A coal injection control method, system and apparatus for a blast furnace

Abstract

This invention discloses a method, system, and apparatus for blast furnace pulverized coal injection control, relating to the technical field of blast furnace pulverized coal injection control. The technical solution of this invention involves acquiring a set of infrared images of the blast furnace tuyeres under different operating conditions. A preset image classification model is trained based on this image set to identify the operating state of the blast furnace tuyeres corresponding to different infrared images. The current infrared image of the blast furnace tuyeres is input into the image classification model to obtain the current operating state of the blast furnace tuyeres. Since the operating states include at least normal combustion, raw material descent, trolley deflection, trolley burnout, and trolley coal shortage, corresponding control actions are executed on the blast furnace trolleys according to the current operating state. This ensures that the trolleys accurately perform normal pulverized coal injection, stop pulverized coal injection, and pipeline unblocking. This control method enables accurate control of each abnormal state of the blast furnace tuyeres, improving the reliability of blast furnace pulverized coal injection control.

Description

Technical Field

[0001] This invention relates to the technical field of blast furnace pulverized coal injection control, and more particularly to a blast furnace pulverized coal injection control method, system, and apparatus. Background Technology

[0002] During the operation of a blast furnace, pulverized coal is injected for combustion. Various abnormal states may occur during the combustion process. Existing monitoring methods mostly rely on manual visual observation, which cannot accurately identify the type of abnormal state, resulting in inadequate abnormal state handling strategies.

[0003] Therefore, improving the reliability of blast furnace pulverized coal injection control is a technical problem that urgently needs to be solved. Summary of the Invention

[0004] The present invention provides a method, system and apparatus for controlling pulverized coal injection in a blast furnace, which improves the reliability of pulverized coal injection control in a blast furnace.

[0005] The embodiments of the present invention provide the following solutions:

[0006] In a first aspect, embodiments of the present invention provide a method for controlling pulverized coal injection in a blast furnace, the method comprising:

[0007] Acquire a set of tuyere images of the blast furnace tuyere, wherein the set of tuyere images is a collection of infrared images of the blast furnace tuyere under different operating conditions;

[0008] A pre-set image classification model is trained based on a set of tuyere images to enable the image classification model to identify the operating status of the blast furnace tuyere corresponding to different infrared images. The operating status includes at least the normal combustion state, the raw material descent state, the coal gun deflection state, the coal gun burnout state, and the coal gun coal shortage state.

[0009] The current infrared image of the blast furnace tuyeres is input into the image classification model to obtain the current operating status of the blast furnace tuyeres.

[0010] The corresponding control actions are performed on the blast furnace coal guns according to the current operating status. These control actions include at least normal coal injection, stopping coal injection, and pipeline unblocking.

[0011] In one optional embodiment, the current operating state is a raw material descent state; according to the current operating state, corresponding control actions are performed on the blast furnace lances, including:

[0012] When the current operating status is determined to be the raw material descent state, control the coal gun to stop injecting coal and continue to monitor the operating status of the blast furnace tuyeres;

[0013] When the blast furnace tuyeres are in normal combustion mode, the coal guns are controlled to resume normal coal injection.

[0014] In one optional embodiment, the current operating state is either a tungsten skew state or a tungsten burnout state; according to the current operating state, corresponding control actions are performed on the blast furnace's tungsten, including:

[0015] When the current operating state is determined to be a coal gun skew state, the skew level of the coal gun is determined based on the tuyeres image of the coal gun skew state.

[0016] When the deviation level is less than the preset first level threshold, the coal gun is controlled to spray coal normally according to the deviation level and the corresponding deviation level coal injection quantity.

[0017] When the skew level is greater than or equal to the first level threshold, control the coal gun to stop injecting coal;

[0018] When the current operating status is determined to be that the coal gun is in a burnt-out state, the burnt-out level of the coal gun is determined based on the tuyeres image of the burnt-out state.

[0019] When the burn-off level is less than the preset second-level threshold, the coal gun is controlled to inject coal normally according to the burn-off level and the corresponding coal injection amount.

[0020] When the burn-off level is greater than or equal to the second-level threshold, the coal gun should be stopped from injecting coal.

[0021] In one optional embodiment, the current operating state is a coal lance cut-off state; according to the current operating state, corresponding control actions are performed on the blast furnace's coal lances, including:

[0022] When the current operating status is determined to be "coal gun cut-off state", control the coal gun to stop spraying coal.

[0023] When the coal gun stops spraying coal, the air valve for clearing the pipeline of the coal gun is opened to remove the blockage inside the coal gun.

