A refining furnace heating gas combustion safety intelligent monitoring method and system

Abstract

The application discloses a kind of refining furnace oven coal gas combustion safety intelligent monitoring method, comprising: S1: before refining furnace oven, install the gas combustion safety detection unit suitable for the front end of oxygen lance;S2: oxygen lance and gas combustion safety detection unit are inserted into to refining furnace, from gas ignition to the time period of extinguishing, synchronous acquisition flame color image and atmosphere concentration in refining furnace;S3: the flame color image collected is handled, to judge flame ignition state and monitor gas combustion flame state;S4: different oven time flame ignition state, gas combustion flame state and atmosphere concentration are set different alarm rules, if abnormal, then alarm;If normal, then end detection.The application further discloses a kind of refining furnace oven coal gas combustion safety intelligent monitoring system.The application greatly improves the safety control level of refining furnace oven process, prolongs the service life of refining furnace, improves refining furnace smelting level and improves steel product quality.

Description

Technical Field

[0001] This invention relates to the field of safety monitoring technology for refining furnaces in the iron and steel metallurgical process, and particularly to an intelligent monitoring method and system for the safe combustion of furnace gas during the baking process of a refining furnace. Background Technology

[0002] The refining furnace is a crucial piece of equipment for improving the quality of steel products and increasing smelting efficiency. Its main functions include deoxidation, decarburization, desulfurization, degassing, removal of rolled impurities, modification of inclusions, adjustment of molten steel temperature and composition, and fine-tuning of alloys. Because the refractory bricks inside the refining furnace are subject to corrosion from the high-temperature molten steel, their lifespan is limited. Brick replenishment and baking are necessary during each furnace cycle, an essential process for the refining furnace to proceed to the next cycle. Baking is typically achieved by supplying gas through a top lance and igniting it to bake the refractory materials inside the furnace. Due to the enclosed environment of the refining furnace and the limitations of monitoring methods, flame detectors are generally used to detect whether the flame inside the furnace is ignited.

[0003] Currently, flame detectors can only detect whether the gas supplied by the top lance in the furnace is in an ignition state. They cannot provide online assessments of whether the gas is completely combusted or whether the furnace atmosphere exceeds acceptable levels. Occasionally, incomplete combustion of the gas can lead to safety accidents such as low combustion efficiency, gas leaks, and furnace explosions. This significantly reduces the smelting efficiency and service life of refining furnaces, and greatly increases the production costs of steel companies. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a method and system for intelligent monitoring of the combustion safety of refining furnace drying gas. This invention combines machine vision, temperature detection, image processing, and deep learning algorithms to achieve safe monitoring and audible / visual early warning of refining furnace drying gas combustion.

[0005] To achieve the above objectives, an embodiment of the present invention provides the following technical solution:

[0006] A method for intelligent monitoring of combustion safety of gas in a refining furnace includes:

[0007] S1: Install a gas combustion safety detection unit adapted to the front end of the oxygen lance before the refining furnace is baked.

[0008] S2: Insert the oxygen lance and gas combustion safety detection unit into the refining furnace and simultaneously collect color images of the flame and atmosphere concentration inside the refining furnace from the time of gas ignition to the time of flameout.

[0009] S3: Process the color image of the flame acquired in step S2 to determine the flame ignition status and monitor the state of the gas combustion flame.

[0010] S4: Set different alarm rules for the flame ignition status, gas combustion flame status and atmosphere concentration at different times of furnace baking. If an abnormality occurs, an alarm will be triggered; if normal, the detection will end.

[0011] As a further improvement of the present invention, in step S3, the flame color image acquired in step S2 is processed to determine the flame ignition state, including:

[0012] S3.1: Collect images of gas combustion flames at different flow rates as a training set for ignition status;

[0013] S3.2: Separate the R channel image from the flame color image acquired in step S2, and use it as the image for judging the flame ignition status using deep learning methods;

[0014] S3.3: Construct a deep learning ignition point recognition model pre-trained with a training set of gas ignition states to determine whether the gas flame in the furnace is in an ignition state.

