Efficient determination of flatness in planar rolling materials
A closed feedback control loop system using a camera-based acquisition and evaluation device corrects flatness errors in planar rolling materials by analyzing surface datasets, addressing environmental challenges and ensuring consistent material quality through real-time adjustments.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- PRIMETALS TECH GERMANY GMBH
- Filing Date
- 2022-08-09
- Publication Date
- 2026-06-23
AI Technical Summary
Existing methods for correcting flatness errors in planar rolling materials face challenges due to harsh environmental conditions and the need for reliable and efficient identification and correction of flatness errors in metal strips, particularly in high-temperature and vibration environments.
A closed feedback control loop system using a camera-based acquisition device, evaluation device, and control device to determine and correct flatness errors in real-time by analyzing two-dimensional datasets of the metal strip surface, employing Fourier transforms to identify intensity and spatial frequency of local oscillations, and adjusting control variables for the roll stand.
Enables precise and efficient correction of flatness errors in planar rolling materials, ensuring consistent material quality by continuously adjusting the roll stand's control elements based on real-time feedback, even in challenging industrial conditions.
Smart Images
Figure 0007879225000001 
Figure 0007879225000002 
Figure 0007879225000003
Abstract
Description
Technical Field
[0001] The present invention relates to an operating method for a roller assembly, - A flat metal rolling material extending in the width direction across the width of the rolling material is rolled using the roll stand of the roller assembly, and the flat rolling material leaves the roll stand in the transport direction after being rolled, - At least one two-dimensional dataset of the surface of the flat rolling material is repeatedly acquired non-contactively at the output side of the roll stand using an acquisition device that operates without mechanical action on the flat rolling material, and the values of the dataset depend at least on the outer surface flatness locally spreading at each corresponding location of the flat rolling material, - Each two-dimensional dataset is received by an evaluation device of the roller assembly, - The evaluation device determines an error value depending on the flatness error for each strip of the flat rolling material extending in the transport direction, using the strip of the respective two-dimensional dataset corresponding to the strip, - The evaluation device supplies the determined error value to a control device of the roller assembly, and the control device then takes the determined error value into account in determining the control variables for the flatness control elements of the roll stand, - As a result of the cooperation of the acquisition device, the evaluation device, the control device, and the roll stand, an operating method for a roller assembly results in a closed feedback control loop operating in real time.
[0002] In order to enable the feedback control loop to operate in real time, the aforementioned components of the roller assembly must repeatedly perform their functions in a fixed operating cycle. The operating cycle is generally in the range of milliseconds, mostly in the range of two-digit milliseconds, and in exceptional cases, in the range of lower three-digit milliseconds. This applies both within the prior art and within the scope of the present invention.
[0003] The present invention relates to a computer program comprising a machine code that can be directly processed by an evaluation device for a roller assembly, wherein the evaluation device repeatedly operates a control device for the roll stand and an acquisition device that operates non-contact without mechanical action on the planar rolling material of a metal during the operation of the roll stand as the planar rolling material is rolled and then exits in the transport direction after the planar rolling material has been rolled. - The acquisition device receives at least one two-dimensional dataset acquired by the acquisition device of the surface of the planar rolling material on the output side of the roll stand, where the values of each two-dimensional dataset depend at least on the locally extending outer flatness at each corresponding location of the planar rolling material. - For strips of planar rolling material extending in the transport direction, the error value dependent on the flatness error for each strip is determined using the strips in the corresponding two-dimensional dataset. - For consideration in determining the control variables for the flatness control element of the roll stand, the determined error value is supplied to the control device. Thus, it has a collaborative effect, This further results from a computer program that, as a result of the cooperation of acquisition, evaluation, and control devices, yields a closed feedback control loop that operates in real time.
[0004] The present invention relates to an evaluation device for a roller assembly, further comprising an evaluation device for a roller assembly, wherein the evaluation device is programmed with the computer program described above, and thereby the evaluation device cooperates with an acquisition device and a control device for the roller stand of the roller assembly in accordance with the operating method described above.
[0005] The present invention relates to a roller assembly, - The roller assembly has a roll stand, the roll stand is equipped with a flatness control element, and the roll stand rolls a planar rolling material of metal that extends in the width direction across the width of the rolling material, and after the planar rolling material is rolled, the roll stand is moved away in the transport direction. - The roller assembly has an acquisition device, which operates non-contact without mechanical action on the planar rolling material, and the acquisition device iteratively acquires at least one two-dimensional dataset of the surface of the planar rolling material at the output side of the roll stand, the values of the datasets depending at least on the localized outer flatness at each corresponding location of the planar rolling material, - The roller assembly has an evaluation device, which is connected to the acquisition device for data transfer to receive repeated two-dimensional datasets of the surface of the planar rolling material acquired using the acquisition device, and for each strip of planar rolling material extending in the transport direction, it determines an error value dependent on the flatness error for each strip using the strip of the two-dimensional dataset corresponding to each strip, and supplies the determined error value to the control device of the roller assembly. - The control device further originates from the roller assembly, taking into account the determined error value in determining the control variable for the flatness control element of the roll stand. [Background technology]
[0006] Such operating methods and associated corresponding articles are known from Patent Document 1. Specifically, Patent Document 1 describes acquiring images of each portion of a metal strip line by line using one or more cameras. During the acquisition of each line, the corresponding area of the metal strip is illuminated with a constant intensity. Repeated acquisition of lines generates a two-dimensional image of the surface of the metal strip. The image is divided into individual strip pieces. The strip pieces can extend in the longitudinal direction of the rolling material. The strip pieces are evaluated in terms of whether hollow bulges are identified therein. If such hollow bulges are identified, this is recorded as a local flatness error. The determined flatness error can be supplied to an upstream roll stand as a correction value to determine the operation of its flatness control element.
