Method and device for diagnosing abnormalities in continuous casting machines

Abstract

The present invention provides an abnormality diagnosis method and an abnormality diagnosis device for a continuous casting machine that can accurately detect abnormalities in roll gaps. The abnormality diagnosis method for a continuous casting machine (1) detects abnormalities in roll gaps between pairs of rolls (3a) that face each other in the thickness direction of a slab in a support roll group that is made up of a plurality of support rolls (2) that sandwich a cast slab (S) and guide it in the casting direction within the continuous casting machine (1), in which a roll gap meter (11) attached to a dummy bar (9) that moves with the cast slab (S) being withdrawn measures the roll gaps between the plurality of roll pairs (3a) at a plurality of positions in the slab width direction, and detects abnormalities in the roll gaps using the roll gaps at the same position in the slab width direction for two roll pairs (3a) of the same type, in which the support rolls (3) have a plurality of split rolls and roll chocks, and the two roll pairs (3a) of the same type have at least the same bearing positions.

Description

[Technical Field]

[0001] The present invention relates to an abnormality diagnosis method and an abnormality diagnosis device for a continuous casting machine. [Background technology]

[0002] A group of guide rolls (support rolls) arranged on the outlet side of the mold in a continuous casting machine supports and guides the cast slab by pairs of rolls that sandwich the cast slab in the thickness direction during drawing. In such guide roll groups, the roll gap between the roll pairs is strictly controlled to achieve stable continuous casting while avoiding bulging and internal cracks that lead to quality degradation of the cast slab.

[0003] Typically, the roll gap of the guide roll group is measured for each casting run using a measuring device attached to a dummy bar, which is a piece of equipment that guides the initial casting portion of the cast slab cast in the mold to the pinch rolls located downstream of the continuous casting machine.

[0004] As prior art related to this point, Patent Document 1 discloses a method for determining whether or not there is an abnormality in the roll spacing by comparing the peak value of the roll spacing between each roll obtained by a roll spacing meter with a preset roll spacing.

[0005] Furthermore, Patent Document 2 discloses a method for measuring the roll spacing at at least two points for a pair of rolls, comparing the difference between the measurements at the two points with the previous value, and determining that the roll spacing is abnormal if the difference exceeds a threshold value. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-245531 [Patent Document 2] Japanese Patent Application Laid-Open No. 2006-247687 Summary of the Invention [Problem to be solved by the invention]

[0007] However, with the above-mentioned conventional method, the roll gap measurement may not be performed correctly due to reasons such as the position of the measuring instrument being shifted due to meandering of the dummy bar during the withdrawal movement, or the dummy bar riding up onto the roll chock, and this may result in the roll gap being recognized as abnormal.

[0008] Furthermore, if the deviation in the roll gap exceeds a threshold value due to deviations and variations in the measured values ​​that depend on the characteristics of the sensor and the measurement environment, it may be determined that an abnormality in the roll gap has occurred.

[0009] Furthermore, even when an abnormality has been detected, there have been many cases where an actual inspection of the support rolls of the continuous casting equipment revealed that there was no abnormality in the rolls.

[0010] Therefore, the present invention has been made in consideration of the above-mentioned problems, and aims to provide an abnormality diagnosis method and an abnormality diagnosis device for a continuous casting machine that can accurately detect abnormalities in the roll gap. [Means for solving the problem]

[0011] (1) According to one aspect of the present invention, there is provided an abnormality diagnosis method for a continuous casting machine that detects abnormalities in the roll spacing between pairs of rolls that face each other in the thickness direction of a slab in a support roll group that is composed of a plurality of support rolls that sandwich a cast slab and guide it in the casting direction within the continuous casting machine, wherein the roll spacing between the plurality of roll pairs is measured at a plurality of positions in the width direction of the slab using a roll spacing meter that is installed on a dummy bar that moves in conjunction with the withdrawal movement of the cast slab, and the roll spacing abnormality is detected for two roll pairs of the same type using the roll spacing at the same position in the width direction of the slab, wherein the support rolls have a plurality of split rolls and roll chocks, and the two roll pairs of the same type have at least the same bearing positions.