[0015] As a further improvement of the present invention, the training set of step S3.1 includes images of ten different fire states, with 200 images for each fire state.

[0016] As a further improvement of the present invention, in step S3, the color image of the flame acquired in step S2 is processed to monitor the state of the gas combustion flame, including:

[0017] S3.4: Separate the B channel image from the flame color image acquired in step S2, and use it as the image basis for monitoring the gas combustion flame status during the furnace baking process;

[0018] S3.5: Perform noise filtering preprocessing and edge extraction on the image to obtain the edge of the combustion flame area, and obtain the gas combustion flame parameters based on the pixel values ​​and actual calibration results.

[0019] S3.6: A long short-term memory network model is adopted to compare the gas combustion flame parameters at the previous moment with the parameters at the current moment and calculate the absolute difference, so as to realize real-time monitoring of the gas combustion flame status during the furnace baking period.

[0020] As a further improvement of the present invention, the gas combustion flame parameters include flame length, flame width, flame abundance, flame combustion center region, and flame color.

[0021] As a further improvement of the present invention, the calculation formulas for flame length, flame width, and flame abundance are as follows:

[0022]

[0023]

[0024]

[0025] Where Li(fire) is the length of the i-th flame at a certain time; L1 is the actual pixel corresponding to the region with the longest flame at time 1 in the image; Ln is the actual pixel corresponding to the region with the longest flame at time n in the image; α is the pixel resolution in the length direction; n is the number of images acquired per second; Wi(fire) is the width of the i-th flame at a certain time; W1 is the actual pixel corresponding to the region with the widest flame at time 1 in the image; Wn is the actual pixel corresponding to the region with the widest flame at time n in the image; β is the pixel resolution in the width direction; and Fi(fire) is the abundance of the i-th flame at a certain time.

[0026] As a further improvement of the present invention, in step S2, a color image of the flame and the atmosphere concentration within a 360° range inside the refining furnace are acquired.

[0027] As a further improvement of the present invention, if an abnormality occurs in step S4, manual intervention confirmation is also included.

[0028] A refining furnace drying gas combustion safety intelligent monitoring system, characterized in that the monitoring method described above includes:

[0029] A gas combustion safety detection unit is used to acquire real-time color images of flames and atmosphere concentration during the furnace baking process of a refining furnace. The gas combustion safety detection unit includes a gas concentration detector, a visible light color industrial CCD, and a communication module.

[0030] The communication unit is used to transmit color images of the flame and atmosphere concentration during the furnace drying process of the refining furnace to the main server;

[0031] The main server is used to receive color images of flames and atmosphere concentration data during the refining furnace drying process, analyze and monitor them online, set different alarm rules, and provide alarms.

[0032] As a further improvement of the present invention, the gas combustion safety detection unit further includes a protection module, a cooling module, and a rotation module.

[0033] As a further improvement of the present invention

[0034] The beneficial effects of this invention are:

[0035] (1) The present invention can realize an online monitoring method for whether the gas flame in the furnace is in an ignition state during the furnace baking process of the refining furnace, which greatly improves the safety control level of the furnace baking process of the refining furnace, lays a good foundation for extending the service life of the refining furnace, improving the smelting level of the refining furnace, and improving the quality of steel products. It also helps to improve the smelting efficiency of the refining furnace, improve the quality of molten steel, reduce smelting costs and promote low-carbon smelting.

[0036] (2) The intelligent monitoring system for gas combustion safety in refining furnace baking of the present invention can simultaneously obtain gas combustion monitoring images and furnace atmosphere concentration data in the furnace. It judges the gas combustion safety status in the refining furnace baking process from two dimensions. At the same time, it sets different alarm rules and provides interface abnormal sound and light alarm and manual intervention confirmation functions. It not only solves the problem of real-time monitoring in the refining furnace baking process, but also provides online early warning and manual access confirmation functions. It provides good support for improving the safety management level of the refining process, and also lays a good foundation for improving the intelligent level of safety monitoring.

[0037] (3) The intelligent monitoring method and system for safe combustion of refining furnace gas proposed in this invention are highly operable, highly integrated and highly intelligent, and have important application value. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a flowchart of a preferred embodiment of the monitoring method of the present invention;

[0040] Figure 2 This is a flowchart illustrating the monitoring method according to a preferred embodiment of the present invention.

