EQUIPMENT FOR DETECTING MOLTEN METAL LEAKS, METHODS FOR DETECTING MOLTEN METAL LEAKS, AND METHODS FOR CONTINUOUS METAL CASTING
Patent Information
- Authority / Receiving Office
- VN · VN
- Patent Type
- Applications
- Current Assignee / Owner
- JFE STEEL CORP
- Filing Date
- 2024-09-20
- Publication Date
- 2026-07-01
AI Technical Summary
Existing leak detection methods for molten metal in continuous casting facilities often result in false detections due to the similarity in color between molten metal splashes during healthy operations and actual leaks, leading to delayed response times and equipment damage.
A leak detection device that captures image data of the molten metal handling equipment at predetermined intervals, acquires time series data of pixels indicating a specific color, and detects leaks based on duration or integrated value exceeding predetermined thresholds, thereby reducing false positives.
The solution effectively suppresses false detection of molten metal leaks, allowing for timely intervention and preventing equipment damage by accurately distinguishing between normal operations and actual leaks.
Smart Images
Figure VN1202603081_0
Abstract
Description
Molten metal leak detection device, molten metal leak detection method, and metal continuous casting method
[0001] The present invention relates to a molten metal leak detection device for detecting molten metal leaks from molten metal handling equipment, a molten metal leak detection method, and a metal continuous casting method using the molten metal leak detection method.
[0002] In facilities that handle molten metal, accidents such as molten metal leaks that cause serious damage to the facility are inevitable. Therefore, when a molten metal leak occurs, it is necessary to detect it quickly and take measures to prevent the damage to the facility from spreading. A typical response is to take emergency measures to stop casting, such as stopping the supply of molten metal to the affected facility.
[0003] In the past, detecting abnormal conditions such as molten metal leaks and taking measures to prevent damage from spreading were typically handled by workers. The worker would perceive the abnormal condition visually or by sound, determine that an abnormal condition had occurred, and press a button to immediately stop the supply of molten steel to the affected equipment. When relying on workers to detect abnormal conditions and prevent their spread, the detection of the abnormal condition could be delayed due to factors such as the worker's characteristics or the workload at the time of the abnormality, resulting in significant damage to the equipment.
[0004] In response to such problems, Patent Document 1 discloses a method and apparatus for automatically detecting a steel leakage accident (breakout), which is an abnormal condition that occurs during a continuous casting process, and taking measures to prevent the damage from spreading. Here, a breakout refers to an abnormal condition in which the outer surface of a slab breaks for some reason, causing the molten steel inside to erupt. According to Patent Document 1, an image of a slab during continuous casting is captured, and if the number of pixels that exhibit a predetermined specific color among the pixels constituting the obtained image data exceeds a predetermined threshold, it is determined that a breakout has occurred.
[0005] Japanese Patent Application Laid-Open No. 2001-269770
[0006] In Patent Document 1, the occurrence of a breakout, i.e., the occurrence of an abnormal state, in a continuous casting facility is detected using instantaneous pixel values. However, in a continuous casting facility, even during normal casting operations where no breakout has occurred, molten metal splashes and sparks occur around the slab during casting.
[0007] The color of these flying sparks is almost the same as that of the molten metal leaking when a breakout occurs. Therefore, there is a problem that flying sparks from molten metal during normal casting operations may be mistakenly detected as a breakout. The present invention has been made in consideration of such problems in the prior art, and its object is to provide a molten metal leak detection device and a molten metal leak detection method that can suppress mistaken detection of molten metal leaks. Another object of the present invention is to provide a continuous metal casting method that uses the molten metal leak detection method.
