Automatic monitoring method and system for steel ladle

Abstract

The invention relates to the field of intelligent detection of a smelting technology, in particular to an automatic monitoring method and system for a steel ladle. An automatic monitoring method for a steel ladle comprises the steps that three-dimensional modeling is conducted on the steel ladle in production circulation, and a three-dimensional point cloud model of the steel ladle is obtained; the steel ladle three-dimensional point cloud model and a preset reference model are positioned and matched, so that the steel ladle three-dimensional point cloud model and the reference model are aligned based on the same standard; dividing the steel ladle three-dimensional point cloud model into n layers along the depth direction, and calculating the radius of each layer according to the reference to obtain the actual radius of each point on the inner wall of the current steel ladle; obtaining the residual thickness value of the refractory brick through the actual radius and the initial radius of the corresponding point in the reference model; the steel ladle three-dimensional point cloud model is divided into multiple layers in the depth direction, the radius of each layer is calculated, the actual radius of each point of the inner wall of the current steel ladle can be obtained, and therefore comprehensive and accurate measurement of the residual thickness of the refractory bricks on the inner wall of the whole steel ladle is achieved.

Description

Technical Field

[0001] The present application relates to the field of intelligent detection of smelting technology, and in particular to an automatic monitoring method and system for a ladle. Background Art

[0002] In the steel production process, the refining of molten steel is a crucial link. When traditional steel mills produce molten steel, they widely use ladles made of refractory bricks to carry high-temperature molten steel. Refractory bricks not only need to withstand the high temperature of molten steel (close to 1000 degrees Celsius), but also need to resist the erosion of various chemical components in the molten steel to ensure the safety and continuity of production. However, after repeated use, refractory bricks will gradually become thinner due to high temperature and chemical erosion. Once the thickness drops below the safety critical value, molten steel may penetrate the ladle, causing serious production interruptions and safety hazards.

[0003] At present, the detection of refractory brick residual thickness faces multiple challenges. First, the thickness of refractory bricks usually ranges from 180 to 230 mm. In order to prevent accidents of puncturing the ladle and ensure maximum cost-effectiveness, the detection accuracy of the residual thickness is extremely high. However, the high temperature environment inside the ladle greatly limits the selection and application of detection methods. Traditional detection methods mainly rely on manual observation, that is, operators judge the degree of corrosion of refractory bricks by naked eyes. This method is not only highly subjective, but also difficult to ensure the accuracy and consistency of the test results.

[0004] Another common detection method is to use infrared detection equipment. However, in the detection of refractory brick residual thickness, due to the complex light and heat environment inside the ladle and the thermal conductivity characteristics of the refractory material itself, infrared detection is often difficult to accurately distinguish the small temperature difference caused by the thickness change, resulting in insufficient reliability of the detection results. In addition, infrared detection equipment also has limitations for the detection of refractory brick residual thickness in all directions. Usually, it can only measure a few preset points, making it difficult to achieve a comprehensive and accurate measurement of the residual thickness of the refractory bricks on the entire inner wall of the ladle.

[0005] Based on this, this application is filed. Summary of the invention

[0006] In view of the above technical problems, the purpose of the present application is to provide an automatic monitoring method and system for a ladle.

[0007] In the first aspect of the present application, an automatic monitoring method for a ladle is provided, comprising: performing three-dimensional modeling on a ladle in production flow to obtain a three-dimensional point cloud model of the ladle; positioning and matching the three-dimensional point cloud model of the ladle with a preset reference model so that the three-dimensional point cloud model of the ladle and the reference model are aligned based on the same reference; dividing the three-dimensional point cloud model of the ladle into n layers along the depth direction, and calculating the radius of each layer according to the reference to obtain the actual radius of each point on the inner wall of the current ladle; and obtaining the residual thickness value of the refractory brick through the actual radius and the initial radius of the corresponding point in the reference model.

[0008] In a further scheme of the present application, three-dimensional modeling is performed on the ladle in the production flow to obtain a three-dimensional point cloud model of the ladle, including: receiving a position signal when the ladle in the flow arrives at a preset process, controlling at least one acquisition device to obtain point cloud images of the ladle in at least two orientations; splicing the point cloud images in at least two orientations to obtain original point cloud data containing the ladle; performing regional boundary segmentation in space according to the position of a preset reference model to segment the ladle point cloud data from the original point cloud data; and reconstructing the three-dimensional point cloud model of the ladle through the ladle point cloud data.

[0009] In a further scheme of the present application, the three-dimensional point cloud model of the ladle and the preset reference model are positioned and matched so that the three-dimensional point cloud model of the ladle and the reference model are aligned based on the same reference, including: establishing a three-dimensional model of a new ladle as a reference model, and obtaining a first feature point set in the reference model; obtaining a second feature point set of the three-dimensional point cloud model of the ladle, and using a positioning matching algorithm to obtain a conversion matrix between the first feature point set and the second feature point set; changing the three-dimensional point cloud model of the ladle or the reference model according to the conversion matrix so that the three-dimensional point cloud model of the ladle and the reference model are based on the same center axis.

[0010] In a further scheme of the present application, the three-dimensional point cloud model of the ladle is divided into n layers along the depth direction, and the radius of each layer is calculated according to the benchmark to obtain the actual radius of each point on the inner wall of the current ladle, including: setting n parallel planes of equal height in the depth direction of the ladle, and making the n parallel planes intersect with the ladle point cloud model respectively to obtain n point cloud slices; performing radius calculation at several measuring points on the inner circumference of each point cloud slice to obtain the actual radius of each point on the inner wall of the current ladle.