[0041] Figure 3 This is a schematic diagram of a monitoring system according to a preferred embodiment of the present invention. Detailed Implementation

[0042] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0043] Please see Figure 1 , Figure 2 This application provides an intelligent monitoring method for the safety of combustion of gas in a refining furnace, comprising:

[0044] S1: Before baking the refining furnace, install a gas combustion safety detection unit adapted to the front end of the oxygen lance.

[0045] S2: Insert the oxygen lance and gas combustion safety detection unit into the refining furnace and simultaneously collect color images of the flame and atmosphere concentration inside the refining furnace during the time period from gas ignition to flameout.

[0046] Preferably, acquiring color images of the flame and atmosphere concentration within a 360° range inside the refining furnace can provide a more comprehensive view of the flame and atmosphere concentration, thus improving accuracy.

[0047] Preferably, the collected atmosphere concentrations include CO, CO2, O2, N2, and H2 concentrations, which better reflect the safety inside the refining furnace.

[0048] S3: Process the color image of the flame acquired in step S2 to determine the flame ignition status and monitor the flame status of the gas combustion.

[0049] Preferably, processing the flame color images acquired in step S2 to determine the flame ignition state includes: First, acquiring flame images of gas combustion at different flow rates as a training set for ignition states. This training set includes images of ten different ignition states, with 200 images for each state, resulting in a total of 2000 image samples for gas ignition state detection. Since the color, shape, and size of the flame differ under different flow rates, the combustion process at different flow rates is used for classification to form images of different ignition states. Next, the R-channel image is separated from the flame color images acquired in step S2, and this R-channel image is used to determine the flame ignition state using a deep learning method. Finally, a deep learning ignition point recognition model pre-trained with the ignition state training set is constructed to determine whether the furnace gas flame is in an ignition state.

[0050] Preferably, processing the color flame image acquired in step S2 to monitor the state of the gas combustion flame includes: first, separating the B-channel image from the color flame image acquired in step S2, and using the B-channel image as the image basis for monitoring the state of the gas combustion flame during the furnace drying process; next, performing noise filtering preprocessing and edge extraction on the image to obtain the edge of the combustion flame area, and obtaining the gas combustion flame parameters based on the pixel values ​​and actual calibration results; finally, using a long short-term memory network model to compare the gas combustion flame parameters at the previous moment with the parameters at the current moment, and calculating the absolute difference to achieve real-time monitoring of the state of the gas combustion flame during the furnace drying process.

[0051] Among them, the parameters of gas combustion flame include flame length, flame width, flame abundance, flame combustion center area, and flame color.

[0052] Specifically, the formulas for calculating flame length, flame width, and flame abundance are as follows:

[0053]

[0054] Where Li(fire) is the length of the i-th flame at a certain time, mm; L1 is the actual pixel corresponding to the region with the longest flame at time 1 in the image, pixel; Ln is the actual pixel corresponding to the region with the longest flame at time n in the image, pixel; α is the pixel resolution in the length direction, mm / pixel; n is the number of images acquired per second, - (dimensionless).

[0055]

[0056] Where Wi(fire) is the width of the i-th flame at a certain time, mm; W1 is the actual pixel corresponding to the widest region occupied by the flame at time 1 in the image, pixel; Wn is the actual pixel corresponding to the widest region occupied by the flame at time n in the image, pixel; β is the pixel resolution in the width direction, mm / pixel; n is the number of images acquired per second, - (dimensionless).

[0057]

[0058] Where Fi(fire) is the fire abundance at a certain moment, - (dimensionless).

[0059] S4: Set different alarm rules for the flame ignition status, gas combustion flame status and atmosphere concentration at different times of furnace baking. If an abnormality occurs, an alarm will be triggered; if normal, the detection will end.

[0060] To enhance safety, if any abnormality occurs, an audible and visual alarm will be triggered via the computer interface on the control terminal, requiring manual intervention for confirmation and handling to ensure safe operation.