[0007] Patent Document 2 describes a method for obtaining the flatness of a metal strip on the output side of a roll stand in a spatially resolved manner across the width of the metal strip, using a segmented measuring roller. The control elements of the roll stand are operated depending on the obtained flatness in order to bring the flatness as close as possible to the desired flatness.
[0008] Patent Document 3 describes how to acquire images of the surface of a metal strip using a camera and how to determine the local surface condition by evaluating these images. Surface roughness, hollow ridges, color, brightness, chemical composition, and other properties have been mentioned as examples of local surface conditions. Based on the determined local surface conditions, the control of an upstream device is influenced. The device may be a cold rolling mill.
[0009] Patent Document 4 discloses an operating method for a roller assembly in which a planar rolling material of metal extending across the width of the rolling material is rolled using a roll stand, and the planar rolling material is moved away from the roll stand in the transport direction after rolling. Using a camera, images of the surface of the planar rolling material are repeatedly acquired at the output side of the roll stand, and the values depend on the locally distributed outer flatness at each corresponding location of the planar rolling material. The images are fed to an evaluation device, which determines an error value for the flatness of the strip of planar rolling material extending in the transport direction from the images. The evaluation device supplies the determined error value to a control device, which takes the error value into consideration when determining the control variables for the flatness control elements of the roll stand. Thus, as a result of the cooperation of the acquisition device, evaluation device, control device, and roll stand, a closed feedback control loop that operates in real time is obtained. To determine the error value, the evaluation device performs a local frequency analysis of each two-dimensional dataset and determines the respective error value based on the local frequency analysis.
[0010] Patent Document 5 discloses an operating method for a roller assembly in which a planar rolling material of metal, extending in the width direction across the width of the rolling material, is rolled using a roll stand, and the planar rolling material is moved away from the roll stand in the transport direction after rolling. Using an acquisition device that operates non-contact without mechanical action on the planar rolling material, at least one two-dimensional dataset of the surface of the planar rolling material is repeatedly acquired at the output side of the roll stand, the values of which depend on the external flatness locally spread at each corresponding location of the planar rolling material, and / or on the internal stress locally spread at each corresponding location of the planar rolling material. Each two-dimensional dataset is received by an evaluation device of the roller assembly. For strips of planar rolling material extending in the transport direction, the evaluation device determines an error value for each strip that depends on the flatness error. The evaluation device supplies the determined error value to a control device of the roller assembly, which then takes the determined error value into consideration when determining control variables for the flatness control elements of the roll stand. Therefore, as a result of the cooperation of the detection device, evaluation device, control device, and roll stand, a closed feedback control loop that operates in real time is obtained. [Prior art documents] [Patent Documents]
[0011] [Patent Document 1] U.S. Patent Application Publication No. 2015 / 0116727 [Patent Document 2] International Publication No. 2019 / 068376 Brochure [Patent Document 3] International Publication No. 2021 / 105364 brochure [Patent Document 4] Japanese Patent Application Publication No. 4-279208 [Patent Document 5] European Patent Application Publication No. 2258492 Specification [Overview of the project] [Problems that the invention aims to solve]
[0012] For example, when rolling metal strips or other metal materials for planar rolling, there has been a constant attempt to produce a material that is stress-free within itself, free from external distortion, and therefore flat. This is true when the material is rolled perfectly uniformly in the width direction, that is, when rolled with a consistent relative reduction in each pass in the width direction. If the rolling is not uniform in the width direction, some longitudinal strips (when viewed in the width direction) of the material will not be rolled as intensely as other longitudinal strips. As a result, the material will form waves in the longitudinal strips that are rolled more intensely. The material will not be flat, and its external (visible) flatness will be non-zero. In the case of small length differences, or when a planar rolling material is subjected to tension, even if the outer surface flatness is equal to 0, the length difference will result in a difference in internal tensile stress. Therefore, in this case, the internal stress of the planar rolling material is not 0.
[0013] Various procedures are known in the prior art for uniformizing and correcting flatness errors. See Patent Document 1 and Patent Document 2, which have already been mentioned.
[0014] The environmental conditions when rolling metal materials for planar rolling are extremely harsh, resulting in factors such as high temperatures, vibrations, water or steam, oil vapor, dust particles, and rust particles. This often hinders the operation of measuring devices for obtaining error values and imposes high spatial requirements.
[0015] The object of the present invention is to provide a possibility for identifying and correcting flatness errors in planar rolling materials in a simple and reliable manner.
[0016] The object is achieved by an operating method having the features of claim 1. Advantageous embodiments of the operating method are the subject matter of dependent claims 2 to 12.