[0011] In a further scheme of the present application, radius calculation is performed at several measuring points on the inner circumference of each point cloud slice to obtain the actual radius of each point on the inner wall of the current ladle, including: starting from the point where the reference is located on the plane as the center, dividing into m radial angle lines from 0 degrees to 360 degrees; according to the first intersection of the angle line and the point cloud slice, obtaining m inner wall point clouds of each point cloud slice; calculating the planar distance from the inner wall point cloud to the reference to obtain n layers of parallel planes and the actual radius of each point on the inner wall of the m ladle.

[0012] In a further scheme of the present application, the residual thickness value of the refractory brick is obtained by comparing the actual radius with the initial radius of the corresponding point in the reference model, including: comparing the actual radius of each point on the inner wall of the current ladle with the initial radius of the corresponding point in the reference model, and calculating the radius difference, which is the corrosion value of the refractory brick at the corresponding point; and subtracting the corrosion value of the corresponding point from the initial radius data to obtain the residual thickness value of the corresponding point of the refractory brick.

[0013] In a further scheme of the present application, the automatic detection method also includes: comparing the residual thickness value with a preset critical value, if the residual thickness value of a certain point is lower than the critical value, marking the corresponding point as a risk point; classifying each risk point according to its angle value in the range of 0 degrees to 360 degrees; traversing the angle values ​​of all risk points, and extracting corresponding risk points with continuous angle values; if the number of risk points with continuous angle values ​​is greater than a preset threshold number, extracting the angle value area corresponding to these risk points as the risk area of ​​package penetration.

[0014] In a further scheme of the present application, the automatic detection method also includes: when there is a risk area for ladle penetration, the ladle circulation operation is stopped and a visual warning of the risk area for ladle penetration is issued; if the number of risk points with consecutive angle values ​​is less than a threshold number, the ladle circulation operation is stopped and a data log is recorded, and the infrared detection equipment is controlled to re-inspect the area where the corresponding risk point is located.

[0015] In a further embodiment of the present application, n is 25 to 40; and m is 90 to 720.

[0016] The second aspect of the present application also provides an automatic monitoring system for a ladle, comprising: a three-dimensional modeling module, used to perform three-dimensional modeling of a ladle in production flow to obtain a three-dimensional point cloud model of the ladle; a positioning and matching module, used to position and match the three-dimensional point cloud model of the ladle with a preset reference model to ensure that the three-dimensional point cloud model of the ladle and the reference model are aligned based on the same reference; a radius calculation module, used to divide the three-dimensional point cloud model of the ladle into n layers along the depth direction, and perform radius calculation on each layer according to the reference to determine the actual radius of each point on the inner wall of the current ladle; a residual thickness value calculation module, used to calculate the residual thickness value of the refractory brick by comparing the actual radius with the initial radius of the corresponding point in the reference model. The risk point detection module is used to detect points where the residual thickness value is lower than the critical residual thickness value, and to determine whether there are risk points whose number is continuously greater than a preset threshold value; the bag penetration risk area extraction module, if the risk point detection module detects risk points whose number is continuously greater than a preset threshold value, then the angle value area corresponding to these risk points is extracted as the bag penetration risk area; the early warning and visualization module, when there is a bag penetration risk area, stops the ladle circulation operation, and visualizes the bag penetration risk area and issues an early warning.

[0017] In summary, the offline calibration device and method provided by the embodiments of the present invention include at least the following technical effects:

[0018] The automatic monitoring method for the ladle provided in the present application obtains a three-dimensional point cloud model through high-precision three-dimensional scanning technology, and more accurately identifies the small temperature difference caused by thickness changes; the scheme divides the three-dimensional point cloud model of the ladle into multiple layers along the depth direction, and calculates the radius of each layer, so as to obtain the actual radius of each point on the inner wall of the current ladle, thereby realizing a comprehensive and accurate measurement of the residual thickness of the refractory bricks on the inner wall of the entire ladle; by real-time monitoring of the residual thickness of the refractory bricks on the inner wall of the ladle, it is helpful to timely discover potential safety hazards and improve the operating stability of the ladle.

[0019] Other features and advantages of the embodiments of the present invention will be described in the following specific implementation examples. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 A flow chart of an automatic monitoring method for a ladle provided in an embodiment of the present application;

[0022] Figure 2 A flowchart of step S10 in the automatic monitoring method for a ladle provided in an embodiment of the present application;

[0023] Figure 3 A flow chart of step S20 in the automatic monitoring method for a ladle provided in an embodiment of the present application;

[0024] Figure 4 A flowchart of step S30 in the automatic monitoring method for a ladle provided in an embodiment of the present application;

[0025] Figure 5 A schematic diagram of a point cloud of a ladle 3D point cloud model and a reference model is shown in the automatic monitoring method for a ladle provided in an embodiment of the present application, wherein the red one is the 3D point cloud model of the ladle and the yellow one is the reference model;

[0026] Figure 6 A schematic diagram of dividing a three-dimensional point cloud model of a steel ladle along a depth direction in the automatic monitoring method for a steel ladle provided in an embodiment of the present application;

[0027] Figure 7A flowchart of step S32 in the automatic monitoring method for a ladle provided in an embodiment of the present application;

[0028] Figure 8 The automatic monitoring method for a ladle provided in the embodiment of the present application demonstrates the point cloud section of a certain layer of a three-dimensional point cloud model of a ladle and a reference model;

[0029] Fig. 9 A schematic diagram of the distribution of angle lines when m=360 is provided for demonstrating the automatic monitoring method for a ladle provided in an embodiment of the present application;