[0008] Means for solving the above problems are as follows. [1] A molten metal leak detection device for detecting a molten metal leak from molten metal handling equipment, the molten metal leak detection device having: an imaging unit that images the molten metal handling equipment at predetermined time intervals to generate image data; a data acquisition unit that acquires time-series data on the number of pixels showing a specific color included in a predetermined area of the image data; and a detection unit that detects a molten metal leak using the time-series data. [2] The molten metal leak detection device described in [1], wherein the detection unit uses the time-series data to measure a duration during which the number of pixels showing the specific color continues to be equal to or greater than a reference number, and detects a molten metal leak when the duration exceeds a predetermined threshold. [3] The molten metal leak detection device described in [1], wherein the detection unit uses the time-series data to measure an integrated value of the number of pixels showing the specific color, and detects a molten metal leak when the integrated value within a predetermined time period exceeds a predetermined threshold. [4] The molten metal leak detection device according to any one of [1] to [3], wherein the molten metal handling equipment is a continuous metal casting machine, and the imaging unit is provided at a position where it can image an area below a casting mold. [5] A molten metal leak detection method for detecting a molten metal leak from molten metal handling equipment, comprising: an imaging step of imaging the molten metal handling equipment at predetermined time intervals to generate image data; a data acquisition step of acquiring time-series data on the number of pixels exhibiting a specific color included in a predetermined area of the image data; and a detection step of detecting a molten metal leak using the time-series data. [6] The molten metal leak detection method according to [5], wherein the detection step measures the duration during which the number of pixels exhibiting a specific color continues to be equal to or greater than a reference number using the time-series data, and detects the molten metal leak when the duration is equal to or greater than a predetermined threshold. [7] The molten metal leakage detection method according to [5], wherein in the detection step, an integrated value of the number of pixels showing a specific color is measured using the time series data, and if the integrated value becomes equal to or greater than a predetermined threshold within a predetermined time, it is detected that the molten metal has leaked.[8] The molten metal leak detection method according to any one of [5] to [7], wherein the molten metal handling equipment is a continuous metal casting machine, and the imaging step generates image data by imaging an area below a mold. [9] A continuous metal casting method, wherein if a molten metal leak is detected by the molten metal leak detection method according to [8], the supply of molten metal to the mold is stopped.
[0009] The molten metal leak detection device and molten metal leak detection method of the present invention detect molten metal leaks using time series data of the number of pixels showing a specific color, thereby making it possible to reduce false detections of molten metal leaks.
[0010] Fig. 1 is a cross-sectional schematic diagram showing an example in which a molten metal leak detection device according to this embodiment is applied to a continuous metal casting machine having a mold. Fig. 2 is a schematic diagram showing an example of the configuration of an image analysis device. Fig. 3 is a graph showing the time change in the number of pixels of a specific color in a continuous steel casting machine.
[0011] The molten metal leak detection device according to this embodiment will be described below using an example in which it is applied to a continuous metal casting machine. A continuous metal casting machine is an example of molten metal handling equipment. Other examples of molten metal handling equipment include a transfer vessel or transfer ladle that transports molten metal, and the molten metal leak detection device according to this embodiment can also be applied to these equipment.
[0012] First, a continuous metal casting machine 10 will be described. FIG. 1 is a cross-sectional schematic diagram showing an example in which a molten metal leak detection device 40 according to this embodiment is applied to a continuous metal casting machine 10 having a mold 12. The continuous metal casting equipment 10 includes the mold 12, a tundish 14 installed above the mold 12, and a plurality of strand support rolls 16 arranged in a row below the mold 12. Although not shown, a ladle containing molten metal 18 is installed above the tundish 14, and the molten metal 18 is poured into the tundish 14 from the bottom of the ladle. A sliding shutter 19 and an immersion nozzle 20 are installed at the bottom of the tundish 14. When the sliding shutter 19 slides and the closure by the sliding shutter 19 is released, the molten metal 18 is poured into the mold 12 through the immersion nozzle 20. The molten metal 18 is heat-extracted from the inner surface of the mold 12 and solidifies, forming a solidified shell 22. As a result, a cast piece 26 is formed, which has the solidified shell 22 as an outer shell and an unsolidified layer 24 made of the molten metal 18 inside.
[0013] In the gaps between adjacent strand support rolls 16 in the casting direction, multiple secondary cooling zones 28, each equipped with spray nozzles (not shown), are installed along the casting direction from directly below the mold 12. The strand 26 is cooled as it is withdrawn by cooling water sprayed from the spray nozzles in the secondary cooling zones 28. While the strand 26 is transported by the strand support rolls 16 and passes through the multiple secondary cooling zones 28, the solidified shell 22 is appropriately cooled, solidification of the unsolidified layer 24 progresses, and solidification of the strand 26 is completed.