Means for Solving the Problems
[0017] According to the invention, in the operating method of the type first mentioned, an evaluation device for determining the respective error value of each strip determines the intensity of the local oscillation and the spatial frequency of the data values of the strip of each two-dimensional data set corresponding to each strip, and determines the respective error value based on the intensity and / or the spatial frequency.
[0018] The intensity and the spatial frequency can be determined, for example, using the Fourier transform of the data values.
[0019] In the simplest case, the acquisition device is in the form of a camera device, and by the camera device, each two-dimensional image of the surface of the flat rolling material is acquired as each two-dimensional data set, or is determined based on the acquired image data. The camera device can specifically be in the form of a "normal" camera used for directly acquiring two-dimensional images. Alternatively, a plurality of such cameras may be used, and thus a plurality of two-dimensional data sets can also be acquired accordingly. The camera device can also be a line scan camera. Alternatively, a plurality of such line scan cameras can also be used. Cameras that can be used in rolling mills are generally known to those skilled in the art. The camera can be used depending on the form of the camera for detecting light in the visible spectrum and / or the infrared range. From a structural perspective, the camera can be in the form of, for example, a CCD camera.
[0020] The acquired image may be a normal intensity image. Alternatively, the acquired image may be a so-called depth image, in which depth information is associated with each location in the two-dimensional image or dataset, thus ultimately yielding a three-dimensional image. Optionally, multiple "normal" two-dimensional images can be acquired by multiple cameras, from which a depth image is determined by fusing the acquired images or similar methods and transmitted to an evaluation device.
[0021] The camera device has an associated illumination device, which can be used to illuminate the image area acquired by the camera device in a predetermined manner. Optionally, the illumination device can modulate the illumination of the acquired image area. As a result, the signal-to-noise ratio can be improved in several situations.
[0022] Preferably, a two-dimensional dataset is used to capture the surface of the planar rolling material across its entire width. As a result, the evaluation of each two-dimensional dataset can be improved.
[0023] The phrase "across the entire width" means that the two-dimensional dataset includes the lateral edges of the material for planar rolling. Therefore, within the scope of evaluating the two-dimensional dataset, the position of the lateral edges may also be considered.
[0024] Preferably, the acquisition device is positioned centrally above the planar rolling material when viewed in a plane defined by the width and transport directions. Thus, the positioning of the acquisition device can be such that a line extending from the acquisition device to the planar rolling material and oriented perpendicular to the surface of the planar rolling material intersects with the planar rolling material at its center. In this case, the acquisition device is positioned directly or nearly directly above the planar rolling material, or above the acquired area of the planar rolling material.
[0025] In many cases, the flatness control elements of a roll stand include locally acting control elements that, in each case, affect only a portion of the upper and / or lower working rollers of the roll stand. Such locally acting control elements may specifically be cooling devices used so that the coolant is applied only to the respective portions of the corresponding working rollers. When locally acting control elements are present, each strip of planar rolling material preferably corresponds to a portion of the upper and / or lower working rollers. As a result, a one-to-one relationship can be established between a determined error value and the associated control variable for the locally acting control element.
[0026] Preferably, the evaluation device for determining the error value of each strip selects a region of each strip, the region extending over the entire length of the strip when viewed in the transport direction of the planar rolling material, and extending over only a portion of the width of the strip when viewed in the width direction of the planar rolling material. In this case, the evaluation device determines the intensity and spatial frequency with respect to the region of each strip. This simplifies the determination of the respective error values.
[0027] It is possible for the evaluation device to preprocess each two-dimensional dataset before determining the intensity and spatial frequency.
[0028] Preprocessing may include, for example, smoothing (frequency filtering). Alternatively or additionally, preprocessing may include artifact removal. Artifacts may be errors caused by the acquisition configuration itself, for example. Alternatively, artifacts may be caused by material on the surface (e.g., scale). It is equally possible to identify the lateral edges of the planar rolling material and to consider the position of those lateral edges in the evaluation of the two-dimensional dataset. Specifically, only strips that are entirely located within the lateral edges of the planar rolling material can be formed and / or evaluated. Alternatively, evaluation can be performed only when the area of each strip used to determine the error value is entirely located within the lateral edges of the planar rolling material.
[0029] The data values are often intensity values. In such cases, preprocessing may include normalizing the intensity values with respect to the maximum possible range of values in a two-dimensional dataset, and adjusting them by averaging the data values of each segment or region based on each segment or region of each segment. As a result, the evaluation is standardized.
[0030] The evaluation device can similarly verify the validity of the error values between the determination of intensity and spatial frequency and the determination of their respective error values. For example, if the strips have relatively small widths, the error values of adjacent strips can be compared with each other. If the error value determined for a particular strip differs considerably from the determined error values of adjacent strips, this may indicate an inaccurate evaluation. Similarly, if the determined spatial frequencies for adjacent strips differ considerably from each other, this may indicate an inaccurate evaluation. The reason for such differences could be, for example, a scale patch. Furthermore, the error values should exhibit a distribution of values that resembles a bell-shaped curve when viewed across multiple strips, specifically, a Gaussian distribution.