[0061] Please see Figure 3 This application also discloses an intelligent monitoring system for the safety of gas combustion in a refining furnace during the furnace drying process, applicable to the aforementioned monitoring method. The system includes a gas combustion safety detection unit 1, used to collect real-time color images of the flame and atmosphere concentration during the furnace drying process. The gas combustion safety detection unit 1 includes a gas concentration detector 11, a visible light color industrial CCD 12, and a communication module 13. A communication unit 2 is used to transmit the color images of the flame and atmosphere concentration during the furnace drying process to a main server 3. The main server 3 is used to receive the color images of the flame and atmosphere concentration during the furnace drying process, perform analysis and online monitoring, set different alarm rules, and provide alarms.

[0062] Among them, the gas concentration detector 11 can realize real-time online detection of the concentration of CO, CO2, O2, N2, and H2 in the atmosphere inside the refining furnace, providing basic support for judging the abnormal concentration of a certain dangerous gas in the furnace atmosphere. The visible light color industrial CCD 12 can realize real-time acquisition of color images of the gas flame inside the furnace, providing image data support for subsequent flame ignition status analysis and gas combustion flame status monitoring. The flame color image is composed of images from three channels: R (red), G (green), and B (blue). The communication module 13 can send the detection data acquired by the gas concentration detector 11 and the visible light color industrial CCD 12 to the outside of the gas combustion safety detection unit 1, receive it through the communication unit 2, and transmit it to the main server 3.

[0063] The gas combustion safety detection unit 1 also includes a protection module 14, a cooling module 15, and a rotation module 16. The protection module 14 protects the gas concentration detector 11 and the visible light color industrial CCD 12 from interference and impact caused by high temperatures, dust, and corrosive gases within the furnace. The cooling module 15 provides cooling for the gas concentration detector 11 and the visible light color industrial CCD 12, and provides negative pressure protection for the detection window to prevent dust particles from adhering to the window and causing interference. The rotation module 16 enables flame imaging and furnace atmosphere concentration sampling within a 360° range for the gas concentration detector 11 and the visible light color industrial CCD 12 within the refining furnace.

[0064] First, the refining furnace preparation and start-up signal is acquired. Then, the gas combustion safety detection unit 1, adapted to the oxygen lance front end, is installed. The system starts detecting signals, and the gas combustion safety detection unit 1 begins operation. On one hand, the gas concentration detector 11 performs real-time online detection of the furnace atmosphere concentrations such as CO, CO2, O2, N2, and H2. On the other hand, the visible light color industrial CCD 12 acquires flame color images. Image processing methods are used to analyze the gas combustion flame state displayed in the flame color images, and a long short-term memory network model is used to achieve online monitoring of the gas combustion flame state. Simultaneously, the R-channel image in the visible light flame color image is separated, and the flame ignition state is determined using deep learning methods. Different alarm rules are formulated by an expert group composed of the refining furnace operator, safety personnel, and other relevant personnel for different furnace start-up times regarding the flame ignition state, gas combustion flame state, and atmosphere concentration. If any of these conditions is not met, an alarm is triggered. The result display and early warning unit in main server 3 displays the safety status of the refining furnace gas combustion according to rules. If an abnormality occurs, an audible and visual alarm is triggered via the operating terminal computer interface, requiring manual intervention for confirmation and safe operation. For example, the flame ignition status primarily detects the presence of a flame. If the flame is not ignited, it is displayed as "no flame ignition" on main server 3. In this case, the system alarms, requiring manual intervention for relevant safety measures, such as closing gas valves. If the gas combustion flame is incomplete and the atmosphere concentration reaches a dangerous level, the system alarms, requiring operators to decide on the final handling measures, such as increasing the oxygen ratio. Finally, the system issues a termination signal after the refining furnace drying process is completed.