[0012] (2) In the abnormality diagnosis method for a continuous casting machine according to (1) above, the two roll pairs of the same type are roll pairs in the same segment.

[0013] (3) In the method for diagnosing an abnormality in a continuous casting machine according to (1) or (2), when detecting an abnormality in the roll gap, the abnormality in the roll gap is detected using current comparison difference data, which is comparison difference data whose element is the difference in the roll gap measured at the same position of two roll pairs of the same type.

[0014] (4) In the method for diagnosing abnormalities in a continuous casting machine described in (3) above, the roll gap measurement is repeatedly performed at the timing when continuous casting by the continuous casting machine is started, and abnormalities in the roll gap are detected using the current comparison difference data and past comparison difference data, which is comparison difference data measured in multiple past continuous castings.

[0015] (5) In the method for diagnosing an abnormality in a continuous casting machine described in (4) above, the past comparison difference data is consecutive comparison difference data from the most recent multiple times.

[0016] (6) In the method for diagnosing abnormalities in a continuous casting machine described in (5) above, vector data having elements of data (n1 x n3 points, where n3 = n2 + 1) is calculated from multiple points (n1 points) of data in the current comparison difference data and multiple points (n1 x n2 points) of data in the comparison difference data measured in the most recent multiple (n2) continuous castings, and abnormalities are detected by statistically evaluating the degree of deviation of the vector data from the normal state.

[0017] (7) According to one aspect of the present invention, there is provided an abnormality diagnosis device for a continuous casting machine that detects abnormalities in the roll spacing between pairs of rolls that face each other in the thickness direction of a slab in a support roll group that is composed of a plurality of support rolls that sandwich the cast slab and guide it in the casting direction inside the continuous casting machine, the abnormality diagnosis device comprising: a roll spacing meter that is installed on a dummy bar that moves in conjunction with the withdrawal movement of the cast slab and that measures the roll spacing between the plurality of roll pairs at a plurality of positions in the width direction of the slab; and a diagnostic unit that detects abnormalities in the roll spacing for two pairs of rolls of the same type using the roll spacing at the same position in the width direction of the slab, wherein the support rolls have a plurality of split rolls and roll chocks, and the two pairs of rolls of the same type have at least the same bearing positions. [Effects of the Invention]

[0018] According to one aspect of the present invention, there are provided an abnormality diagnosis method and an abnormality diagnosis device for a continuous casting machine, which are capable of accurately detecting abnormalities in the roll gap. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a schematic diagram showing a continuous casting machine according to an embodiment of the present invention. [Figure 2] FIG. 2 is a schematic diagram illustrating the configuration of a support roll. [Figure 3] FIG. 10 is a side view showing a method for measuring the roll gap using a roll gap meter. [Figure 4] FIG. 1 is a schematic diagram illustrating the concept of a Q value. [Figure 5] 10 is a graph showing the trend of the roll gap deviation and the Q value when a malfunction occurs. [Figure 6] 10 is a graph showing the trend of the deviation of the roll interval and the Q value when a false detection occurs. [Figure 7] 10 is a graph showing the trend of deviation in roll spacing when a false detection occurs, where (A) shows a pair of rolls where a false detection occurred, and (B) shows a pair of rolls that are the same type as the support roll 3 in (A) and are close to each other in the same segment. [Figure 8] 10 is a graph showing the number of false detections in an example and a comparative example. [Figure 9] 10 is a graph showing the abnormality detection rate and the breakdown of detection in an example and a comparative example. DETAILED DESCRIPTION OF THE INVENTION

[0020] In the following detailed description, embodiments of the present invention will be described with reference to the drawings. In the description of the drawings, identical or similar parts are designated by identical or similar reference numerals, and redundant explanations will be omitted. The drawings are schematic and may differ from the actual product. Furthermore, the embodiments shown below exemplify devices and methods for embodying the technical concept of the present invention, and the technical concept of the present invention does not specify the materials, structure, arrangement, etc. of component parts as described below. The technical concept of the present invention can be modified in various ways within the technical scope defined by the claims.

[0021] <Configuration of a continuous casting machine> The present invention will now be described in more detail with reference to the drawings. Fig. 1 is a diagram schematically illustrating a continuous casting machine 1 according to one embodiment of the present invention.