[0031] Preferably, the evaluation device determines each error value using at least the intensity and / or spatial frequency of the largest local vibration. For example, in the simplest case, only the intensity of the largest local vibration may be used. Alternatively, only the spatial frequency of the largest local vibration may be used. However, preferably, each error value is determined based on a combination of the intensity and spatial frequency of the largest local vibration. In either case, the corresponding characteristic curve can be stored in the evaluation device.
[0032] The material for planar rolling may be hot-rolled or cold-rolled in a roll stand, as needed.
[0033] Generally, there are no other roll stands between the roller assembly's roll stand and the acquisition device. This is true regardless of whether the roller assembly's roll stand is the only roll stand on the rolling mill, the last roll stand in a multi-stand rolling mill train, or any roll stand other than the last roll stand in a multi-stand rolling mill train.
[0034] The objective is further achieved by a computer program having the features of claim 13. According to the present invention, the processing of the computer program has the effect that an evaluation device for determining the error value of each strip determines the intensity and spatial frequency of the local vibration of the strip data value of each strip in each two-dimensional dataset corresponding to each strip, and determines the respective error value based on the intensity and / or spatial frequency.
[0035] The processing of the computer program may also have the effect that the evaluation device performs some of the advantageous embodiments of the method of operation described above.
[0036] The objective is further achieved by an evaluation device having the features of claim 15. According to the present invention, the evaluation device is programmed with a computer program according to the present invention, thereby the evaluation device cooperates with an acquisition device and a control device for the roll stand of a roller assembly in accordance with the operating method according to the present invention.
[0037] The objective is further achieved by a roller assembly having the features of claim 16. According to the present invention, an evaluation device for a roller assembly is a form of evaluation device according to the present invention.
[0038] The above-mentioned characteristics, features, and advantages of this invention, and the manner in which they are achieved, will be more clearly understood in conjunction with the following description of exemplary embodiments, which will be explained in more detail together with the schematic drawings. [Brief explanation of the drawing]
[0039] [Figure 1] This is a side view of the roller assembly. [Figure 2] This is a view of the roller assembly shown in Figure 1 from above. [Figure 3] This is a diagram of a two-dimensional dataset. [Figure 4] This is a flowchart. [Figure 5] This is a diagram of a two-dimensional dataset. [Figure 6] This is a flowchart. [Figure 7] This is a map of the area. [Figure 8] This is a flowchart. [Figure 9] This is a diagram showing the data profile. [Figure 10] This is a diagram of the local spectrum. [Figure 11] This is a diagram of a rolling material with non-flat regions. [Figure 12] This is a diagram of a single step in a flowchart. [Figure 13] This is a diagram of a rolling mill array with multiple stands. [Figure 14] This is a diagram of a rolling mill array with multiple stands. [Modes for carrying out the invention]
[0040] According to Figures 1 and 2, the roller assembly has a roll stand 1. The rolling material 2 is rolled using the roll stand 1. After being rolled on the roll stand 1, the rolling material 2 leaves the roll stand 1 in the transport direction x.
[0041] The rolling material 2 consists of a metal, often steel. Alternatively, the rolling material 2 may consist of, for example, aluminum or copper. The rolling material 2 can be cold-rolled in the roll stand 1. However, generally, the rolling material 2 is hot-rolled.
[0042] As illustrated in Figure 2, the rolling material 2 is a planar rolling material, that is, a strip or plate. This implicitly follows from the illustration of the roll stand 1 in Figure 1, as a roll stand having additional rollers 4, specifically support rollers, in addition to its operating rollers 3. The rolling material 2 extends in the width direction y across the rolling material width b.
[0043] The roll stand 1 includes a flatness control element 5 that acts globally and / or a flatness control element 6 that acts locally. The flatness of the rolling material 2 leaving the roll stand 1 can be adjusted using both the flatness control element 5 and the flatness control element 6. The flatness control element 5 is a control element whose operation necessarily affects the flatness of the rolling material 2 over the entire width b of the rolling material. Examples of such control elements include a bending device for bending the working roller 3, a pushing device for axially displacing the working roller 3 and / or further rollers 4, and other control elements such as control elements for so-called paired intersections. The flatness control element 6 can exist as a substitute or addition to the flatness control element 5. Using the flatness control element 6, individual parts of the upper working roller 3 and / or lower working roller 3 can be individually affected. This is shown in Figure 2 for the upper working roller 3. Such locally acting control elements may specifically be cooling devices used so that the coolant 7 is applied only to each portion of the corresponding operating roller 3.
[0044] The roller assembly further includes an acquisition device 8. As shown in Figure 2, the acquisition device 8 is preferably centrally located above the rolling material 2 when viewed in a plane defined by the width direction y and the transport direction x. Generally, there are no other roll stands between the roll stand 1 and the acquisition device 8.
[0045] By using the acquisition device 8, at least one two-dimensional dataset D of the surface of the rolling material 2 is repeatedly acquired, mostly at a fixed period time T (see Figure 4). The period time T can correspond to several image rates per second, such as 24 images / second, 30 images / second, or 60 images / second. Other values are also possible. As a result of the placement of the acquisition device 8 on the output side of the roll stand 1, the dataset D also relates to the rolling material 2 on the output side of the roll stand 1.