[0024] After the coal gun completes the pipeline dredging, control the coal gun to resume normal coal injection.

[0025] In one optional embodiment, training a preset image classification model based on a set of wind vent images includes:

[0026] The operating status is classified based on the image features of each infrared image in the vent image set to obtain multiple infrared images for each operating status.

[0027] According to a preset ratio, multiple infrared images for each operating state are divided into training and testing sets for each operating state;

[0028] The image classification model is trained based on the training set for each operating state;

[0029] When the image classification model has completed training, input the test set into the trained image classification model;

[0030] When the test results meet the preset conditions, the image classification model is considered to have completed training.

[0031] In one optional embodiment, the operating status is classified based on the image features of each infrared image in the vent image set to obtain multiple infrared images for each operating status, including:

[0032] Based on the image features represented by each infrared image, determine the coal combustion area of ​​each infrared image, as well as the coal gun projection located within the coal combustion area;

[0033] When the brightness of the coal combustion area is greater than the preset brightness threshold, and the coal gun tip represented by the coal gun projection is located at the center of the coal combustion area, and the coal gun tip contains coal injection characteristics, the infrared image is determined to be in a normal combustion state.

[0034] When the brightness of the coal combustion zone is less than or equal to the brightness threshold, the infrared image is determined to be in a raw material descent state.

[0035] When the brightness of the coal combustion area is greater than the brightness threshold, and the coal gun tip represented by the coal gun projection is not located in the center of the coal combustion area, the infrared image is determined to be in a coal gun skew state.

[0036] When there is a gap in the coal gun projection, the infrared image is determined to be in the state of coal gun burnout;

[0037] When the tip of the coal gun does not contain coal injection characteristics, the infrared image is determined to be in the state of coal gun cut-off.

[0038] Secondly, embodiments of the present invention also provide a blast furnace pulverized coal injection control system, the system comprising:

[0039] Infrared thermal imagers are used to acquire infrared images of blast furnace tuyeres under different operating conditions.

[0040] The host computer's input terminal is connected to the output terminal of the infrared thermal imager.

[0041] The host computer is used to store multiple infrared images as a set of tuyere images for the blast furnace tuyere;

[0042] And train a preset image classification model based on the tuyere image set so that the image classification model can identify the operating status of the blast furnace tuyere corresponding to different infrared images. The operating status includes at least the normal combustion state, raw material descent state, coal gun tilt state, coal gun burnout state, and coal gun coal shortage state.

[0043] The current infrared image of the blast furnace tuyeres is then input into the image classification model to obtain the current operating status of the blast furnace tuyeres.

[0044] The controller's input is connected to the output of the host computer.

[0045] The controller is used to perform corresponding control actions on the coal guns of the blast furnace according to the current operating status. The control actions include at least normal coal injection, stopping coal injection, and pipeline unblocking.

[0046] Thirdly, embodiments of the present invention also provide a pulverized coal injection control device for a blast furnace, the device comprising:

[0047] The acquisition module is used to acquire a set of tuyere images of the blast furnace tuyere, wherein the set of tuyere images is a collection of infrared images of the blast furnace tuyere under different operating conditions;

[0048] The training module is used to train a preset image classification model based on the set of tuyere images, so that the image classification model can identify the operating status of the blast furnace tuyere corresponding to different infrared images. The operating status includes at least the normal combustion state, the raw material descent state, the coal gun tilting state, the coal gun burnout state, and the coal gun coal shortage state.

[0049] The acquisition module is used to input the current infrared image of the blast furnace tuyeres into the image classification model to obtain the current operating status of the blast furnace tuyeres.

[0050] The control module is used to perform corresponding control actions on the coal guns of the blast furnace according to the current operating status. The control actions include at least normal coal injection, stopping coal injection, and pipeline unblocking.

[0051] Fourthly, embodiments of the present invention also provide an electronic device, including a processor and a memory, the memory being coupled to the processor, the memory storing instructions that, when executed by the processor, cause the electronic device to perform the steps of any of the methods in the first aspect.

[0052] Fifthly, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of any of the methods in the first aspect.