[0030] Fig.10 A schematic diagram of an inner wall point cloud extracted by the automatic monitoring method for a ladle provided in an embodiment of the present application;

[0031] Fig.11 A flowchart of step S40 in the automatic monitoring method for a ladle provided in an embodiment of the present application;

[0032] Fig.12 A flow chart of step S50 in the automatic monitoring method for a ladle provided in an embodiment of the present application;

[0033] Fig.13 A schematic diagram of the modules of the automatic monitoring system for a ladle provided in an embodiment of the present application. DETAILED DESCRIPTION

[0034] The terms "second direction", "first direction", "third direction", "inside", "outside" and the like that appear below to indicate orientation or positional relationships, unless otherwise specified, are to be understood as being based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0035] In addition, if there is a feature defined as "first" or "second", it is only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Features defined as "first" or "second" may explicitly or implicitly include at least one of the defined features. If there is a description of "plurality", the general meaning is to include at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0036] In this application, unless otherwise clearly specified and limited, the terms such as "installed", "connected", "connected", "fixed" and so on should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integrated connection; it can be a mechanical connection or an electrical connection, it can be a direct connection, or it can be an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For those skilled in the art, the specific meanings of the above terms in this application can be understood according to the specific circumstances.

[0037] In the description of this specification, if the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" appear, it means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0038] Reference Figure 1 The present application embodiment first provides an automatic monitoring method for a ladle, comprising:

[0039] Step S10, performing three-dimensional modeling on the ladle in production flow to obtain a three-dimensional point cloud model of the ladle;

[0040] Step S20, positioning and matching the ladle three-dimensional point cloud model with a preset reference model, so that the ladle three-dimensional point cloud model and the reference model are aligned based on the same reference;

[0041] Step S30, dividing the ladle three-dimensional point cloud model into n layers along the depth direction, and calculating the radius of each layer according to the benchmark to obtain the actual radius of each point on the inner wall of the current ladle;

[0042] Step S40, obtaining the residual thickness value of the refractory brick by comparing the actual radius with the initial radius of the corresponding point in the reference model.

[0043] Based on the general concept, firstly, three-dimensional coordinate data of the surface of the ladle in the production flow is obtained through three-dimensional scanning, and a three-dimensional point cloud model of the ladle is formed; then, it is ensured that the three-dimensional point cloud model of the ladle and the preset reference model are aligned based on the same reference in space for subsequent comparison and analysis. For example, feature point matching, ICP (Iterative Closest Point) algorithm and other methods are used for rough matching, and the three-dimensional point cloud model of the ladle is aligned with the preset reference model to keep the reference of the two in spatial position consistent.

[0044] The ladle 3D point cloud model is divided into n layers along the depth direction (i.e., the direction perpendicular to the bottom of the ladle), and each layer contains a certain amount of point cloud data. Then, the actual radius of each point in each layer is calculated based on the aligned benchmark; the actual radius calculated in step S30 is compared with the initial radius of the corresponding point in the reference model, and the difference between the two is the corrosion value of the refractory brick at that point; and the residual thickness value of the refractory brick at each point on the inner wall of the ladle, and then its overall condition is evaluated.

[0045] In the corresponding step S20, there are many methods for positioning and matching, including feature point matching, ICP algorithm (Iterative Closest Point), NDT algorithm (Normal Distributions Transform), etc.; the two point cloud models are based on the same benchmark. It should be noted that the three-dimensional point cloud of the ladle is a set of point cloud data obtained directly from the surface of the ladle in the actual production flow through three-dimensional scanning technology, and the reference model is a geometric model generated based on the design of the original new ladle, which is used for comparison and analysis with the three-dimensional point cloud data of the actual ladle.

[0046] In summary, the automatic monitoring method for the ladle provided in this application obtains a three-dimensional point cloud model through high-precision three-dimensional scanning technology. In contrast, infrared detection equipment usually requires experience to fix and measure several preset points, which is difficult to achieve such comprehensiveness and accuracy; and this solution divides the three-dimensional point cloud model of the ladle into multiple layers along the depth direction, and calculates the radius of each layer, so as to obtain the actual radius of each point on the inner wall of the current ladle, thereby realizing a comprehensive and accurate measurement of the residual thickness of the refractory bricks on the inner wall of the entire ladle. By real-time monitoring of the residual thickness of the refractory bricks on the inner wall of the ladle, it is helpful to timely discover potential safety hazards and improve the operational stability of the ladle.

[0047] like Figure 2 In the embodiment of the present application, an example of step S10 is: performing three-dimensional modeling on the ladle in production flow to obtain a three-dimensional point cloud model of the ladle, including:

[0048] Step S11, receiving a position signal of a ladle in circulation arriving at at least one preset process, and controlling at least one acquisition device to obtain point cloud images of at least two positions of the ladle during the circulation process;

[0049] Ensure the accuracy and reliability of position signal sensors (such as RFID tags, infrared sensors, etc.) to trigger the acquisition equipment in time; when the ladle in circulation arrives at the preset process, the system will receive a position signal. This signal can be an electrical signal, wireless signal or other form of trigger signal to start the subsequent point cloud data acquisition process.

[0050] In this application, "preset process" can refer to one process position or multiple process positions; correspondingly, there can be one or multiple acquisition devices installed at different process positions. That is, the acquisition device, such as a laser scanner or a structured light scanner, should be able to cover at least two positions of the ladle during the process to ensure that the acquired point cloud images have sufficient overlapping areas for subsequent splicing.

[0051] In one example, if the preset process refers to only a single specific process position, at least one acquisition device can be installed at this position. The single acquisition device can be carried by the robot arm to cover the scanning of different directions of the ladle in a surrounding manner.