[0046] The acquisition device 8 operates without mechanical action on the rolling material 2 and without contact. For example, the acquisition device 8 can take the form of a camera device, and using the camera device, each two-dimensional image of the surface of the rolling material 2 is acquired as a respective two-dimensional dataset D. If there are multiple cameras, the acquisition device 8 can either use the two-dimensional dataset D supplied by the cameras itself, or determine the resulting two-dimensional image of the surface of the rolling material 2 as the resulting two-dimensional dataset D based on the acquired image data from the multiple cameras. See the corresponding comments in the introduction to this description. If necessary, an illumination device operating in a modulated manner at the optional rate may be further associated with the camera device.
[0047] Dataset D may occur individually or sequentially. If the acquisition device 8 is in the form of a camera device, the transmitted dataset D or images may be individual images in, for example, JPEG format or other suitable format, or sequentially occurring video images in, for example, MPEG format or mp4 format.
[0048] As shown in Figure 3, dataset D comprises numerous data values DW. To avoid making Figure 3 unnecessarily excessive, only a portion of the data values DW are shown in Figure 3. However, regardless of the specific form of the acquisition device 8, each data value DW corresponds to a corresponding location on the rolling material 2, depending on the depiction in Figure 3, i.e., depending on their location in the two-dimensional dataset D in which they are involved. Thus, the surface of the rolling material 2 is mapped to the location in dataset D. Each data value DW itself specifically depends on the external flatness of the rolling material 2, which extends locally at each corresponding location on the rolling material 2. Optionally, each data value DW itself additionally depends on the internal stress of the rolling material 2, which extends locally at each corresponding location on the rolling material 2.
[0049] The roller assembly further includes an evaluation device 9. The evaluation device 9 is connected to an acquisition device 8 for data transfer. As a result of the connection for data transfer, the evaluation device 9 can receive a dataset D from the acquisition device 8. The structure and operating principle of the evaluation device 9 is a core subject of the present invention.
[0050] The evaluation device 9 is generally in the form of a software-programmable device, as indicated in Figure 1 by "μP" within the evaluation device 9. The evaluation device 9 is programmed with a computer program 10, which includes machine code 11 that can be processed directly by the evaluation device 9. As a result of programming with the computer program 10, or as a result of processing the machine code 11, the evaluation device 9 performs a series of steps described below in conjunction with Figure 4.
[0051] In step S1, the evaluation device 9 receives each dataset D (or, optionally, multiple datasets D) from the acquisition device 8.
[0052] In step S2, the evaluation device 9 performs preprocessing of dataset D. Step S2 is optional; therefore, it may be omitted. For this reason, step S2 is simply indicated by a dashed line in Figure 3.
[0053] In step S3, the evaluation device 9 divides the dataset D into strips 12 (see Figure 3, which shows one of the strips 12). As is clear, the strips 12 (or the region of the rolling material 2 corresponding to the strips 12) extend in the transport direction x. The strips 12 may have the same width as each other. This is true in most cases. However, this is not absolutely necessary.
[0054] In step S4, the evaluation device 9 performs a local frequency analysis of each strip 12 individually in the transport direction x. For example, the evaluation device 9 can perform a Fourier transform in step S4, as indicated by "FOU" in Figure 4. However, in either case, the evaluation device 9 uses the frequency analysis to determine the intensity of the local vibrations and the spatial frequencies of each local vibration, thereby ultimately determining the local spectra as well.
[0055] Based on local frequency analysis, and subsequently using local frequency analysis, the evaluation device 9 determines the error value PF for each band segment 12 individually in step S5. Thus, the evaluation device 9 evaluates the local spectrum determined for each band segment 12.
[0056] As a result of the mapping rules in acquiring dataset D, strip 12 in dataset D corresponds to the corresponding strip 13 of rolling material 2. One of the strips 13 is shown in Figure 2. As a result of the given correspondence between strip 12 and strip 13, the error value PF can also be directly assigned to the corresponding strip 13 of rolling material 2.
[0057] As already mentioned, there are flatness control elements 5 that act on the whole and / or flatness control elements 6 that act only locally. When flatness control elements 5 that act on the whole are present, the number of strips 12 can be determined as needed. This is also possible in principle when there are only locally acting flatness control elements 6 that act only on each portion of the working roller 3, or in particular. However, in this case, the strips 12 are preferably determined such that the corresponding strips 13 of the rolling material 2 correspond to such portions of the upper working roller 3 and / or the lower working roller 3, respectively, as shown in the illustration in Figure 2. The term “corresponding” is not necessarily understood in this regard to a 1:1 correspondence. Rather, the term is meant so that each individual strip 12 / 13 can be associated with exactly one portion, and therefore, corresponding individual strips 12 / 13 cannot be part of multiple portions at the same time. However, conversely, it is entirely possible for multiple strips 12 / 13 to be in an individual portion.
[0058] The flatness error of the rolling material 2 is defined as δL / L, where L is the minimum length of each corresponding strip 13 of the rolling material 2 in a stress-free state, and δL is the difference in length of each strip 13 of the rolling material 2 that is longer than the minimum length. The error value PF is not generally the same as the flatness error, but depends on the flatness error.