[0053] The pulverized coal injection control method, system, and apparatus for a blast furnace of the present invention have the following advantages compared with the prior art:

[0054] The technical solution of this invention acquires a set of infrared images of the blast furnace tuyere under different operating conditions. A preset image classification model is trained based on this image set to identify the operating state of the blast furnace tuyere corresponding to different infrared images. By inputting the current infrared image of the blast furnace tuyere into the image classification model, the current operating state of the blast furnace tuyere can be obtained. Since the operating state includes at least normal combustion, raw material descent, trolley deflection, trolley burnout, and trolley coal shortage, corresponding control actions are executed on the blast furnace trolleys based on the current operating state. This ensures that the trolleys accurately perform normal coal injection, stop coal injection, and pipeline unblocking. This control method reduces the error of human visual observation, enabling accurate control of each abnormal state of the blast furnace tuyere, thereby improving the reliability of blast furnace coal injection control. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 A flowchart of a blast furnace pulverized coal injection control method provided in an embodiment of the present invention;

[0057] Figure 2-1 Infrared images of normal combustion state provided in embodiments of the present invention;

[0058] Figure 2-2 Infrared image of raw material falling state provided in an embodiment of the present invention;

[0059] Figure 2-3 Infrared image of the coal gun in a deflected state provided in an embodiment of the present invention;

[0060] Figure 2-4 Infrared images of the burnt state of coal guns provided in embodiments of the present invention;

[0061] Figure 2-5 Infrared image of the coal gun in the coal-cut state provided in an embodiment of the present invention;

[0062] Figure 3 A flowchart illustrating the training and application of the image classification model provided in this embodiment of the invention;

[0063] Figure 4 This is a schematic diagram of the pulverized coal injection control device for a blast furnace provided in an embodiment of the present invention. Detailed Implementation

[0064] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the protection scope of the embodiments of the present invention.

[0065] Please see Figure 1 , Figure 1 The flowchart of a blast furnace pulverized coal injection control method provided in this embodiment of the invention can be applied to the blast furnace control terminal to implement blast furnace operation control. The control terminal can be a PLC (Programmable Logic Controller) or other devices capable of running this method; no specific limitations are made here. The control method includes:

[0066] S11. Obtain a set of tuyere images of the blast furnace tuyere, wherein the set of tuyere images is a collection of infrared images of the blast furnace tuyere under different operating conditions.

[0067] Specifically, infrared images can be acquired using an infrared thermal imager or a CCD (Charge Coupled Device) camera. Taking an infrared thermal imager as an example, it can be mounted outside the blast furnace tuyeres. While the blast furnace is in operation, it acquires infrared images of the tuyeres at preset intervals. Because the tuyeres have a small viewing area but a high temperature in the tuyeres' vortex zone, the infrared thermal imager should have a narrow field of view, with an imaging area of ​​approximately 3.14 × 10⁻⁶. -4 m 2 The infrared thermal imager has a temperature measurement range of 0–2400℃. The acquired infrared images include images of the blast furnace tuyeres under different operating conditions. All infrared images can be classified, and a preset number of images from each operating condition can be selected and combined into a tuyeres image set. Alternatively, video data of the blast furnace tuyeres can be recorded using a CCD camera, and the video data can be broken down into image frames of the tuyeres area's operating state, and all image frames can be combined into a tuyeres image set. After obtaining the tuyeres image set, proceed to step S12.

[0068] S12. Train a preset image classification model based on the tuyere image set so that the image classification model can identify the operating status of the blast furnace tuyere corresponding to different infrared images. The operating status includes at least the normal combustion status, raw material descent status, coal gun deflection status, coal gun burnout status, and coal gun coal shortage status.

[0069] Specifically, the image classification model can be selected based on actual needs. A CNN (Convolutional Neural Network) model can be chosen, dividing the vent image set into training and testing sets. Each infrared image in the training set is labeled with its corresponding operating state. During training, the image classification model identifies the features corresponding to the operating state of each infrared image, enabling it to classify images based on their operating states. After training, the image classification model is tested using the testing set to determine its accuracy. An accuracy rate greater than 90% indicates that the image classification model is applicable and training is complete.

[0070] For example, training a pre-defined image classification model based on a set of images of wind vents includes:

[0071] The first step is to classify the operating status based on the image features of each infrared image in the tuyere image set, thereby obtaining multiple infrared images for each operating status. The image features represent the characteristics of the operating status, and the infrared images can be labeled based on the characteristics of different operating statuses of the blast furnace tuyere, which facilitates subsequent training and testing.

[0072] Based on the image features characterized by each infrared image, the coal combustion area and the coal lance projection within that area can be determined for each infrared image. Infrared images are taken at the blast furnace tuyeres. Since the coal lances are located outside the blast furnace tuyeres, they are photographed simultaneously with the infrared images. Due to the high brightness of pulverized coal during combustion, a coal lance projection is formed on the lance. Based on these features, the coal combustion area and coal lance projection can be determined for each infrared image. The following section will explain how to implement marking for different operating states based on these features.