[0052] Furthermore, multiple collection devices may be directly installed at different positions corresponding to the specific process position, that is, multiple collection devices are fixed at different positions respectively, and when the ladle arrives at the process position, multiple collection devices perform synchronous collection.

[0053] In another example, if the preset process includes multiple process positions, there are also multiple acquisition devices, and the acquisition devices should be installed at the corresponding process positions; that is, the acquisition devices should be arranged at positions that can cover multiple directions that the ladle passes through during the circulation process. Specifically, at each process position, the installation position of the acquisition device is accurately located according to the circulation path of the ladle and the direction when it may pass through the process. Ensure that the field of view of the acquisition device at each process position can cover multiple directions that the ladle passes through during the circulation process: for example, the top, the left half side, the right half side, etc., so as to obtain comprehensive point cloud data; to ensure that the acquired point cloud images have sufficient overlapping areas for subsequent splicing, and to ensure that the ladle does not need to affect its original process during the circulation process, so as to avoid affecting the circulation efficiency.

[0054] Step S12, stitching the point cloud images in at least two directions to obtain original point cloud data including the ladle;

[0055] It can be understood that the image stitching algorithm is used to stitch point cloud images from different directions to obtain a complete original point cloud data set containing the ladle. The stitching process needs to consider the spatial relationship and overlapping parts between the point cloud images to ensure that the stitched point cloud data is continuous and accurate. In an example: the stitching algorithm can be expressed as:

[0056] Pcombined=Stitch(P1,P2,…,Pn)

[0057] Among them, Pcombined represents the complete point cloud dataset after splicing, and P1, P2, …, Pn represent point cloud images from different orientations.

[0058] Step S13, performing region boundary segmentation in space according to the position of the preset reference model to segment the ladle point cloud data from the original point cloud data;

[0059] Among them, the original point cloud data still provides an accurate three-dimensional representation of the ladle and its surroundings, providing a basis for subsequent point cloud data segmentation and model reconstruction.

[0060] Step S14, performing region boundary segmentation in space according to the position of the preset reference model to segment the ladle point cloud data from the original point cloud data;

[0061] It should be noted that the "reference model" mentioned above is the unused ladle model mentioned above, and its position and size data are known; using the position and size information of the preset reference model, the position information of the preset reference model is converted into a boundary in space, and the original point cloud data is divided into two parts using the crop point cloud data command: one part is the ladle point cloud data, and the other part is the point cloud data of the surrounding environment or other objects; through the guidance of the preset reference model, the ladle point cloud data can be accurately segmented from the original point cloud data to avoid interference from the surrounding environment. In some embodiments, other segmentation methods can also be used instead, such as clustering-based segmentation, feature-based segmentation, or deep learning.

[0062] Subsequently, the segmented ladle point cloud data is subjected to preprocessing operations such as denoising and filtering. The filtering operation can further smooth the point cloud data, remove outliers or irregular points, and make the point cloud data more uniform and continuous, so as to improve the quality of the collected point cloud data; the preprocessed ladle point cloud data is reconstructed into a three-dimensional point cloud model using a point cloud processing algorithm.

[0063] like Figure 3 In the embodiment of the present application, an example of step S20 is: positioning and matching the ladle three-dimensional point cloud model with a preset reference model so that the ladle three-dimensional point cloud model and the reference model are aligned based on the same reference, including:

[0064] Step S21, establishing a three-dimensional model of a new bag as a reference model, and obtaining a first feature point set in the reference model;

[0065] According to the original shape and size of the ladle, a three-dimensional model of a new ladle is established as a reference model; key feature points are extracted from the reference model to form a first feature point set, such as a point set in the bottom plane of the ladle.

[0066] Step S22, obtaining a second feature point set of the three-dimensional point cloud model of the ladle, and obtaining a conversion matrix between the first feature point set and the second feature point set by using a positioning matching algorithm;

[0067] Feature points corresponding to the reference model are extracted from the three-dimensional point cloud model of the ladle to form a second feature point set. Then, a positioning matching algorithm (such as ICP algorithm, NDT algorithm, etc.) is used to calculate the transformation matrix between the first feature point set and the second feature point set; this transformation matrix describes how to transform the three-dimensional point cloud model of the ladle to a position aligned with the reference model.

[0068] Step S23, changing the ladle three-dimensional point cloud model or the reference model according to the conversion matrix, so that the ladle three-dimensional point cloud model and the reference model are based on the same central axis.

[0069] The algorithm will output a transformation matrix, which describes how to transform the ladle point cloud model to a position aligned with the reference model. According to the calculated transformation matrix, the ladle 3D point cloud model or reference model is transformed so that the two are aligned based on the same central axis. The central axis is usually the symmetry center or rotation center of the ladle, which facilitates subsequent monitoring and analysis work, such as calculating the residual thickness of refractory bricks and monitoring the deformation of the ladle.

[0070] It can be understood that by following the principle of consistent center of circle and aligning the ladle 3D point cloud model and the reference model based on the same center axis, it can be ensured that the two are comparable in subsequent comparison and analysis.

[0071] like Figures 4 to 6 In the embodiment of the present application, an example of step S30 is: dividing the three-dimensional point cloud model of the ladle into n layers along the depth direction, and calculating the radius of each layer according to the benchmark to obtain the actual radius of each point on the inner wall of the current ladle, including:

[0072] Step S31, setting n parallel planes of equal height in the depth direction of the ladle, and making the n parallel planes intersect with the ladle point cloud model respectively to obtain n point cloud slices;

[0073] Set n parallel planes of equal height in the depth direction of the ladle (i.e. from the top to the bottom of the ladle, or from the bottom to the top); the number and height of these planes can be adjusted according to actual needs, preferably to just reflect the changes of each refractory brick, so as to ensure that the shape and changes of the inner wall of the ladle can be fully reflected. Then, each plane is intersected with the three-dimensional point cloud model of the ladle to obtain n point cloud slices containing the information of the inner wall of the ladle.