[0014] Downstream in the casting direction, a plurality of transport rolls 17 are installed for continuing to transport the slab 26. A slab cutter 30 is disposed above the transport rolls 17 for cutting the slab 26. After solidification is complete, the slab 26 is cut by the slab cutter 30, and slabs 26a of a predetermined length are continuously cast.
[0015] In such a continuous metal casting machine 10, a breakout may occur in which the molten metal 18 leaks to the outside from the solidified shell 22 formed in the cast slab 26. The molten metal leak detection device 40 according to this embodiment is used to detect this breakout in which the molten metal 18 leaks to the outside.
[0016] The molten metal leak detection device 40 includes a camera 42 and an image analyzer 44. The camera 42 performs an imaging step, capturing an image of the continuous metal casting machine 10 including the cast piece 26 below the mold 12, for example, every 100 msec to generate image data, and transmits the generated image data to the image analyzer 44. Here, 100 msec is an example of a predetermined time interval.
[0017] The cast slab 26 immediately after being withdrawn from the mold 12 has a thin solidified shell 22, making it prone to breakout. Therefore, it is preferable that the range imaged by the camera 42 be the range directly below the mold 12. However, depending on the configuration of the continuous metal casting machine 10, it may be difficult to image the range directly below the mold 12. In this case, it is sufficient to image the range below the mold 12 that is close to directly below the mold 12.
[0018] The image analysis device 44 acquires time-series data, which is the change over time in the number of pixels exhibiting a specific color, by measuring the number of pixels exhibiting a specific color contained in a predetermined region of the image data acquired from the camera 42. Here, the specific color is the color that represents the molten metal 18 contained in the predetermined region of the image data. The image analysis device 44 detects the leakage of molten metal using the time-series data. The camera 42 is an example of an imaging unit, and is, for example, a digital camera or video camera that has a CCD image sensor or a CMOS image sensor and is capable of generating color image data.
[0019] The molten metal leak detection device 40 may have two or more cameras 42 to capture images of the entire circumference below the mold 12. In this case, the image analysis device 44 acquires time-series data, which is the change over time in the number of pixels showing a specific color, from the image data acquired from each camera, and detects the leakage of the molten metal 18 using the time-series data.
[0020] Next, the image analysis device 44 will be described. Fig. 2 is a schematic diagram showing an example configuration of the image analysis device 44. The image analysis device 44 is, for example, a general-purpose computer such as a workstation or a personal computer. The image analysis device 44 has a control unit 46, an input unit 48, an output unit 50, and a storage unit 52.
[0021] The control unit 46 is, for example, a CPU, and executes various programs stored in the storage unit 52 to cause the control unit 46 to function as a data acquisition unit 54 and a detection unit 56. The input unit 48 is, for example, a keyboard, a touch panel integrated with a display, or the like. The output unit 50 is, for example, an LCD, a CRT display, or a patrol lamp. The storage unit 52 is, for example, an updatable flash memory, a built-in hard disk or a hard disk connected via a data communication terminal, an information recording medium such as a memory card, and a read / write device for the same. The storage unit 52 stores programs and data used to acquire time-series data from image data received from the camera 42 and detect leakage of the molten metal 18 using the time-series data.
[0022] Next, the processing executed by the data acquisition unit 54 and the detection unit 56 will be described. When the data acquisition unit 54 receives image data generated by the camera 42, it executes a data acquisition step and measures the number of pixels that exhibit a specific color included in a predetermined area of the image data. By repeatedly receiving image data and performing this measurement, the data acquisition unit 54 acquires time-series data that is the change over time in the number of pixels that exhibit a specific color.
[0023] Specifically, when the data acquisition unit 54 receives the image data, it reads information indicating the predetermined area from the storage unit 52 and identifies the predetermined area. The predetermined area is set to eliminate disturbances that inevitably enter the imaging area depending on the installation position of the camera. The predetermined area is set in advance during tuning performed when the molten metal leak detection device 40 is installed, and the information indicating the predetermined area is input from the input unit 48 and stored in the storage unit 52.