[0022] The continuous casting machine 1 includes a continuous casting mold 2, a plurality of support rolls 3, a plurality of segments 4, a tundish 5, a sliding nozzle 6, a submerged nozzle 7, a pinch roll 8, a dummy bar 9, and an abnormality diagnosis device 10.

[0023] The multiple support rolls 3 constitute a support roll group, which sandwich the cast slab S and guide it in the casting direction within the continuous casting machine 1. The casting direction is the longitudinal direction of the cast slab S in FIG. 1, and is the direction in which the cast slab S including the solidified shell moves within the continuous casting machine 1. Furthermore, within the continuous casting machine 1, the width direction of the cast slab S (the front-to-back direction in FIG. 1) is referred to as the slab width direction, and the thickness direction of the cast slab S is referred to as the slab thickness direction. Furthermore, the two support rolls 3 facing each other in the slab thickness direction are also referred to as a roll pair 3a.

[0024] The support rolls 3 each have a plurality of split rolls 31 and at least one roll chock (bearing) 32. FIG. 2 shows an example of the structure of the support rolls 3 in one segment 4. In the example shown in FIG. 2, five roll pairs 3a are provided in the segment 4, and only five support rolls 3 on one side of the roll pairs 3a in the thickness direction of the slab are shown in FIG. 2. The five support rolls 3 on the other side of the thickness direction of the slab also have the same shape as in FIG. 2. As shown in FIG. 2, the support rolls 3A to 3E each have three split rolls 31 and a plurality of roll chocks 32 provided at the ends of the split rolls 31. The outer diameter, material, etc. of each split roll 31 are all the same. In the segment 4 shown in FIG. 2, the support rolls 3A to 3E are classified into two types. Here, the type of the support rolls 3 is classified at least based on the bearing position, which is the position of the roll chocks 32. In the example of FIG. 2, the support rolls 3A, 3C, and 3E are roll pairs 3a of the same type having the same bearing positions, and the support rolls 3B and 3D are roll pairs 3a of the same type having the same bearing positions.

[0025] The segment 4 is a frame body in which a plurality of roll pairs 3a are provided. A group of support rolls is formed in the continuous casting machine 1 by incorporating a plurality of segments 4. The segment 4 is also provided with a cooling means (not shown) that injects cooling water.

[0026] The tundish 5 is an intermediate vessel that holds molten steel (molten metal) M transferred from a ladle (not shown). The sliding nozzle 6 is set at the bottom of the tundish 5, and the flow rate of the molten steel M is adjusted by adjusting its opening. The submerged nozzle 7 is set on the underside of the sliding nozzle 6. The molten steel M held in the tundish 5 is poured into the mold 2 through the sliding nozzle 6 and the submerged nozzle 7.

[0027] The pinch rolls 8 are provided downstream in the casting direction of the continuous casting machine 1, and rotate to withdraw the cast slab S in the casting direction. The dummy bar 9 is used in the early stages of continuous casting, and is equipment that withdraws the initial casting portion of the cast slab S cast in the mold 2 to the pinch rolls 8, and moves along with the withdrawal movement of the cast slab S.

[0028] The dummy bar 9 is also provided with a roll spacing meter 11 that measures the roll spacing, which is the distance between the roll pairs 3a in the thickness direction of the slab. The roll spacing meter 11 is a measuring means that measures the roll spacing using a known measuring means such as a rotary encoder or a differential transformer. A plurality of roll spacing meters 11 are provided in the width direction of the slab, and measure the roll spacing at multiple positions in the width direction of the slab of the roll pair 3a. For example, three roll spacing meters 11 are provided on the dummy bar 9, and measure the roll spacing at three positions in the width direction of the slab. As shown in FIG. 3 , the roll spacing meter 11 measures the roll spacing of the roll pair 3a by coming into contact with the roll pair 3a when the dummy bar 9 passes through the support roll group.

