system, control method and control system |
| CN102392095A (en) * | 2011-10-21 | 2012-03-28 | 湖南镭目科技有限公司 | Termination point prediction method and system for converter steelmaking |
| CN205024254U (en) * | 2015-10-16 | 2016-02-10 | 唐山钢铁集团有限责任公司 | Converter mouth flame image collection system |
| CN105925750A (en) * | 2016-05-13 | 2016-09-07 | 南阳理工学院 | Steelmaking end point prediction method based on neural networks |
| CN110309973A (en) * | 2019-07-01 | 2019-10-08 | 中冶赛迪重庆信息技术有限公司 | A kind of converter splash prediction technique and system based on video intelligent algorithm |
| CN111104856A (en) * | 2019-11-18 | 2020-05-05 | 中冶赛迪技术研究中心有限公司 | Converter smelting splash monitoring method, system, storage medium and equipment |
| CN111753597A (en) * | 2019-03-29 | 2020-10-09 | 中国安全生产科学研究院 | Splash early warning system based on image recognition |
-
2021
- 2021-11-08 CN CN202111314402.5A patent/CN116083676A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0598335A (en) * | 1991-10-09 | 1993-04-20 | Nkk Corp | Method for predicting slopping in converter |
| CN102206727A (en) * | 2011-05-31 | 2011-10-05 | 湖南镭目科技有限公司 | Converter steelmaking endpoint determination method and system, control method and control system |
| CN102392095A (en) * | 2011-10-21 | 2012-03-28 | 湖南镭目科技有限公司 | Termination point prediction method and system for converter steelmaking |
| CN205024254U (en) * | 2015-10-16 | 2016-02-10 | 唐山钢铁集团有限责任公司 | Converter mouth flame image collection system |
| CN105925750A (en) * | 2016-05-13 | 2016-09-07 | 南阳理工学院 | Steelmaking end point prediction method based on neural networks |
| CN111753597A (en) * | 2019-03-29 | 2020-10-09 | 中国安全生产科学研究院 | Splash early warning system based on image recognition |
| CN110309973A (en) * | 2019-07-01 | 2019-10-08 | 中冶赛迪重庆信息技术有限公司 | A kind of converter splash prediction technique and system based on video intelligent algorithm |
| CN111104856A (en) * | 2019-11-18 | 2020-05-05 | 中冶赛迪技术研究中心有限公司 | Converter smelting splash monitoring method, system, storage medium and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110438284B (en) | Intelligent tapping device of converter and control method | |
| CN110031477A (en) | Bridge key component disease early warning system and method based on image monitoring data | |
| CN102428700B (en) | Monitoring Apparatus | |
| CN105925750A (en) | Steelmaking end point prediction method based on neural networks | |
| CN110490867A (en) | Metal increasing material manufacturing forming dimension real-time predicting method based on deep learning | |
| CN101966083A (en) | Abnormal skin area computing system and computing method | |
| CN113469231A (en) | Fault diagnosis method, fault diagnosis system, computer device, and storage medium | |
| CN105223932A (en) | Nuclear plant safety method for early warning, system and nuclear power station emulation technology platform | |
| CN114707970B (en) | Electrolytic production parameter determining method | |
| CN116083676A (en) | Method, device, equipment and system for monitoring converter steelmaking process | |
| CN114723675A (en) | Photovoltaic module detection method, device, equipment and storage medium | |
| CN116567395A (en) | Security equipment linkage operation regulation and control method and computer storage medium | |
| CN113469098B (en) | Intelligent visual monitoring device for organic hazardous chemical leakage | |
| CN111354496B (en) | Nuclear power plant accident online diagnosis and state tracking prediction method | |
| CN210765379U (en) | Device for intelligent tapping of converter | |
| Park et al. | Improving image monitoring performance for underwater laser cutting using a deep neural network | |
| CN116774904A (en) | Method and device for correcting data curve and nonvolatile storage medium | |
| CN116221643B (en) | Automatic switch control of colour temperature and power LED lamp of making a video recording | |
| CN117020457A (en) | Welding system and spot inspection method of welding system | |
| CN111753597A (en) | Splash early warning system based on image recognition | |
| CN111556106A (en) | Electric power online communication optimization system based on cloud computing | |
| CN114266286A (en) | Online detection method and device for welding process information | |
| CN112968790B (en) | Communication protection method for laser vision sensor and external equipment | |
| CN116086608A (en) | In-furnace reaction monitoring system and method in converter steelmaking process | |
| KR20240044317A (en) | Program, information processing device, information processing method, method of generating learning model, and molten steel treatment method |
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 |