[0073] Please see Figure 2-1 , Figure 2-1 This is a schematic diagram of an infrared image under normal combustion conditions. When the brightness of the coal combustion area exceeds a preset brightness threshold, and the coal lance tip, as represented by the coal lance projection, is located at the center of the coal combustion area, and the coal lance tip needs to inject pulverized coal into the blast furnace, the presence of pulverized coal injection characteristics at the coal lance tip indicates that the pulverized coal is in a normal combustion state. The coal lance position is accurate, and normal pulverized coal injection can be implemented, thus confirming the infrared image as a normal combustion state. The pulverized coal injection characteristics can be determined based on the grayscale characteristics within the coal lance tip. The pulverized coal ejected from the coal lance affects the projection of the pulverized coal combustion light onto the coal lance; therefore, the grayscale of the central area at the coal lance tip will differ from other areas of the coal lance projection. This difference can be used to determine whether the coal lance tip contains pulverized coal injection characteristics.

[0074] Please see Figure 2-2 , Figure 2-2The infrared image diagram illustrating the raw meal descent state indicates that the brightness of the coal combustion zone is less than or equal to a brightness threshold. Blast furnace slag is generally composed of gangue from the ore, ash from the fuel, and solvents. As the slag transitions from a solid to a liquid state within the furnace, it continuously drips through the coke column and into the hearth. However, when slag located at the edge of the blast furnace drips to the tuyeres, due to the high cooling intensity at the tuyeres' front end, it transforms back into a solid slag layer – a phenomenon known as "raw meal descent." If pulverized coal is continuously injected at this point, it will cause pulverized coal to continuously enter the solid slag layer, ultimately leading to difficulty in air intake at the tuyeres and affecting normal blast furnace production. Therefore, when the brightness of the coal combustion zone is less than or equal to a brightness threshold, it indicates that the injected pulverized coal has entered a solid slag layer state, and this infrared image's operating state is determined to be the raw meal descent state.

[0075] Please see Figure 2-3 , Figure 2-3 This is a schematic diagram of an infrared image showing a coal lance in a deflected state. When the brightness of the coal combustion area exceeds a brightness threshold, and the coal lance tip, as represented by the coal lance projection, is not located at the center of the coal combustion area, the infrared image is determined to be in a deflected state. While coal injection can still be carried out after the coal lance is deflected, the injection direction is not ideal, which may lead to incomplete coal combustion. The position of the coal lance projection on the infrared image differs from that of the coal lance in a normal combustion state. Therefore, when the coal lance tip is not located at the center of the coal combustion area, the infrared image is determined to be in a deflected state.

[0076] Please see Figure 2-4 , Figure 2-4 This is a schematic diagram of an infrared image showing the burnt state of a coal gun. When a gap exists in the coal gun projection, the infrared image is determined to be in the burnt state. The burnt state of the coal gun refers to the state after coal dust has eroded the coal gun. The projection contour can be extracted based on the characteristics of the coal gun projection. A smooth projection contour indicates that there is no gap in the coal gun projection, and the coal gun has not been eroded; conversely, a gap in the coal gun projection indicates localized erosion of the coal gun, thus confirming the infrared image as the burnt state.

[0077] Please see Figure 2-5 , Figure 2-5 This is a schematic diagram of an infrared image showing the coal gun in a coal-free state. When the tip of the coal gun does not contain any coal injection characteristics, the infrared image is determined to be in a coal-free state. The coal gun being in a coal-free state occurs because impurities in the coal dust clog the coal injection holes, preventing normal coal injection. In other words, the infrared image lacks the grayscale characteristics of the coal dust ejection location, thus confirming the coal gun as being in a coal-free state.

[0078] Based on the image features of the infrared image under each operating state, state marking can be implemented. The marking process can be carried out by computer or manually, without specific restrictions. After the state marking is completed, proceed to the next step.

[0079] The second step involves dividing multiple infrared images for each operating state into training and testing sets according to a preset ratio. The preset ratio can be set based on actual needs, such as 7:3 or 8:2, or different preset ratios can be set for different operating states. When the difficulty of identifying the operating state is relatively low, a preset ratio of 8:2 can be set, for example, for the raw material descent state or the coal gun deflection state. When the difficulty of identifying the operating state is relatively high and it is easy to cause confusion, the training and testing sets for that operating state are set to 7:3 to facilitate accurate training in the future.

[0080] The third step is to train the image classification model using the training set for each operating state. The training set for each operating state can be fed into the image classification model, which then performs feature recognition.

[0081] The fourth step is to input the test set into the image classification model after it has completed training. The purpose of this step is to test the training results of the image classification model. The test set can be used to determine whether the image classification model can accurately identify the image features of each operating state.