[0074] For example, Figure 6 :Set n parallel planes of equal height, so as to obtain n point cloud slices. The position of each slice can be represented by its coordinate z_i in the depth direction of the ladle. The value of z_i depends on the height h of the slice and the slice number i, as well as the starting coordinates of the bottom or top of the ladle. If the coordinate of the bottom of the ladle is z=0, and the slice height is h, then the coordinate of the first slice is z_1=h, the coordinate of the second slice is z_2=h+h=2h, and so on; each slice contains a set of point cloud data, which is located in the intersection area of ​​the slice plane and the ladle point cloud model. Point cloud data can be represented as a series of three-dimensional coordinate points (x_j, y_j, z_j), where j is the point number in the point cloud. In the slice, the z coordinates of all points are approximately equal to the z_i coordinates of the slice (taking into account the slice thickness and measurement errors). Point cloud slice data are usually stored in specific file formats, such as PLY, PCD, OBJ, etc.

[0075] By dividing the ladle 3D point cloud model into multiple point cloud slices of equal height, the shape and changes of the ladle inner wall can be analyzed in more detail. This layered processing helps to find small deformations or damages and provides more accurate data for subsequent radius calculations.

[0076] It can be understood that point cloud slicing refers to cutting point cloud data through a set of parallel planes to obtain point cloud data located in the same plane dimension. The z value of the point cloud data in a point cloud slice can be kept consistent or within a certain tolerance range.

[0077] Step S32, performing radius calculation at a number of measurement points on the inner periphery of each point cloud slice to obtain the actual radius of each point on the inner wall of the current ladle.

[0078] In each point cloud slice, select several measurement points at the inner circumference. These measurement points should be evenly distributed to ensure that the shape of the inner wall of the ladle at the slice can be fully reflected. Subsequently, three-dimensional geometric measurement technology (such as least squares method, projection method, etc.) is used to calculate the horizontal distance from each measurement point to the center of the ladle (i.e., the central axis), i.e., the actual radius.

[0079] By evenly selecting measurement points in each point cloud slice and performing radius calculation, the actual radius values ​​of each point on the inner wall of the current ladle can be obtained; these radius values ​​not only reflect the shape and changes of the inner wall of the ladle, but can also be used for subsequent analysis and comparison, such as calculating the residual thickness of refractory bricks.

[0080] like Figure 7 In the embodiment of the present application, an example of step S32 is: performing radius calculation at a number of measurement points at the inner periphery of each point cloud slice to obtain the actual radius of each point on the inner wall of the current ladle, including:

[0081] Step S321, starting from the reference point on the plane as the center, divide it into m radial angle lines from 0 degrees to 360 degrees;

[0082] This step is to divide the two-dimensional plane into m radial angle lines from 0 to 360 degrees with the reference point (i.e., the point where the central axis of the circle is located) as the center, that is, to divide the plane into m areas with equal angles; each area is defined by two adjacent angle lines, which start from the reference point and extend radially to the boundary of the point cloud slice;

[0083] In the present application, n is 25 to 40; m is 90 to 720, depending on the specifications of the ladle and the accuracy required for the project. In the present application, n is optionally 38 and m is optionally 360.

[0084] For example, Figures 8 to 10 , assuming m = 360, the plane is divided into 360 equal sectors, and the angle range of each sector is 1 degree (360 degrees / 360).

[0085] Step S322: Obtain m inner wall point clouds of each point cloud slice according to the first intersection point of the angle line and the point cloud slice;

[0086] The first intersection point of each angle line and the point cloud slice generates m intersection points; these intersection points represent the position of the inner wall of the ladle; in simple terms, the m inner wall point cloud coordinates on each point cloud slice are obtained by extracting these intersection points.

[0087] For example, Fig. 9 : For the case of m=360 (i.e., 360 angle lines), 360 inner wall point cloud points will be extracted from each point cloud slice.

[0088] Step S323, calculate the plane distance from the inner wall point cloud to the reference to obtain the actual radius of each point on the inner wall of the m ladle in n layers of parallel planes.

[0089] For example, 360 inner wall point cloud coordinates are extracted from the point cloud slices (e.g. Fig.10), for each inner wall point cloud coordinate, calculate its distance to the center of the circle and obtain the actual radius value at 360 different angles.

[0090] That is, the present application divides the three-dimensional point cloud model of the ladle into multiple point cloud slices of equal height along the depth direction, thereby fully covering the inner wall of the ladle from top to bottom, and divides 0 to 360 degrees into multiple radial angle lines with the center of the circle as the center on the plane, thereby ensuring comprehensive coverage and detailed division from both depth and angle dimensions, and significantly improving the accuracy of measuring the radius of the inner wall of the ladle.

[0091] like Fig.11 , further in step S40, the residual thickness value of the refractory brick is obtained by comparing the actual radius with the initial radius of the corresponding point in the reference model, including:

[0092] Step S41, comparing the actual radius of each point on the inner wall of the current ladle with the initial radius of the corresponding point in the reference model, and calculating the radius difference, which is the corrosion value of the refractory brick at the corresponding point;

[0093] Step S42, subtract the corrosion value of the corresponding point from the initial radius data to obtain the residual thickness value of the corresponding point of the refractory brick.