[0029] The abnormality diagnosis device 10 is a means for detecting an abnormality in the roll spacing from the roll spacing measurement results obtained by the roll spacing meter 11, and is configured, for example, by a computer. The abnormality diagnosis device 10 includes an acquisition unit 100, a storage unit 101, and a diagnosis unit 102. When the abnormality diagnosis device 10 is a computer, an arithmetic processing device such as a CPU (Central Processing Unit) executes a program to function as the acquisition unit 100 and the diagnosis unit 102. The acquisition unit 100 acquires the roll spacing measurement results obtained by the roll spacing meter 11 (roll spacing data) and, as necessary, acquires operating conditions and the like from a host computer. The storage unit 101 stores the roll spacing data acquired by the acquisition unit 100, as well as data necessary for abnormality determination, such as past roll spacing data obtained under normal conditions and the results of principal component analysis based on the roll spacing data under normal conditions, which will be described later. The storage unit 101 may be, for example, a primary storage device such as a RAM (Random Access Memory) or a secondary storage device capable of storing data, such as a hard disk drive (HDD) or a solid state drive (SSD). The diagnosing unit 102 diagnoses the roll gap abnormality based on the roll gap data. The method of diagnosing the roll gap abnormality by the diagnosing unit 102 will be described in detail later.

[0030] In this embodiment, the continuous casting machine 1 is a vertical bending type. In this continuous casting machine 1, continuous casting begins when molten steel M is poured into the mold 2 with a portion of the head of the dummy bar 9 inserted into the bottom of the mold 2. The molten steel M is then cooled in the mold 2 to form a solidified shell, resulting in a semi-solidified cast slab S. The dummy bar 9 is then withdrawn in the casting direction, resulting in the semi-solidified cast slab S being withdrawn in the casting direction. The cast slab S then passes through a group of support rolls and is secondarily cooled by a cooling means until it is completely solidified. When the tip of the cast slab S downstream in the casting direction reaches the pinch roll 8, it is separated from the dummy bar 9 and then cut to a predetermined length to become a cast slab or the like.

[0031] <Abnormality diagnosis method> Next, an abnormality diagnosis method for the continuous casting machine 1 according to this embodiment will be described. In this embodiment, first, the roll gap meter 11 measures the roll gap of each roll pair 3a when the dummy bar 9 passes through the support rolls at the beginning of casting (measurement process). In the measurement process, the roll gap meter 11 measures the roll gap of each roll pair 3a at multiple positions in the slab width direction. In this embodiment, for example, the roll gap meter 11 measures the roll gap of each roll pair 3a at three positions in the slab width direction. The measurement results are transmitted to the acquisition unit 100 and stored in the storage unit 101. The measurement process ends when the roll gaps of all roll pairs 3a for which abnormality diagnosis of the roll gap is to be performed have been measured. Note that, as an example, the following will describe abnormality diagnosis of the roll gap of the support rolls 3 included in one segment 4 shown in FIG. 2. In the example shown in FIG. 2, three positions P in the slab width direction indicated by dashed lines are measured. a ,P b ,P c That is, for example, for the support roll 3A, the roll gap is measured at a position where the support roll 3A faces the other support roll 3 (not shown) forming the roll pair 3a, and at three positions P a ,P b ,P c Position P, which is the intersection point with a1 ,P b1 ,P c1 The roll gap is measured at each point.

[0032] After the measurement step is completed, the diagnosis unit 102 detects the presence or absence of an abnormality in the roll gap for two roll pairs 3a of the same type in the same segment 4, using the roll gap measured in the measurement step at the same position in the slab width direction (diagnosis step). The two roll pairs 3a of the same type are not particularly limited as long as they are in the same segment 4, but in order to improve the accuracy of detecting abnormalities, it is preferable to use roll pairs 3a that are close to each other. For example, in the case of Figure 2, the two roll pairs 3a of the same type may be the roll pair 3a of backing roll 3A and backing roll 3E, but it is preferable to use the roll pair 3a of backing roll 3A and backing roll 3C (or backing roll 3C and backing roll 3E).