[0059] Therefore, in step S6, the evaluation device 9 supplies the determined error value PF to the control device 14. For this purpose, the evaluation device 9 is connected to the control device 14 for data transfer, also referencing Figure 1.
[0060] The control device 14 is also part of the roller assembly. The control device 14 considers the error value PF transmitted to it in determining the control variables S for the flatness control elements 5 and 6 of the roll stand 1. The error value is considered in such a way that the error value PF is corrected as much as possible, that is, so that the resulting flatness of the rolling material 2 is as close as possible to the desired flatness. The control device 14 outputs the control variables S to the flatness control elements 5 and 6.
[0061] From step S6, the evaluation device 9 returns to step S1. Thus, the evaluation device 9 repeatedly performs steps S1 to S6. Generally, the steps are performed with a fixed period time T. This period time T should preferably not exceed the control frequency of the control device 14.
[0062] Ultimately, the cooperation of the acquisition device 8, evaluation device 9, control device 14, and roll stand 1 (or its flatness control elements 5, 6) results in a closed feedback control loop that operates in real time and is used to ensure that the error value PF can be corrected and eliminated as much as possible.
[0063] The acquisition range of the acquisition device 8 is preferably determined such that the surface of the rolling material 2 is acquired over the entire width b of the rolling material 2 using the two-dimensional dataset D, as shown in the illustration in Figure 5. Therefore, each two-dimensional dataset D also includes the lateral edges 15 (or images thereof) of the rolling material 2.
[0064] Possible procedures used for frequency analysis of the error value PF and the determination therefrom are described below in conjunction with Figure 6. Figure 6 shows possible implementations of steps S4 and S5 in Figure 4.
[0065] As shown in Figure 6, in step S11, the evaluation device 9 selects one of the strips 12. In step S12, the evaluation device 9 selects a region 16 of the strip 12 selected in step S11. As shown in Figure 7, region 16 extends over the entire length of the strip 12 selected in step S11 when viewed in the transport direction x. In contrast, when viewed in the width direction y, region 16 extends over only a portion of the width of the strip 12 selected in step S11. In step S13, the evaluation device 9 performs a frequency analysis on the data value DW of only that region 16 (see step S4), and based on this, determines the error value PF for the strip 12 selected in step S11 (see step S5).
[0066] In general, the procedure in Figure 6 yields better results as the area in the width direction y selected in step S12 becomes narrower. In extreme cases, it is possible for area 16 to extend in the width direction y to only span a single line of the two-dimensional dataset D.
[0067] In step S14, the evaluation device 9 checks whether steps S11-S13 have already been performed for all strips 12. If not, the evaluation device 9 returns to step S11. In this case, when the evaluation device 9 performs step S11 again, it selects a different strip 12 for which steps S11-S13 have not yet been performed. Otherwise, the procedure in Figure 6 is completed.
[0068] Within the scope of the embodiments shown in Figures 6 and 7, the evaluation device 9 can even select multiple regions 16 for each individual strip 12. In this case, the evaluation device 9 evaluates each region 16 individually and then determines an error value PF for the corresponding strip 12 based on the evaluation results of each region 16. For example, the evaluation device 9 can determine a preliminary error value for each of the regions 16 and then determine the resulting error value PF based on the preliminary error value. Specifically, the evaluation device 9 can use the largest preliminary error value as the resulting error value PF, or it can use a weighted average or unweighted average of the determined preliminary error values.
[0069] The pretreatment mentioned in step S2 can be performed if necessary. Possible procedures have already been described. Further possible pretreatments are described in more detail below, in conjunction with Figure 8. This pretreatment may be performed as needed, as an alternative or addition to other possible pretreatments.
[0070] Within the scope of Figure 8, it is assumed that each dataset D is a "normal" intensity image (grayscale image) of the rolling material 2. Therefore, the data value DW is an intensity value. In this case, as illustrated in Figure 8, for example, in step S21, the normalization of the intensity value with respect to the maximum possible range of values of the two-dimensional dataset D, which maps to a value range between 0 and 1, can be performed first, i.e., it is almost linear. Just as an example, if the data value DW is a gray value with an 8-bit data depth, that is, if the data value DW can be located between 0 (=black) and 255 (=white), then the data value DW of 0 is assigned the value 0 without change, the data value DW of 1 is assigned the value 1 / 255, the data value DW of 2 is assigned the value 2 / 255, and so on. Step S21 can be performed simultaneously across all strips 12 because it is independent of the positioning of the data value DW in dataset D. The newly determined data value DW, that is, a value in the range of 0 to 1, then appears in place of the original data value DW, that is, a value in the range of 0 to 255, for example.
[0071] In step S22, the evaluation device 9 selects a region of the dataset D. This region may be a strip 12, or, in the (preferred) embodiment shown in Figures 6 and 7, a region 16 within the strip 12. In step S23, the evaluation device 9 determines the average M of the data values DW for the region selected in step S22. Thus, the evaluation device 9 forms a sum of the data values DW and divides this sum by the number n of data values DW in the sum. In step S24, the evaluation device 9 subtracts the average M from the data values DW of the region selected in step S22.