[0024] Furthermore, the data acquisition unit 54 converts the RGB values of each pixel into HSV values. This is because the range of the specific color is set as a range of HSV values. The range of the specific color may be set as a range of color space coordinates such as RGB values or HSV values. In this case, the data acquisition unit 54 converts the pixels so that they correspond to the color space in which the range of the specific color is set.
[0025] The data acquisition unit 54 reads information indicating the range of a specific color from the storage unit 52, compares the HSV values of all pixels included in the specified area with the information indicating the range, and measures the number of pixels that fall within the range of the specific color.
[0026] The range of the specific color is specified in advance during tuning that is performed when the molten metal leak detection device 40 is installed, and information indicating the range of the specific color is input from the input unit 48 and stored in the storage unit 52. The data acquisition unit 54 repeatedly receives image data and measures the number of pixels that exhibit the specific color, thereby acquiring time-series data that is the change over time in the number of pixels that exhibit the specific color.
[0027] The detection unit 56 executes a detection step and detects leakage of the molten metal 18 using time-series data, which is a change over time in the number of pixels showing a specific color acquired by the data acquisition unit 54. The detection unit 56 uses the time-series data to measure the duration during which the number of pixels showing a specific color continues to be equal to or greater than a reference number. The detection unit 56 reads information indicating the reference number and a threshold value from the storage unit 52, and detects leakage of the molten metal 18 when the duration during which the number of pixels showing a specific color continues to be equal to or greater than the reference number exceeds the threshold value. On the other hand, even if the number of pixels showing a specific color is equal to or greater than the reference number, the detection unit 56 does not detect leakage of the molten metal 18 when the duration during which the number of pixels showing a specific color continues to be equal to or greater than the reference number is not equal to or greater than a predetermined threshold value.
[0028] The reference number is, for example, 1000, and the threshold for the duration for which the reference number or more continues is, for example, 2 seconds. The reference number and threshold are set in advance during tuning that is performed when the molten metal leak detection device 40 is installed, and information indicating the reference number and information indicating the threshold are input from the input unit 48 and stored in the storage unit 52.
[0029] Figure 3 is a graph showing the change over time in the number of pixels of a specific color in a continuous steel casting machine. The horizontal axis of Figure 3 represents time (seconds), and the vertical axis represents the number of pixels (number). The solid line in Figure 3 represents the change over time in the number of pixels of a specific color when a breakout occurs in the continuous steel casting machine and molten steel leaks, while the dashed line in Figure 3 represents the change over time in the number of pixels of a specific color when no molten steel leaks from the slab but sparks fly off. The dotted line represents the reference number (1,000) of pixels of a specific color.
[0030] The color of flying sparks caused by splashes of molten steel during normal operation is the same as that of leaked molten steel. As shown in Figure 3, even when flying sparks of molten steel occur, the number of pixels of a specific color may instantaneously increase to the same number of pixels as when molten steel leaks. Therefore, if an attempt is made to detect a molten steel leak using the reference number of pixels of a specific color as a threshold, it can be seen that the occurrence of flying sparks of molten steel alone will result in a false detection of a molten steel leak.
[0031] In contrast, the molten metal leak detection device 40 according to this embodiment uses time-series data, which is a change over time in the number of pixels exhibiting a specific color, to detect a leak of molten metal 18 based on the duration for which the number of pixels exhibiting a specific color remains equal to or exceeds a reference number. As shown in Figure 3, flying sparks occur suddenly, whereas when a breakout occurs and molten steel leaks, the molten metal flows out continuously. Therefore, by using time-series data, which is a change over time in the number of pixels exhibiting a specific color, to detect a leak of molten metal 18 based on the duration for which the number of pixels exhibiting a specific color remains equal to or exceeds a reference number, it becomes possible to detect a leak of molten metal 18 while suppressing false detection of a leak of molten metal.