[0094] The above-mentioned actual radius of each point on the inner wall of the current ladle is compared with the initial radius of the corresponding point in the reference model, and the corrosion value of the refractory brick at the corresponding point is obtained by calculating the difference between the two; the corrosion value intuitively reflects the degree of wear or corrosion of the refractory brick at this point due to use; specifically, the actual radius must be greater than the initial radius, and the radius difference = actual radius - initial radius; the radius difference is the corrosion value of the refractory brick at the corresponding point.

[0095] After obtaining the corrosion value of each point, subtract these corrosion values ​​from the initial radius data to obtain the residual thickness value of the refractory brick at each corresponding point. The residual thickness value refers to the thickness of the refractory brick that remains after use; specifically, the residual thickness value = actual radius = initial radius - corrosion value. This value reflects the actual thickness of the refractory brick in its current state.

[0096] like Fig.12 As shown, in a further embodiment of the present application, the automatic detection method further includes:

[0097] Step S51, comparing the residual thickness value with a preset critical value, if the residual thickness value of a certain point is lower than the critical value, marking the point as a risk point;

[0098] Compare the calculated residual thickness of refractory bricks with the preset critical value, which is set based on the safety standards for the use of refractory bricks, the operating conditions of the ladle and previous experience. If the residual thickness of a corresponding point is lower than this critical value, we consider that there is a safety risk at that point and mark it as a risk point.

[0099] Step S52, classifying each risk point according to its angle value within the range of 0 degrees to 360 degrees;

[0100] For all positions marked as risk points, record their angle values ​​in the range of 0 to 360 degrees, so as to classify the risk points according to their distribution on the circumference of the inner wall of the ladle, which is convenient for subsequent analysis;

[0101] Step S53, traverse the angle values ​​of all risk points, and extract corresponding risk points with continuous angle values;

[0102] Step S54: If the number of risk points with continuous angle values ​​is greater than a preset threshold, the angle value areas corresponding to these risk points are extracted as bag penetration risk areas.

[0103] It can be understood that the angle values ​​of all risk points are traversed to check whether there are continuous risk points that exceed a certain threshold number; the "threshold number" is set according to the design of the ladle and the material of the refractory bricks. If the number of risk points with continuous angle values ​​exceeds this threshold, it indicates that these continuous risk points have formed a potential risk area, that is, marked as a puncture risk area. This area represents a potential position on the circumference of the inner wall of the ladle that may cause puncture due to severe wear or corrosion of the refractory bricks.

[0104] For the sake of clarity, an example is given as follows: the critical residual thickness value is set to 100 mm, the number of thresholds is set to 5, and the positive direction of the x-axis is set to 0 degrees and 360 degrees, and a set of data is provided as shown in Table 1 below:

[0105] Table 1

[0106] Inner wall point cloud number Is it a risk point? Angle value (°) Residual thickness(mm) Critical value (mm) 1 no 1 110mm 100mm ... ... ... ... 100mm 36 yes 36 85mm 100mm 37 yes 37 87mm 100mm 38 yes 38 90mm 100mm 39 yes 39 91mm 100mm 40 yes 40 89mm 100mm ... ... ... ... 100mm

[0107] In the above case, the risk points are those located between the angles of 36° and 40° and their number is greater than the preset threshold (≥5). Therefore, the area between 36° and 40° can be extracted as the bag penetration risk area, and other risk points, if they are not continuous, are not included in the bag penetration risk area.

[0108] By dividing the angle into equal intervals of 1 degree, the specific location of the risk area on the circumference can be more accurately identified, which helps to more accurately assess the location and severity of the risk.

[0109] The technical advantage of the above method is that if the judgment is made only based on the data of a single risk point, misjudgment or missed detection may occur due to data fluctuations or point cloud acquisition errors; according to actual conditions, corrosion is relatively not concentrated at a certain point or a very small range. Corrosion is a gradual development process and is usually not concentrated at a certain point or a very small range. Therefore, it is insufficient to judge the corrosion situation based only on the data of a single point; by considering the data of multiple consecutive risk points, potential high-risk areas can be more reliably identified, the possibility of misjudgment and missed detection can be reduced, and more reasonable and reliable maintenance warnings can be formulated.

[0110] In this application, the automatic detection method also includes:

[0111] Step S55: when there is a ladle penetration risk area, stop the ladle circulation operation and issue a visual warning of the ladle penetration risk area;

[0112] That is, once a risk area for ladle penetration is detected, the system stops the current circulation operation of the ladle by sending a command to prevent possible safety accidents caused by continued use in a risky state.

[0113] At the same time, the system should intuitively display to the operator through a graphical interface or other visualization means, combined with the camera image of the location of the ladle penetration risk area calculated above, the three-dimensional model simulation, and the corresponding data and other information. For example, the risk area can be highlighted in the three-dimensional model of the ladle, or the relevant alarm information can be displayed on the monitoring screen; by intuitively displaying the location and severity of the ladle penetration risk area, the operator can formulate a maintenance plan more effectively and give priority to high-risk areas, thereby extending the service life of the ladle and reducing maintenance costs.

[0114] Step S56: If the number of risk points with consecutive angle values ​​is less than the threshold value, the ladle circulation operation is stopped and the data log is recorded, and the infrared detection equipment is controlled to re-inspect the area where the corresponding risk point is located.

[0115] On the contrary, if the number of consecutive risk points of angle values ​​is less than the threshold number, it indicates that there may be misjudgment in the data collection or calculation process, or there are structural problems with individual fireproof bricks, etc. By recording abnormal data and related information, the circulation operation of the ladle is stopped and the early warning system is triggered at the same time, and infrared detection equipment is used to conduct actual observations of the corresponding positions. Combined with the above method, it is highly targeted and only a small number of measurements are required in the suspected problem area to obtain accurate results, thereby improving monitoring efficiency.