[0033] In the diagnosis process, first, the difference in roll gap between two roll pairs 3a of the same type at the same position in the slab width direction is calculated. The difference in roll gap between two roll pairs 3a of the same type at the same position in the slab width direction is calculated at a plurality of positions in the width direction, preferably at all positions, measured in the measurement process. Note that data that uses this difference in roll gap as an element is also called comparative difference data. For example, in the case shown in Figure 2, the roll pair 3a (3a1) including the support roll 3A and the roll pair 3a (3a3) including the support roll 3C are roll pairs 3a of the same type. Furthermore, in the measurement process, the difference in roll gap between the roll pair 3a1 and the roll pair 3a3 is calculated at position P a1 ,P b1 ,P c1 Roll spacing (e.g., d a1 ,d b1 ,d c1 ) and position P a3 ,P b3 ,P c3 Roll spacing (e.g., d a3 ,d b3 ,d c3 The difference in the roll gap at the same position in the width direction of the slab (for example, the position P a1 Roll spacing and position P a3 Difference between the roll spacing at (d a1 -d a3 )) is calculated. In addition, the calculation of the difference in the roll interval is performed at the position P a ,P b ,P c (In the above example, for example, Δd a =d a1 -d a3 ,Δd b =d b1 -d b3 ,Δd c =d c1 -d c3 Here, the comparison difference data calculated in the diagnosis process is also referred to as current comparison difference data.

[0034] Next, feature quantities are extracted from the calculated current comparative difference data and the most recent past comparative difference data. The most recent past comparative difference data is comparative difference data based on the roll gap differences measured in the most recent past multiple continuous castings, and is for the same support roll 3 and slab width direction position as the target for abnormality diagnosis. The most recent past comparative difference data, like the current comparative difference data, includes roll gap differences at multiple positions in the slab width direction. Furthermore, the most recent past comparative difference data uses comparative difference data collected in the most recent past multiple continuous castings, but it is preferable to use comparative difference data collected in at least the most recent two continuous castings, and to use three comparative difference data in total, including the current comparative difference data. As past comparative difference data, data from up to the most recent five castings may be sufficient. The above Δd a For example, when the most recent past is considered to be up to two measurements before the current comparison difference data, the index representing the current comparison difference data is k. Then, the index of the most recent previous comparison difference data is k-1, and the index of the comparison difference data before the most recent previous data is k-2. Then, each comparison difference data is Δd a (k),Δd a (k-1),Δd a It can be written as (k-2).

[0035] In the continuous casting machine 1, casting is performed using a dummy bar 9 every time continuous casting is started, so the roll gap is measured and comparison difference data is calculated every time continuous casting is started. In addition, in so-called serial casting operations in which continuous casting is performed continuously using multiple ladles, measurement is performed only when a dummy bar is inserted during the first continuous casting. Past comparison difference data is stored in the memory unit 101.

[0036] For extracting feature quantities, it is preferable to use dimension reduction by principal component analysis. Specifically, a data vector is calculated by combining multiple points (n1 points) of data in the current comparison difference data with multiple points of data in the comparison difference data obtained during multiple past (n2) continuous castings. The number of measurements, including the current measurement and the multiple most recent past measurements (n2 times), is n3 (=n2+1). Since the number of points of comparison difference data is n1, the data vector is vector data with data (n1 × n3 points) as elements.

[0037] In the above example, n1 and n3 are both 3. In other words, the comparison difference data (Δd a ,Δd b ,Δd c ), the index of the current comparison difference data is k, and the indexes of the two most recent comparison difference data are k-1 and k-2, respectively, to create vector data that includes data from the three data acquisition timings. If this vector data is represented as ΔD, then in this example, ΔD is composed of the following elements: ΔD=(Δd a (k) Δd b (k) Δd c (k) Δd a (k-1) Δd b (k-1) Δd c (k-1) Δd a (k-2) Δd b (k-2) Δd c (k-2) T (T stands for transpose.)

[0038] Then, for ΔD, principal component analysis is performed in advance using vector data obtained during past operations under normal conditions with no abnormalities in the roll gap. From the principal component vectors (n1 × n3) obtained as the analysis results, one or more principal component vectors (fewer than n1 × n3, if multiple) that capture the characteristics of the normal state are determined as feature vectors (hereinafter simply referred to as feature vectors). This method is called dimensionality reduction because it represents features with fewer than n1 × n3 vectors. Furthermore, a hyperplane used to calculate Q statistics (also referred to as Q value) is calculated using the feature vectors, which are principal component vectors that capture the characteristics of the normal state previously determined. The coupling coefficients of the principal component vectors used to calculate this hyperplane are also calculated. Data from these analysis results is stored in the storage unit 101. Note that principal component analysis and calculation of Q statistics are well-known statistical methods, and for details, refer to statistical analysis literature.