[0032] Referring again to Figure 2, the detection unit 56 detects a leak of molten metal 18 when the number of pixels showing a specific color continues to be equal to or greater than a reference number for a time period equal to or greater than a threshold. When the detection unit 56 detects a leak of molten metal 18, it displays a message indicating that molten metal 18 has leaked on the LCD or CRT display of the output unit 50, or turns on a patrol lamp. This makes it possible to alert those in the vicinity that molten metal 18 has leaked.
[0033] If a leak of molten metal 18 occurs and the molten metal 18 continues to be supplied to the mold 12, the molten metal 18 will continue to leak, causing the damage to worsen. Therefore, when a leak of molten metal 18 is detected by the detection unit 56, it is preferable to slide the sliding shutter 19 in the reverse direction to stop the supply of molten metal 18 to the mold 12. In this way, by stopping the supply of molten metal 18 when a leak of molten metal 18 is detected, the damage caused by the leak of molten metal 18 can be prevented from worsening in a continuous casting method using a continuous metal casting machine.
[0034] As described above, the molten metal leak detection device 40 according to this embodiment uses time-series data representing the change over time in the number of pixels showing a specific color to detect a leak of the molten metal 18 based on the duration for which the number of pixels showing a specific color remains equal to or exceeds a reference number. This makes it possible for the molten metal leak detection device 40 according to this embodiment to suppress false detection of a molten metal leak.
[0035] In the above embodiment, the detection unit 56 detects a leak of molten metal 18 based on the duration over which the number of pixels exhibiting a specific color remains equal to or greater than a reference number. However, this is not limiting. The detection unit 56 may use time-series data, which is a time change in the number of pixels exhibiting a specific color, to measure an integrated value of the number of pixels exhibiting a specific color and detect a leak of molten metal 18 based on this integrated value within a predetermined time (e.g., 0.5 seconds). In this case, the detection unit 56 integrates the time-series data and detects a leak of molten metal 18 when the integrated value exceeds a predetermined threshold. The threshold value within this predetermined time is also specified in advance during tuning performed when the molten metal leak detection device 40 is installed, and information indicating this threshold value is input from the input unit 48 and stored in the storage unit 52.
[0036] 3, when a flying spark of molten metal occurs, the increase in the number of pixels of a specific color is instantaneous, and therefore the number of pixels of the specific color rapidly decreases over time. Therefore, the leakage of molten metal 18 is detected by the integrated value of the number of pixels of a specific color in multiple image data. This makes it possible to detect the leakage of molten metal 18 while suppressing false detection due to flying sparks of molten metal 18, compared to detecting the leakage of molten metal 18 by the integrated value of the number of pixels of a specific color in a single image data.
[0037] Furthermore, by detecting a leak of molten metal 18 using the integrated value of the number of pixels of a specific color, it is possible to detect a leak of molten metal 18 in a shorter time than when detecting a leak of molten metal 18 based on the duration for which the number of pixels remains above a reference number. The integrated value of the number of pixels of a specific color in the five image data when molten metal 18 leaks is significantly different from the integrated value of the number of pixels of a specific color in the five image data when sparks fly from molten metal 18. Therefore, by detecting a leak of molten metal 18 using a threshold value that can distinguish between the two integrated values, for example, when image data is generated every 100 msec, it is possible to detect a leak of molten metal 18 0.5 seconds after the leak occurs. In this way, it can be seen that detecting a leak of molten metal 18 based on the integrated value of the number of pixels of a specific color in multiple image data allows a leak of molten metal 18 to be detected in a shorter time.
[0038] In the above embodiment, an example has been described in which the image data generated by the camera 42 is transmitted to the data acquisition unit 54, but this is not limiting. The image data generated by the camera 42 may be stored in the storage unit 52, and the data acquisition unit 54 may read the image data stored in the storage unit 52 and count the number of pixels that exhibit a specific color included in a predetermined area of the image data.
[0039] Next, an example will be described in which a molten metal leak detection device 40 was used to detect a leak of molten steel from a continuous casting machine when steel was continuously cast using the continuous metal casting machine 10 shown in Figure 1. In this example, 10 samples of data showing the transition of the number of pixels of a specific color when an image of flying sparks of molten steel generated during normal operation was captured, and 10 samples of data showing the transition of the number of pixels of a specific color when an image of a breakout (leak of molten steel) was captured were prepared. Breakouts were detected from the 10 samples prepared using the following method.