[0065] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0066] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for intelligent monitoring of combustion safety of gas in a refining furnace, characterized in that, include: S1: Install a gas combustion safety detection unit adapted to the front end of the oxygen lance before the refining furnace is baked. S2: Insert the oxygen lance and gas combustion safety detection unit into the refining furnace and simultaneously collect color images of the flame and atmosphere concentration inside the refining furnace from the time of gas ignition to the time of flameout. S3: Process the color image of the flame acquired in step S2 to determine the flame ignition status and monitor the state of the gas combustion flame. S4: Set different alarm rules for the flame ignition status, gas combustion flame status and atmosphere concentration at different times of furnace baking. If an abnormality occurs, an alarm will be triggered; if normal, the detection will end.

2. The intelligent monitoring method for safe combustion of refining furnace gas as described in claim 1, characterized in that, In step S3, the color image of the flame acquired in step S2 is processed to determine the flame ignition status, including: S3.1: Collect images of gas combustion flames at different flow rates as a training set for ignition status; S3.2: Separate the R channel image from the flame color image acquired in step S2, and use it as the image for judging the flame ignition status using deep learning methods; S3.3: Construct a deep learning ignition point recognition model that has been pre-trained using the ignition state training set, and then determine whether the furnace gas flame is in an ignition state.

3. The intelligent monitoring method for safe combustion of refining furnace gas as described in claim 2, characterized in that, The training set in step S3.1 includes images of ten different fire states, with 200 images for each fire state.

4. The intelligent monitoring method for safe combustion of refining furnace gas as described in claim 1, characterized in that, In step S3, the color image of the flame acquired in step S2 is processed to monitor the state of the gas combustion flame, including: S3.4: Separate the B channel image from the flame color image acquired in step S2, and use it as the image basis for monitoring the gas combustion flame status during the furnace baking process; S3.5: Perform noise filtering preprocessing and edge extraction on the image to obtain the edge of the combustion flame area, and obtain the gas combustion flame parameters based on the pixel values ​​and actual calibration results; S3.6: A long short-term memory network model is adopted to compare the gas combustion flame parameters at the previous moment with the parameters at the current moment and calculate the absolute difference, so as to realize real-time monitoring of the gas combustion flame status during the furnace baking period.

5. The intelligent monitoring method and system for safe combustion of refining furnace gas as described in claim 4, characterized in that, The parameters of the gas combustion flame include flame length, flame width, flame abundance, flame combustion center area, and flame color.

6. The intelligent monitoring method and system for safe combustion of refining furnace gas as described in claim 5, characterized in that, The formulas for calculating flame length, flame width, and flame abundance are as follows: Where Li(fire) is the length of the i-th flame at a certain time; L1 is the actual pixel corresponding to the region with the longest flame at time 1 in the image; Ln is the actual pixel corresponding to the region with the longest flame at time n in the image; α is the pixel resolution in the length direction; n is the number of images acquired per second; Wi(fire) is the width of the i-th flame at a certain time; W1 is the actual pixel corresponding to the region with the widest flame at time 1 in the image; Wn is the actual pixel corresponding to the region with the widest flame at time n in the image; β is the pixel resolution in the width direction; and Fi(fire) is the abundance of the i-th flame at a certain time.

7. The intelligent monitoring method for safe combustion of refining furnace gas as described in claim 1, characterized in that, In step S2, a color image of the flame and the atmosphere concentration within a 360° range inside the refining furnace are acquired.

8. The intelligent monitoring method for safe combustion of refining furnace gas as described in claim 1, characterized in that, In step S4, if an abnormality occurs, manual intervention for confirmation is also included.

9. A smart monitoring system for the safety of combustion of gas in a refining furnace, characterized in that, The monitoring method applied to any one of claims 1-8 includes: A gas combustion safety detection unit is used to acquire real-time color images of flames and atmosphere concentration during the furnace baking process of a refining furnace. The gas combustion safety detection unit includes a gas concentration detector, a visible light color industrial CCD, and a communication module. The communication unit is used to transmit color images of the flame and atmosphere concentration during the furnace drying process of the refining furnace to the main server; The main server is used to receive color images of flames and atmosphere concentration data during the refining furnace drying process, analyze and monitor them online, set different alarm rules, and provide alarms.

10. The intelligent monitoring method for safe combustion of refining furnace gas as described in claim 9, characterized in that, The gas combustion safety detection unit also includes a protection module, a cooling module, and a rotation module.

Images