[0039] Furthermore, the presence or absence of an abnormality in the roll spacing is diagnosed by evaluating the degree of deviation of the detected comparison difference data vector ΔD of the roll spacing and the feature amount from a normal state in which there is no abnormality in the roll spacing.

[0040] The evaluation of the deviation from the normal state, that is, the diagnosis of an abnormality in the roll spacing, can be performed as follows. As shown in FIG. 4, the distance between a hyperplane set using feature quantities (principal component vectors) previously extracted under normal conditions and vector data (the above-mentioned ΔD) including comparison difference data measured at multiple locations this time and corresponding comparison difference data from the most recent past is defined as the Q value. This Q value is then compared with a threshold, and if the Q value exceeds the threshold, it is diagnosed that there is an abnormality in the roll spacing. Furthermore, at least one of the two roll pairs 3a is diagnosed as having an abnormality. The threshold can be set appropriately depending on the desired abnormality detection accuracy. For example, by setting a low threshold, it is possible to predict an abnormality in the roll spacing. Furthermore, by normalizing the calculated Q value by the top 80% of all Q values, the threshold can be uniformly set to the normalized value.

[0041] The above-mentioned abnormality diagnosis of the roll gap is preferably performed for all pairs of two rolls 3a of the same type that are close to each other within the same segment 4. For example, in the case of Fig. 2, it is preferable to perform abnormality diagnosis for each of the combinations of support roll 3A and support roll 3C, the combination of support roll 3B and support roll 3D, and the combination of support roll 3C and support roll 3E.

[0042] In the case of a vertical bending continuous casting machine 1 as in this embodiment, the support rolls 3 are divided into a vertical section, a curved section, and a horizontal section, as shown in FIG. 1 . Therefore, when measuring the roll spacing using a roll spacing meter 11, the measurement environment changes depending on the position of the support roll 3, resulting in bias in the measured values ​​depending on the position of the support roll 3. However, if the support rolls 3 are normal, the measured values ​​of support rolls 3 located close to each other, i.e., in the same measurement environment, will change in synchronization. Conversely, if the support roll 3 is abnormal, this synchronization will be lost. In contrast, in this embodiment, multiple roll spacing measurements of the same support roll 3 located close to each other are used, and the synchronized changes between the multiple measured values ​​are captured as a characteristic. The roll spacing data of the target support roll 3 is compared with the roll spacing data of a normal support roll. This makes it possible to determine whether the support roll 3 is normal or abnormal.

[0043] In this embodiment, the roll spacing between two roll pairs 3a that are close to each other in the same segment 4 and have the same type of bearing positions is treated as a single data set, and the comparison difference data is used to detect abnormalities. In this configuration, the difference between the roll spacing data between two roll pairs 3a that are close to each other in the same segment 4 and have the same type of bearing positions is calculated. This makes it possible to eliminate measurement abnormalities caused by the physical shape of the divided rolls 31, such as misalignment of the roll spacing meter 11 or riding on the roll chocks 32, and improves the accuracy of abnormality detection. Furthermore, in this embodiment, the accuracy of abnormality detection can be further improved by using the current comparison difference data and data from multiple recent previous measurements. Furthermore, according to this embodiment, the reduced number of false detections reduces the frequency of workers entering the facility when an abnormality occurs.

[0044] Furthermore, in conventional methods such as those described in Patent Documents 1 and 2, it is necessary to separately monitor three items: a comparison between the front and rear roll pairs, a comparison of time-series data of measurement values ​​of the pair of rolls, and a comparison of measurement results at multiple locations on the pair of rolls, which places a heavy burden on the monitor. However, according to this embodiment, all of these items can be monitored automatically, thereby reducing the burden on the monitor.