[0072] In step S25, the evaluation device 9 checks whether steps S22-S24 have already been performed for all strip segments 12. If not, the evaluation device 9 returns to step S22. In this case, when the evaluation device 9 performs step S22 again, it selects a different region where steps S22-S24 have not yet been performed. Otherwise, the procedure shown in Figure 8 is completed.
[0073] Therefore, the repeated execution of steps S22 to S25 results in an adjustment by the average M of the data values DW of each strip 12 or area 16, based on the strip 12 or area 16 of the strip 12.
[0074] The procedure in Figure 8 is also evident from the illustration in Figure 9. Figure 9 shows, using only an example, the profile of the data value DW in the transport direction x for the area 16 with the minimum width in the width direction y. The scale transformation of the data value DW after steps S21 to S25 are performed is shown on the left side of Figure 9, and the scale transformation of the data value DW before steps S21 to S25 are performed is shown on the right side. The numbers in the x-coordinate can be, for example, the number of cells in the transport direction x.
[0075] Figure 10 shows a typical spectrum in a local region, for example, obtained for each strip 12 by performing steps S4 and S5. The number of vibrations per meter may be plotted on the horizontal axis, for example, and the associated intensity of the frequency in any unit may be plotted on the vertical axis. As illustrated in Figure 10, the spectrum has (1) a significant peak, i.e., the highest intensity I0. This is because, specifically, the associated waveform in the rolling material 2 is very regular, where a lack of flatness occurs, as seen in Figure 11.
[0076] The highest intensity I0 occurs at the associated spatial frequency f0. Preferably, as illustrated in Figure 12, the evaluation device 9 uses the highest intensity I0 and / or the associated spatial frequency f0 to determine the error value PF for the corresponding strip 12. It is possible that only the highest intensity I0 and / or the associated frequency f0 may be used in determining the error value PF. In the simplest case, for example, only the highest intensity I0 may be used. Alternatively, only the spatial frequency f0 of the largest local vibration may be used. However, preferably, each error value PF is determined based on a combination of the highest intensity I0 and the associated spatial frequency f0. Alternatively, other intensities and / or their respective associated spatial frequencies may be used in determining the error value PF. In any case, it is possible to store the corresponding (one-dimensional) characteristic curve or (multidimensional) set of characteristic curves in the evaluation device 9.
[0077] The roll stand 1 of the roller assembly according to the present invention can be the sole roll stand of the rolling mill. Alternatively, the roll stand 1 of the roller assembly according to the present invention may be part of a multi-stand rolling mill train. In this case, the roll stand 1 of the roller assembly according to the present invention may be either the last roll stand of a multi-stand rolling mill train as shown in the (simplified) illustration in Figure 13, or any roll stand other than the last roll stand of a multi-stand rolling mill train as shown in the (similarly simplified) illustration in Figure 14. However, in both cases, there are no other roll stands between the roll stand 1 of the roller assembly according to the present invention and the acquisition device 8.
[0078] The present invention has many advantages. Specifically, the acquisition device 8 is simple, robust, and inexpensive. It can be installed compactly and in a space-saving manner. Furthermore, it is possible to place the acquisition device 8 at a sufficiently large distance from the rolling material 2 so that the load on the acquisition device 8 from dust, water, heat, etc., is relatively small. The structure is modular. Therefore, the individual components, specifically the acquisition device 8, evaluation device 9, and control device 14, can be changed and replaced. Only the interfaces of the replaced components need to be made compatible.
[0079] Although the present invention has been illustrated and described in detail by preferred exemplary embodiments, the present invention is not limited to the examples disclosed, and other modifications can be derived by those skilled in the art without departing from the scope of protection of the present invention. [Explanation of symbols]
[0080] 1 Roll Stand 2. Materials for rolling 3. Operating roller 4. Further Laura 5. Flatness control element 6 Control Elements 7 Coolant 8 Acquisition Devices 9. Evaluation Devices 10 Computer Programs 11 Machine Code 12 strips 13 strips 14 Control Devices 15 Lateral edge 16 areas b. Width of material for rolling D dataset DW Data Values f0 is the spatial frequency with the highest intensity. I0 Highest intensity M average PF error value S control variable S1~S25 Step T period time x transport direction y width direction
Claims
1. A method of operation for a roller assembly, A planar rolling material (2) of metal, extending in the width direction (y) over the width (b) of the rolling material, is rolled using the roll stand (1) of the roller assembly, and after the planar rolling material (2) is rolled away from the roll stand in the transport direction (x). At least one two-dimensional dataset (D) of the surface of the planar rolling material (2) is repeatedly acquired at the output side of the roll stand (1) using an acquisition device (8) that operates non-contact without mechanical action on the planar rolling material (2), and the value (DW) of the two-dimensional dataset depends at least on the locally extending outer flatness at each corresponding location of the planar rolling material (2). Each of the two-dimensional datasets (D) is received by the evaluation device (9) of the roller assembly, and the evaluation device determines a flatness error-dependent error value (PF) for each of the strips (13, 12) of the planar rolling material (2) extending in the transport direction (x), using the strips (12) of the two-dimensional dataset (D) corresponding to the strips (13). The evaluation device (9) supplies the determined error value (PF) to the control device (14) of the roller assembly, and the control device (14) then considers the determined error value (PF) when determining the control variables (S) for the flatness control elements (5, 6) of the roll stand (1). In this operating method, as a result of the cooperation of the acquisition device (8), the evaluation device (9), the control device (14), and the roll stand (1), a closed feedback control loop that operates in real time is obtained. The evaluation device (9) for determining the error value (PF) of each of the strip pieces (13) is characterized by performing a local frequency analysis of the corresponding strip piece (12) in the transport direction (x), thereby determining the intensity and spatial frequency of the local vibration of the data values of the strip piece (12) in the two-dimensional dataset (D) corresponding to each of the strip pieces (13), and determining the respective error value (PF) based on the intensity and / or spatial frequency.