[0082] The fifth step, when the test results meet the preset conditions, indicates that the image classification model can accurately identify the image features of each operating state, and confirms that the image classification model has completed training. The preset conditions can be set based on actual needs, for example, setting the preset condition as a test accuracy rate of over 90%. After the image classification model completes training, proceed to step S13.

[0083] S13. Input the current infrared image of the blast furnace tuyeres into the image classification model to obtain the current operating status of the blast furnace tuyeres.

[0084] Specifically, this step involves the application of an image classification model. During blast furnace operation, real-time images of the blast furnace tuyeres are captured to obtain current infrared images. These images are then input into the image classification model, which determines the current operating state of the blast furnace tuyeres based on the image features. The current operating state can be any one of the following: normal combustion, raw material descent, trolley deflection, trolley burnout, or trolley coal shortage. After determining the current operating state of the blast furnace tuyeres, the process proceeds to step S14.

[0085] S14. Perform corresponding control actions on the coal guns of the blast furnace according to the current operating status. The control actions include at least normal coal injection, stopping coal injection, and pipeline unblocking.

[0086] Specifically, the current operating status characterizes the combustion and pulverized coal injection status of the blast furnace tuyeres. Based on the current operating status, it can be determined whether combustion and pulverized coal injection are normal, as well as the specific types of abnormalities, and then corresponding control strategies can be implemented. Normal pulverized coal injection means continuing to inject pulverized coal into the blast furnace to ensure its normal operation; stopping pulverized coal injection can mean halting pulverized coal injection for maintenance, or suspending pulverized coal injection altogether; pipeline unblocking means cleaning impurities from the pulverized coal injection pipelines to ensure the smooth implementation of normal pulverized coal injection. This control method utilizes an infrared thermal imager to observe the operating status of the tuyeres area, integrating artificial intelligence and automated control technologies to form a closed loop. This reduces the error rate of existing technologies relying on manual visual judgment, and improves the production efficiency and automation rate of ironmaking.

[0087] The following section will detail how to implement corresponding control actions under different current operating states.

[0088] When the current operating state is determined to be a raw material decline state, it indicates that pulverized coal is continuously entering the solid slag layer. If pulverized coal injection continues, it will cause difficulties in air intake at the tuyere. Therefore, the pulverized coal lance is controlled to stop pulverized coal injection, and the operating state of the blast furnace tuyere is continuously monitored. When the current operating state of the blast furnace tuyere is normal combustion, the pulverized coal lance is controlled to resume normal pulverized coal injection. This control strategy can ensure that when raw material decline occurs at the blast furnace tuyere, the pulverized coal lance is controlled to stop pulverized coal injection in a timely manner, and pulverized coal injection is resumed in a timely manner after the raw material decline state subsides.

[0089] When the current operating status is determined to be a coal gun skewed state, it indicates a deviation in the coal injection direction, but coal injection can still be carried out. The skew level of the coal gun is then determined based on the tuyeres image of the skewed state. If the skew level is less than the preset first-level threshold, coal injection can continue, and the coal gun is controlled to inject coal at the corresponding skew level. If the skew level is greater than or equal to the first-level threshold, the skew of the coal gun is more severe, and to prevent further escalation of the abnormal coal injection, the coal gun is controlled to stop injecting coal.

[0090] Similarly, when the current operating status is determined to be coal gun burnout, the burnout level of the coal gun is determined based on the vent image of the burnout status. If the burnout level is less than the preset second-level threshold, the coal gun is controlled to spray coal normally at the corresponding burnout level. If the burnout level is greater than or equal to the second-level threshold, the coal gun is controlled to stop spraying coal. When the coal gun stops spraying coal, an alarm signal can also be output to remind on-site inspection personnel to inspect and handle the coal gun. The alarm signal can be an audible and visual alarm signal, which can remind on-site personnel. It should be noted that the first-level threshold and the second-level threshold can be freely set based on the actual usage requirements of the scenario, and are not specifically limited here.

[0091] When the current operating state is determined to be a coal-stopped state, the coal gun is controlled to stop coal injection. While the coal gun is stopped injecting coal, the gas valve for clearing blockages in the coal gun's pipeline is opened to remove any obstructions. After the pipeline clearing is complete, the coal gun is controlled to resume normal coal injection. The coal gun includes two input pipelines: a pulverized coal input pipeline and a compressed gas input pipeline, each controlled by a corresponding solenoid valve. When the current operating state is determined to be a coal-stopped state, the solenoid valve of the pulverized coal input pipeline is first activated, stopping the coal gun from injecting coal. Then, the solenoid valve of the compressed gas input pipeline is activated, clearing the pipeline with compressed gas. When the pressure relief of the compressed gas or the duration of its flow reaches a preset threshold, it indicates that the pipeline clearing is complete, and the coal gun is controlled to resume normal coal injection.