[0045] Furthermore, in conventional methods such as those described in Patent Documents 1 and 2, if a failure is detected, sudden replacement of the support rolls is required, and the time required to urgently arrange for construction personnel is currently a major cause of lost production time for the continuous caster. However, according to this embodiment, by performing a diagnosis using the Q value, it is possible to predict abnormalities in the roll gap earlier than with the current judgment method that sets a threshold value for the deviation in the roll gap, and construction personnel can be arranged with ample time to spare, thereby reducing lost production time.

[0046] <Modification> Although the present invention has been described above with reference to specific embodiments, it is not intended that the invention be limited by these descriptions. By referring to the description of the present invention, other embodiments of the present invention that include various modifications in addition to the disclosed embodiments will be apparent to those skilled in the art. Therefore, it should be understood that the embodiments of the invention set forth in the claims also encompass embodiments that include these modifications described herein, either alone or in combination.

[0047] For example, in the above embodiment, the continuous casting machine 1 is of a vertical bending type, but the present invention is not limited to this example. For example, the continuous casting machine 1 may be of another type, such as a curved type.

[0048] In the above embodiment, the measurement step is performed when the dummy bar 9 passes through the support rolls in the early stage of casting, but the present invention is not limited to this example. For example, if the roll gap can be measured at the stage when the dummy bar 9 is inserted into the continuous casting machine 1 before continuous casting, the measurement may be performed before continuous casting.

[0049] Furthermore, in the above embodiment, one segment 4 has two types of roll pairs 3a, but the present invention is not limited to this example. One segment 4 may have a plurality of roll pairs 3a of each type. For example, one segment 4 may have three types of roll pairs 3a. [Example]

[0050] Next, we will describe examples conducted by the present inventors. In the examples, the accuracy of abnormality detection by applying the abnormality diagnosis method of the above-described embodiment was verified using roll gap measurement data from eight months of past continuous casting in a continuous casting machine 1 equipped with 158 roll pairs 3a, each consisting of support rolls 3 with diameters of 320 to 345 mm. In the past measurements, the roll gap was measured at three different locations in the slab width direction for the roll pairs 3a. Furthermore, in the examples, for two roll pairs 3a of the same type and closest to each other in the same segment 4, Q statistics were calculated using a data vector created by combining current comparison difference data and multiple recent past data, as in the above-described embodiment. The past comparison difference data used were the measurement data from the two most recent measurements. Furthermore, as a comparative example, abnormalities were also diagnosed using a conventional abnormality diagnosis method similar to that described in Patent Document 1. Specifically, in the comparative example, the difference between the measured roll gap and a reference value was calculated, and the presence or absence of an abnormality was determined by determining whether this difference exceeded a threshold.

[0051] FIG. 5 shows the roll gap deviation and Q value for a support roll 3 where a malfunction has occurred. Note that in FIG. 5 and FIG. 6 described below, the deviation is the difference between a conventional roll gap measurement value (measurement value for one target roll pair 3a) and a reference value. In the malfunction, the Q value also increased at the timing when the malfunction was confirmed, confirming that an abnormality in the roll gap can be detected by setting a threshold value. FIG. 6 also shows the deviation and Q value when a roll abnormality is falsely detected. As shown in FIG. 6, the deviation is excessively large at the timing when the false detection occurs, but it was confirmed that the Q value does not change at the same timing. Therefore, it was confirmed that the abnormality diagnosis method according to the above embodiment can accurately detect abnormalities in the roll gap.

[0052] FIG. 7 shows the trend of roll gap deviation when a false detection occurs. FIG. 7(A) shows the trend for a back roll 3 where a false detection occurred, and FIG. 7(B) shows the trend of roll gap deviation for a back roll 3 of the same type as the back roll 3 in FIG. 7(A) and close to it in the same segment 4, over the same period as FIG. 7(A). In FIGS. 7(A) and (B), the circled timings indicate the timings at which false detection occurred. In addition, in FIGS. 7(A) and (B), three types of data are plotted at three positions in the slab width direction. As shown in FIG. 7(A), the deviation exceeds the threshold at the circled timings. However, the deviation tends to increase with the passage of measurement time, and a similar trend can be seen in the deviation of a back roll 3 of the same type shown in FIG. 7(B). In other words, it can be confirmed that false detection can be reduced by considering the time change in the measured roll gap value and the measured value of a back roll of the same type, as in the above embodiment.