[0116] Specifically, compared to the traditional method of only using infrared detection equipment in conjunction with manual measurement, it is not only time-consuming and labor-intensive, but also depends on the operator's experience and judgment, and may be subjective and erroneous. By combining the above-mentioned monitoring scheme, when the number of risk points with continuous angle values ​​is less than the threshold number, the potential problem area can be quickly located, and only a small amount of measurement is performed at these corresponding points using infrared detection equipment. This method not only reduces unnecessary comprehensive inspections, reduces manpower and time costs, but also improves the pertinence and accuracy of monitoring. In addition, this method also reduces the interference of human factors, making the monitoring results more objective and reliable; through precise data analysis and positioning, potential safety hazards can be discovered and dealt with in a timely manner, ensuring the safe operation of the ladle and reducing potential risks and losses.

[0117] At the same time, step S56 also warns of manual intervention, and the manual intervention ensures the safe operation of the ladle and reduces potential risks and losses by checking the data and the results of the infrared detection equipment.

[0118] Through the above, when the infrared device shows that the refractory bricks at a specific risk point are abnormal in multiple measurement points or continuous measurements, the system will output "the refractory bricks at individual points have structural defects", so that manual maintenance personnel can quickly locate and repair or replace them, which not only improves maintenance efficiency, but also reduces unnecessary comprehensive inspections and reduces maintenance costs. On the contrary, if the results obtained by the infrared device at different measurement points or point cloud computing are inconsistent, the system will output "there are problems in the initial point cloud acquisition process"; this prompts the user that there may be deviations or errors in the data acquisition stage, and the point cloud acquisition equipment needs to be recalibrated.

[0119] Through the above methods, problems in the data collection stage can be discovered and solved in a timely manner to ensure the accuracy of subsequent monitoring. In general, this method provides users with targeted reference information through rapid and accurate diagnosis, improves monitoring and maintenance efficiency, reduces costs, and also ensures the safe operation of the ladle.

[0120] In summary, the present application provides an automatic monitoring method for a ladle, which locates and matches the three-dimensional point cloud model of the ladle and a preset reference model, divides the three-dimensional point cloud model of the ladle into n layers along the depth direction, and calculates the radius of each layer according to the benchmark to obtain the actual radius of each point on the inner wall of the current ladle, and obtains the residual thickness value of the refractory brick through the actual radius and the initial radius of the corresponding point in the reference model. It can reflect the corrosion condition of the inner wall of the ladle more carefully, improve the monitoring accuracy, and more reliably ensure the safe operation of the ladle.

[0121] like Fig.13 As shown, in accordance with the second aspect of the present application, the present application further provides an automatic monitoring system 100 for a ladle, comprising:

[0122] A three-dimensional modeling module 10 is used to perform three-dimensional modeling on the ladle in production flow to obtain a three-dimensional point cloud model of the ladle;

[0123] A positioning and matching module 20 is used to position and match the three-dimensional point cloud model of the ladle with a preset reference model to ensure that the three-dimensional point cloud model of the ladle and the reference model are aligned based on the same reference;

[0124] The radius calculation module 30 is used to divide the three-dimensional point cloud model of the ladle into n layers along the depth direction, and calculate the radius of each layer according to the reference to determine the actual radius of each point on the inner wall of the current ladle;

[0125] The residual thickness value calculation module 40 is used to calculate the residual thickness value of the refractory brick by comparing the actual radius with the initial radius of the corresponding point in the reference model.

[0126] The risk point detection module 50 is used to detect points where the residual thickness value is lower than the critical residual thickness value, and to determine whether there are risk points where the number of residual thickness values ​​is continuously greater than a preset threshold value;

[0127] The bag penetration risk area extraction module 60 extracts the angle value areas corresponding to these risk points as bag penetration risk areas if the risk point detection module detects a number of risk points that is greater than a preset threshold value in succession;

[0128] The early warning and visualization module 70 stops the ladle circulation operation when there is a ladle penetration risk area, and issues an early warning by visualizing the ladle penetration risk area.

[0129] Furthermore, those skilled in the art should understand that if the automatic monitoring system 100 provided in the embodiment of the present application; all or part of the sub-modules involved in each product are combined and replaced through embodiments such as fusion, simple change, mutual transformation, etc., such as the placement and movement of each component, or designing an offline calibration device 100; or a detachable design; or the product it constitutes is set as one, such as an integrated design; all the combined components can form equipment / devices / systems with specific functions, and using such equipment / devices / systems to replace the corresponding components of the present application also falls within the protection scope of the present application.

[0130] Therefore, in the third aspect of the present application, the present application also provides another automatic monitoring system for a ladle, comprising a point cloud acquisition device, an infrared acquisition device and a controller; the controller is electrically connected to the point cloud acquisition device and the infrared acquisition device, and the controller is configured to be able to execute the above-mentioned automatic monitoring method.

[0131] The various technical features described above can be combined arbitrarily. Although all possible combinations of these technical features are not described, any combination of these technical features should be considered to be covered by this specification as long as there is no contradiction in such combination.