[0092] The following embodiments of the present invention will take the identification and control of the raw material falling state as an example to illustrate the overall method of pulverized coal injection control in a blast furnace. Please refer to [link to relevant documentation]. Figure 3 , Figure 3 A flowchart illustrating the training and application of an image classification model. Specific steps include:

[0093] S301, the infrared thermal imager acquires images of the blast furnace tuyeres to obtain multiple infrared images of the raw material descending.

[0094] S302. Label multiple infrared images.

[0095] S303. Divide multiple infrared images into training dataset and test dataset at a preset ratio of 8:2.

[0096] S304. Train and test the image classification model based on the training dataset and the test dataset.

[0097] S305. When the recognition accuracy of the image classification model reaches more than 90%, the image classification model shall be deployed and uploaded to the host computer.

[0098] S306. Monitoring is carried out through an infrared thermal imager at the blast furnace tuyeres to sense the current infrared image of the blast furnace tuyeres, and the data is transmitted via TCP / IP protocol to output the current infrared image to the host computer.

[0099] S307: The host computer analyzes the current infrared image based on the deployed image classification model and outputs the corresponding execution instructions based on the OPC protocol.

[0100] S308: Executes control actions of the coal gun based on instructions from the host computer. The control actions include normal coal injection and stopping coal injection.

[0101] Based on the same technical concept as the control method, this embodiment of the invention also provides a pulverized coal injection control system for a blast furnace, the system including an infrared thermal imager, a host computer, and a controller.

[0102] Infrared thermal imagers are used to acquire infrared images of blast furnace tuyeres under different operating conditions.

[0103] The host computer's input terminal connects to the output terminal of the infrared thermal imager. The host computer stores multiple infrared images as a tuyere image set for the blast furnace tuyere; and trains a pre-defined image classification model based on this image set. This model identifies the operating status of the blast furnace tuyere corresponding to different infrared images. The operating status includes at least normal combustion, raw material descent, coal tweezers deflection, coal tweezers burnout, and coal tweezers failure. The current infrared image of the blast furnace tuyere is input into the image classification model to obtain its current operating status. An industrial control computer (ICC) can be used as the host computer. The ICC integrates the KEPServer software for communication with the controller. Due to the large data transfer volume between the ICC and the infrared thermal imager, the ICC should have high performance requirements: a CPU clock speed of at least 4GHz, at least 32GB of RAM, at least 1TB of hard disk space, and a network card bandwidth of at least 1Gbps.

[0104] The controller's input is connected to the output of the host computer. The controller is used to perform corresponding control actions on the blast furnace's coal guns based on the current operating status. These control actions include at least normal coal injection, stopping coal injection, and pipeline unblocking. A PLC can be used as the controller; the PLC should have an OPC communication interface to communicate with the KEPServer software. Control actions can be generated by the host computer and transmitted to the controller for execution via binary data. For example, normal coal injection and stopping coal injection can be set as a switch value of 0 or 1. Outputting 0 to the controller controls normal coal injection, while outputting 1 controls stopping coal injection.

[0105] Based on the same technical concept as the control method, this invention also provides a pulverized coal injection control device for a blast furnace. Please refer to [link to relevant documentation]. Figure 4 , Figure 4 This is a schematic diagram of the control device. The control device includes:

[0106] The acquisition module 401 is used to acquire a set of tuyere images of the blast furnace tuyere, wherein the set of tuyere images is a set of infrared images of the blast furnace tuyere under different operating conditions;

[0107] Training module 402 is used to train a preset image classification model based on the set of tuyere images, so that the image classification model can identify the operating status of the blast furnace tuyere corresponding to different infrared images. The operating status includes at least the normal combustion state, raw material descent state, coal gun tilting state, coal gun burnout state, and coal gun coal shortage state.

[0108] The module 403 is used to input the current infrared image of the blast furnace tuyeres into the image classification model to obtain the current operating status of the blast furnace tuyeres.

[0109] The control module 404 is used to perform corresponding control actions on the coal guns of the blast furnace according to the current operating status. The control actions include at least normal coal injection, stopping coal injection, and pipeline unblocking.

[0110] In one optional embodiment, the current operating state is a raw material decreasing state; the control module includes:

[0111] The first determining submodule is used to control the coal gun to stop injecting coal when the current operating state is the raw material descent state, and to continue to monitor the operating status of the blast furnace tuyeres;

[0112] The first control submodule is used to control the coal gun to resume normal coal injection when the current operating state of the blast furnace tuyeres is normal combustion.