[0053] 8 and 9 show the results of applying the anomaly diagnosis method according to the embodiment to all roll spacing anomalies and false positives that occurred over an eight-month period. In FIG. 9, the results for anomaly prediction indicate cases in which anomalies were detected two or more days earlier than the comparative example. As shown in FIG. 8, by performing diagnosis using the Q value, it was confirmed that the number of false positives was significantly reduced in the embodiment compared to the comparative example. Furthermore, as shown in FIG. 9, it was confirmed that the anomaly detection rate in the embodiment was 100%, which was higher than the comparative example, and that roll spacing anomalies could be detected earlier than the comparative example. In other words, the anomaly diagnosis method according to the embodiment can significantly reduce false positives and detect anomalies that could not be detected by conventional methods, thereby improving detection accuracy. Additionally, it can also detect roll spacing anomalies earlier than conventional methods. [Explanation of symbols]

[0054] 1. Continuous casting machine 2. Mold 3 Support Roll 3a Roll vs. 31 Divided Roll 32 Roll Chock 4 segments 5 Tundish 6 Sliding Nozzle 7 Submerged Entry Nozzle 8 Pinch Roll 9 Dummy Bar 10. Abnormality diagnosis device 100 Acquisition Department 101 Storage section 102 Diagnostic Department 11 Roll spacing meter S Casting M Molten steel

Claims

1. A method for diagnosing an abnormality in a continuous casting machine, which detects an abnormality in the roll gap between a pair of rolls facing each other in a thickness direction of a slab in a support roll group consisting of a plurality of support rolls that sandwich a cast slab and guide it in a casting direction inside the continuous casting machine, measuring roll intervals between the plurality of roll pairs at a plurality of positions in the width direction of the slab using a roll interval meter installed on a dummy bar that moves in association with the withdrawal movement of the cast slab; detecting an abnormality in the roll gap using the roll gap at the same position in the slab width direction for two roll pairs of the same type; The support roll has a plurality of divided rolls and roll chocks, The two roll pairs of the same type have the same bearing positions at least in the axial direction of the rolls.

2. The method for diagnosing an abnormality in a continuous casting machine according to claim 1 , wherein the two roll pairs of the same type are roll pairs in the same segment.

3. 3. The method for diagnosing an abnormality in a continuous casting machine according to claim 1, wherein, when detecting an abnormality in the roll gap, the abnormality in the roll gap is detected using current comparison difference data, which is comparison difference data in which the difference in the roll gap measured at the same position of the two roll pairs of the same type is used as an element.

4. The measurement of the roll gap is repeatedly performed at the timing when continuous casting of the continuous casting machine is started, 4. The method for diagnosing an abnormality in a continuous casting machine according to claim 3, wherein the roll gap abnormality is detected using the current comparison difference data and past comparison difference data, which is comparison difference data measured in multiple past continuous castings.

5. The abnormality diagnosis method for a continuous casting machine according to claim 4, wherein the past comparison difference data is consecutive comparison difference data from a plurality of most recent times.

6. The plurality of points (n １ The data of the most recent multiple times (n ２ The comparison difference data is measured at multiple points (n １ ×n ２ From the data of points, data (n １ ×n ３ points, but n ３ = n ２ +1) as an element, The method for diagnosing an abnormality in a continuous casting machine according to claim 5, wherein the degree of deviation of the vector data from a normal state is statistically evaluated to detect an abnormality.

7. An abnormality diagnosis device for a continuous casting machine that detects an abnormality in the roll gap between pairs of rolls that face each other in a thickness direction of a slab in a support roll group that is composed of a plurality of support rolls that sandwich a cast slab and guide it in a casting direction inside the continuous casting machine, a roll gap meter that is installed on a dummy bar that moves along with the withdrawal movement of the cast strand and that measures the roll gaps between the plurality of roll pairs at a plurality of positions in the width direction of the strand; a diagnostic unit that detects an abnormality in the roll gap using the roll gap at the same position in the slab width direction for two roll pairs of the same type; Equipped with The support roll has a plurality of divided rolls and roll chocks, The two roll pairs of the same type have the same bearing positions at least in the axial direction of the rolls.

Images