[0132] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still adjust the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these adjustments or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. An automatic monitoring method for a ladle, characterized in that: include: Carry out 3D modeling of the ladle in production flow and obtain the 3D point cloud model of the ladle; Positioning and matching the ladle three-dimensional point cloud model with a preset reference model so that the ladle three-dimensional point cloud model and the reference model are aligned based on the same reference; Divide the ladle three-dimensional point cloud model into n layers along the depth direction, and calculate the radius of each layer according to the reference to obtain the actual radius of each point on the inner wall of the current ladle, where n≥2; The residual thickness value of the refractory brick is obtained by comparing the actual radius with the initial radius of the corresponding point in the reference model.

2. The automatic monitoring method according to claim 1, characterized in that: The three-dimensional modeling of the ladle in the production flow is performed to obtain the three-dimensional point cloud model of the ladle, including: Receiving a position signal of a ladle in circulation arriving at at least one preset process, and controlling at least one acquisition device to obtain point cloud images of at least two positions of the ladle during the circulation process; splicing the point cloud images in at least two directions to obtain original point cloud data containing the ladle; Performing regional boundary segmentation in space according to the position of the preset reference model to segment the ladle point cloud data from the original point cloud data; Reconstruct the 3D point cloud model of the ladle through the ladle point cloud data.

3. The automatic monitoring method according to claim 1, characterized in that: Positioning and matching the ladle three-dimensional point cloud model with a preset reference model so that the ladle three-dimensional point cloud model and the reference model are aligned based on the same reference includes: Establishing a three-dimensional model of a new bag as the reference model, and obtaining a first feature point set in the reference model; Acquire a second feature point set of the three-dimensional point cloud model of the ladle, and acquire a conversion matrix between the first feature point set and the second feature point set by using a positioning matching algorithm; The ladle three-dimensional point cloud model or the reference model is changed according to the conversion matrix so that the ladle three-dimensional point cloud model and the reference model are based on the same central axis.

4. The automatic monitoring method according to claim 1, characterized in that: The three-dimensional point cloud model of the ladle is divided into n layers along the depth direction, and the radius of each layer is calculated according to the reference to obtain the actual radius of each point on the inner wall of the current ladle, including: Setting n parallel planes of equal height in the depth direction of the ladle, and making the n parallel planes intersect with the ladle point cloud model respectively to obtain n point cloud slices; The radius is calculated at several measuring points on the inner periphery of each point cloud slice to obtain the actual radius of each point on the inner wall of the current ladle.

5. The automatic monitoring method according to claim 4, characterized in that: The radius calculation is performed at several measuring points on the inner periphery of each point cloud slice to obtain the actual radius of each point on the inner wall of the current ladle, including: Starting from the reference point on the plane as the center, the angle lines are divided into m radial angle lines from 0 to 360 degrees; wherein m ≥ 4; According to the first intersection point of the angle line and the point cloud slice, m inner wall point clouds of each point cloud slice are obtained; The plane distance from the inner wall point cloud to the reference is calculated to obtain the actual radius of each point on the inner wall of the m ladle in n layers of parallel planes.

6. The automatic monitoring method according to claim 5, characterized in that: The method of obtaining the residual thickness value of the refractory brick by comparing the actual radius with the initial radius of the corresponding point in the reference model includes: Compare the actual radius of each point on the inner wall of the current ladle with the initial radius of the corresponding point in the reference model, and calculate the radius difference, which is the corrosion value of the refractory brick at the corresponding point; The residual thickness value of the refractory brick at the corresponding point is obtained by subtracting the corrosion value of the corresponding point from the initial radius data.

7. The automatic monitoring method according to any one of claims 5 to 6, characterized in that: The automatic detection method further comprises: The residual thickness value is compared with a preset critical value, and if the residual thickness value of a certain point is lower than the critical value, the corresponding point is marked as a risk point; Classifying each of the risk points according to its angle value within the range of 0 degrees to 360 degrees; Traverse the angle values ​​of all risk points and extract the corresponding risk points with continuous angle values; If the number of risk points with continuous angle values ​​is greater than a preset threshold, the angle value areas corresponding to these risk points are extracted as the bag penetration risk areas.

8. The automatic monitoring method according to claim 7, characterized in that: The automatic detection method further comprises: When the ladle penetration risk area exists, the ladle circulation operation is stopped and the ladle penetration risk area is visually warned; If the number of consecutive risk points with angle values ​​is less than the threshold, the ladle circulation operation is stopped and the data log is recorded, and the infrared detection equipment is controlled to re-inspect the area where the corresponding risk point is located.

9. The automatic monitoring method according to claim 5, characterized in that: The n is 25 to 40; the m is 90 to 720.

10. An automatic monitoring system for a ladle, characterized in that: include: A three-dimensional modeling module, used for performing three-dimensional modeling on a ladle in production flow to obtain a three-dimensional point cloud model of the ladle; A positioning and matching module, used for positioning and matching the three-dimensional point cloud model of the ladle with a preset reference model to ensure that the three-dimensional point cloud model of the ladle and the reference model are aligned based on the same reference; A radius calculation module, used for dividing the three-dimensional point cloud model of the ladle into n layers along the depth direction, and performing radius calculation on each layer according to the reference to determine the actual radius of each point on the inner wall of the current ladle; The residual thickness value calculation module is used to calculate the residual thickness value of the refractory brick by comparing the actual radius with the initial radius of the corresponding point in the reference model. The risk point detection module is used to detect points where the residual thickness value is lower than the critical residual thickness value, and to determine whether there are risk points whose number is continuously greater than a preset threshold value; The bag penetration risk area extraction module, if the risk point detection module detects a number of risk points that is greater than a preset threshold value, then the angle value area corresponding to these risk points is extracted as the bag penetration risk area; The early warning and visualization module stops the ladle circulation operation when there is a ladle penetration risk area, and issues an early warning by visualizing the ladle penetration risk area.

Images