[0113] In one optional embodiment, the current operating state is either a coal gun deflection state or a coal gun burnout state; the control module further includes:

[0114] The second determining submodule is used to determine the deflection level of the coal gun based on the tuyeres image of the coal gun deflection state when the current operating state is determined to be the coal gun deflection state.

[0115] The second control submodule is used to control the coal gun to spray coal normally according to the corresponding deflection level when the deflection level is less than the preset first level threshold.

[0116] The third control submodule is used to control the coal gun to stop injecting coal when the skew level is greater than or equal to the first level threshold.

[0117] The third determination submodule is used to determine the burn level of the coal gun based on the vent image of the coal gun burn state when the current operating state is determined to be the coal gun burn state.

[0118] The fourth control submodule is used to control the coal gun to spray coal normally according to the corresponding coal injection amount when the burn-off level is less than the preset second level threshold.

[0119] The fifth control submodule is used to control the coal gun to stop injecting coal when the burn-off level is greater than or equal to the second level threshold.

[0120] In one optional embodiment, the current operating state is a coal gun outage state; the control module further includes:

[0121] The fourth determination submodule is used to control the coal gun to stop spraying coal when the current operating state is the coal gun cut-off state;

[0122] The sixth control submodule is used to control the opening of the gas valve for clearing the pipeline of the coal gun when the coal gun stops spraying coal, so as to clear the blockage in the coal gun.

[0123] The seventh control submodule is used to control the coal gun to resume normal coal injection after the coal gun has completed pipeline unblocking.

[0124] In one optional embodiment, the training module includes:

[0125] The processing and acquisition submodule is used to classify the operating status based on the image features of each infrared image in the vent image set, so as to obtain multiple infrared images for each operating status.

[0126] The sub-module is used to divide multiple infrared images of each operating state into training and test sets for each operating state according to a preset ratio.

[0127] The training submodule is used to train the image classification model based on the training set for each running state;

[0128] The input submodule is used to input the test set into the trained image classification model when the model has finished training.

[0129] The fifth determination submodule is used to determine whether the image classification model has completed training when the test results meet the preset conditions.

[0130] In one alternative embodiment, the processing to obtain the submodule includes:

[0131] The first determining unit is used to determine the coal combustion area of ​​each infrared image and the coal gun projection located within the coal combustion area based on the image features characterized by each infrared image.

[0132] The second determining unit is used to determine that the infrared image is in a normal combustion state when the brightness of the coal combustion area is greater than a preset brightness threshold, the coal gun tip represented by the coal gun projection is located at the center of the coal combustion area, and the coal gun tip contains coal injection features.

[0133] The third determining unit is used to determine that the infrared image is in a raw material descent state when the brightness of the coal combustion area is less than or equal to the brightness threshold.

[0134] The fourth determining unit is used to determine that the infrared image is in a coal gun skew state when the brightness of the coal combustion area is greater than the brightness threshold and the coal gun tip represented by the coal gun projection is not located in the center of the coal combustion area.

[0135] The fifth determining unit is used to determine that the infrared image is in the state of coal gun burnout when there is a gap in the coal gun projection;

[0136] The sixth determining unit is used to determine that the infrared image is in a coal-cut-off state when the coal gun tip does not contain coal injection characteristics.

[0137] Based on the same technical concept as the control method, embodiments of the present invention also provide an electronic device, including a processor and a memory, the memory being coupled to the processor, the memory storing instructions, which, when executed by the processor, cause the electronic device to perform the steps of any of the methods in the control method.

[0138] Based on the same technical concept as the control method, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any one of the methods in the control method.

[0139] The technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages:

[0140] By acquiring infrared images of the blast furnace tuyere under different operating conditions, a pre-set image classification model is trained based on the tuyere image set. This model can then identify the operating state of the blast furnace tuyere corresponding to different infrared images. By inputting the current infrared image of the blast furnace tuyere into the image classification model, the current operating state of the blast furnace tuyere can be obtained. Since the operating state includes at least normal combustion state, raw material descent state, coal gun deflection state, coal gun burnout state, and coal gun coal cut-off state, corresponding control actions are executed on the coal gun of the blast furnace according to the current operating state. This allows the coal gun to accurately perform normal coal injection, stop coal injection, and pipeline unblocking. This control method reduces the error of human observation and judgment, and enables accurate control of each abnormal state of the blast furnace tuyere, thereby improving the reliability of blast furnace coal injection control.