[0040] Specific color: Pixels with HSV values that satisfy H<45 or H>190, 50<S<150, and 50<V<150 were defined as specific colors (0≦H, S, V≦256). Conventional example: A breakout was detected when the number of pixels showing a specific color exceeded a reference number (1,000). Inventive example: A breakout was detected when the number of pixels showing a specific color remained above the reference number (1,000) for two seconds or more. The detection results for the conventional example and inventive example are shown in Table 1 below.
[0041]
[0042] As shown in Table 1, in the conventional example, flying sparks caused by molten steel splashes that occurred during normal operation were erroneously detected as breakouts in 4 / 10 cases. In contrast, in the inventive example, flying sparks caused by molten steel splashes that occurred during normal operation were never erroneously detected as breakouts. These results confirm that the use of the molten metal leak detection device according to this embodiment, which detects leaks of molten metal 18 using time-series data, which is the change over time in the number of pixels showing a specific color, can suppress erroneous detection of molten metal leaks.
[0043] REFERENCE SIGNS LIST 10 Continuous metal casting machine 12 Mold 14 Tundish 16 Strand support roll 17 Conveyor roll 18 Molten metal 19 Sliding shutter 20 Submerged nozzle 22 Solidified shell 24 Unsolidified layer 26 Strand 26a Strand 28 Secondary cooling zone 30 Strand cutting machine 40 Molten metal leak detection device 42 Camera 44 Image analysis device 46 Control unit 48 Input unit 50 Output unit 52 Storage unit 54 Data acquisition unit 56 Detection unit
Claims
1. A molten metal leak detection device for detecting leaks of molten metal from molten metal handling equipment, comprising: an imaging unit for imaging the molten metal handling equipment at predetermined time intervals to generate image data; a data acquisition unit for acquiring time series data on the number of pixels exhibiting a specific color included in a predetermined area of the image data; and a detection unit for detecting the occurrence of a molten metal leak using the time series data.
2. A molten metal leak detection device as described in claim 1, wherein the detection unit uses the time series data to measure the duration during which the number of pixels exhibiting a specific color remains equal to or greater than a reference number, and detects that the molten metal has leaked when the duration reaches or exceeds a predetermined threshold value.
3. A molten metal leakage detection device as described in claim 1, wherein the detection unit measures an integrated value of the number of pixels exhibiting a specific color using the time series data, and detects that the molten metal has leaked when the integrated value within a specified period of time becomes equal to or exceeds a predetermined threshold value.
4. A molten metal leak detection device as described in any one of claims 1 to 3, wherein the molten metal handling equipment is a continuous metal casting machine, and the imaging unit is provided in a position where it can image the area below the mold.
5. A molten metal leakage detection method for detecting leakage of molten metal from molten metal handling equipment, comprising: an imaging step of imaging the molten metal handling equipment at predetermined time intervals to generate image data; a data acquisition step of acquiring time series data on the number of pixels exhibiting a specific color contained in a predetermined area of the image data; and a detection step of detecting leakage of molten metal using the time series data.
6. A molten metal leakage detection method as described in claim 5, wherein in the detection step, the time series data is used to measure the duration during which the number of pixels exhibiting a specific color continues to be equal to or greater than a reference number, and when the duration reaches or exceeds a predetermined threshold, it is detected that the molten metal has leaked.
7. A molten metal leakage detection method as described in claim 5, wherein in the detection step, an integrated value of the number of pixels exhibiting a specific color is measured using the time series data, and if the integrated value becomes equal to or exceeds a predetermined threshold value within a specified time period, it is detected that the molten metal has leaked.
8. A molten metal leakage detection method according to any one of claims 5 to 7, wherein the molten metal handling equipment is a continuous metal casting machine, and the imaging step involves imaging an area below the mold to generate image data.
9. A method for continuous casting of metal, comprising the steps of: stopping the supply of molten metal to the mold when a molten metal leak is detected by the molten metal leak detection method as set forth in claim 8.