[0141] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0142] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (modules, systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0143] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0144] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0145] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0146] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method of coal injection control of a blast furnace, characterized by, The method includes: Acquire a set of tuyere images of the blast furnace tuyere, wherein the set of tuyere images is a collection of infrared images of the blast furnace tuyere under different operating conditions; A preset image classification model is trained based on the set of tuyere images so that the image classification model can identify the operating state of the blast furnace tuyere corresponding to different infrared images. The operating state includes at least the normal combustion state, raw material descent state, coal gun deflection state, coal gun burnout state, and coal gun coal shortage state. The current infrared image of the blast furnace tuyeres is input into the image classification model to obtain the current operating status of the blast furnace tuyeres. According to the current operating status, the corresponding control actions are performed on the coal guns of the blast furnace, wherein the control actions include at least normal coal injection, stopping coal injection, and pipeline unblocking; The current operating state is either a tungsten skew state or a tungsten skew state; the step of performing corresponding control actions on the blast furnace tungsten according to the current operating state includes: When the current operating state is determined to be a coal gun skew state, the skew level of the coal gun is determined based on the vent image of the coal gun skew state. When the deflection level is less than the preset first level threshold, the coal gun is controlled to spray coal normally at the corresponding deflection level according to the deflection level. When the skew level is greater than or equal to the first level threshold, the coal gun is controlled to stop injecting coal. When the current operating state is determined to be a coal gun burnout state, the burnout level of the coal gun is determined based on the vent image of the coal gun burnout state; When the burn-off level is less than the preset second level threshold, the coal gun is controlled to spray coal normally at the corresponding burn-off level coal injection quantity according to the burn-off level. When the burn-off level is greater than or equal to the second level threshold, the coal gun is controlled to stop injecting coal.

2. The coal injection control method of a blast furnace according to claim 1, characterized by, The current operating state is the raw material descent state; the step of performing corresponding control actions on the blast furnace coal guns according to the current operating state includes: When the current operating state is determined to be the raw material descent state, the coal gun is controlled to stop injecting coal, and the operating state of the blast furnace tuyeres is continuously monitored; When the current operating state of the blast furnace tuyeres is the normal combustion state, the coal gun is controlled to resume normal coal injection.

3. The coal injection control method of a blast furnace according to claim 1, characterized by, The current operating state is a coal lance cut-off state; the step of performing corresponding control actions on the blast furnace coal lances according to the current operating state includes: When the current operating state is determined to be a coal gun cut-off state, control the coal gun to stop spraying coal; When the coal gun stops spraying coal, the air valve for clearing the pipeline of the coal gun is opened to clear the blockage in the coal gun. After the coal gun completes the pipeline unblocking, control the coal gun to resume normal coal injection.

4. The coal injection control method of a blast furnace according to claim 1, characterized by, The step of training a preset image classification model based on the set of wind vent images includes: The operating status is classified according to the image features of each infrared image in the vent image set to obtain multiple infrared images for each operating status. According to a preset ratio, multiple infrared images for each operating state are divided into training and testing sets for each operating state; The image classification model is trained based on the training set for each operating state; When the image classification model completes training, the test set is input into the trained image classification model; When the test results meet the preset conditions, the image classification model is considered to have completed training.

5. The coal injection control method of a blast furnace according to claim 4, characterized by, The step of classifying the operating status based on the image features of each infrared image in the vent image set to obtain multiple infrared images for each operating status includes: Based on the image features represented by each infrared image, the coal combustion area of ​​each infrared image and the coal gun projection located within the coal combustion area are determined. When the brightness of the coal combustion area is greater than a preset brightness threshold, and the coal gun tip represented by the coal gun projection is located at the center of the coal combustion area, and the coal gun tip contains coal injection characteristics, the infrared image is determined to be in a normal combustion state. When the brightness of the coal combustion zone is less than or equal to the brightness threshold, the infrared image is determined to be in a raw material descent state. When the brightness of the coal combustion area is greater than the brightness threshold, and the coal gun tip represented by the coal gun projection is not located in the center of the coal combustion area, the infrared image is determined to be a coal gun skewed state. When there is a gap in the projection of the coal gun, the infrared image is determined to be in a state of coal gun burnout; When the tip of the coal gun does not contain coal injection characteristics, the infrared image is determined to be in a coal gun cut-off state.

6. An electronic device, characterized in that, The device includes a processor and a memory, the memory being coupled to the processor, the memory storing instructions that, when executed by the processor, cause the electronic device to perform the steps of the method according to any one of claims 1-5.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that, When executed by a processor, the program implements the steps of the method described in any one of claims 1-5.

Images