2. The operation method according to claim 1, characterized in that the acquisition device (8) is in the form of a camera device, and using the camera device, each two-dimensional image of the surface of the planar rolling material (2) is acquired as each two-dimensional dataset (D), or determined based on the acquired image data.
3. The operating method according to claim 1 or 2, characterized in that the surface of the planar rolling material (2) is acquired over the entire width (b) of the planar rolling material (2) using the two-dimensional dataset (D).
4. The operation method according to claim 1, characterized in that the acquisition device (8) is positioned in the center above the planar rolling material (2) when viewed in a plane defined by the width direction (y) and the transport direction (x).
5. The operation method according to claim 1, characterized in that the flatness control elements (5, 6) of the roll stand include a locally acting control element (6), and using the control element (6), in each case only a portion of the upper operating roller (3) and / or the lower operating roller (3) is affected, and the strip (13) of the planar rolling material (2) each corresponds to a portion of the upper operating roller (3) and / or the lower operating roller (3).
6. The method of operation according to claim 1, wherein the evaluation device (9) for determining the error value (PF) of each of the strips (13) is characterized by selecting a region (16) of each of the strips (12), the region (16) extending over the entire length of each strip (13) when viewed in the transport direction (x) of the planar rolling material (2), and extending over only a portion of the width of each strip (13) when viewed in the width direction (y) of the planar rolling material (2), and the evaluation device (9) determining the intensity and spatial frequency with respect to the region (16) of each of the strips (13).
7. The method of operation according to claim 1, characterized in that the evaluation device (9) performs preprocessing of each of the two-dimensional datasets (D) before determining the intensity and spatial frequency.
8. The method of operation according to claim 7, characterized in that the data value (D) is an intensity value, and the preprocessing includes normalizing the intensity value with respect to the maximum possible range of values in the two-dimensional dataset (D), and adjusting it by the average (M) of the data values (DW) of each of the strips (13) or the area (16) based on each of the strips (13) or the area (16).
9. The operating method according to claim 1, characterized in that the evaluation device (9) determines the respective error values (PF) using at least the intensity (I0) and / or spatial frequency (f0) of the largest local vibration.
10. The operating method according to claim 1, characterized in that the planar rolling material (2) is hot-rolled or cold-rolled in the roll stand (1).
11. The operating method according to claim 1, characterized in that there is no other roll stand between the roll stand (1) and the acquisition device (8) of the roller assembly.
12. The operation method according to claim 11, characterized in that the roll stand (1) of the roller assembly is the only roll stand of a rolling mill, the last roll stand of a rolling mill array of multiple stands, or a roll stand other than the last roll stand of a rolling mill array of multiple stands.
13. A computer program comprising a machine code (11) that can be directly processed by an evaluation device (9) of a roller assembly, wherein the processing of the machine code (11) by the evaluation device (9) is performed repeatedly by the evaluation device (9) during the operation of the roll stand (1) as the planar rolling material (2) of metal is rolled and the roll stand (1) moves out in the transport direction (x) after the planar rolling material (2) has been rolled, with the control device (14) of the roll stand (1) and the acquisition device (8) which operates non-contact without mechanical action on the planar rolling material (2). The acquisition device (8) receives at least one two-dimensional dataset (D) acquired by the acquisition device (8) of the surface of the planar rolling material (2) on the output side of the roll stand (1), and the value (DW) of each of the two-dimensional datasets (D) depends at least on the locally extending outer surface flatness at each corresponding location of the planar rolling material (2), For each strip (13) of the planar rolling material (2) extending in the transport direction (x), an error value (PF) dependent on the flatness error is determined for each strip (13, 12) using the strips (12) of the two-dimensional data set (D) corresponding to each strip (13). For consideration in determining the control variables (S) for the flatness control elements (5, 6) of the roll stand (1), the determined error value (PF) is supplied to the control device (14). Thus, it has a collaborative effect, As a result of the cooperation of the acquisition device (8), the evaluation device (9), and the control device (14), a closed feedback control loop that operates in real time is obtained in a computer program. The evaluation device (9) for determining the error value (PF) of each of the strip pieces (13) is characterized by performing a local frequency analysis of the corresponding strip piece (12) in the transport direction (x), thereby determining the intensity and spatial frequency of the local vibration of the data values of the strip piece (12) in each of the two-dimensional datasets (D) corresponding to each strip piece (13), and determining the error value (PF) of each based on the intensity and / or spatial frequency.
14. The computer program according to claim 13, characterized in that the processing of the machine code (11) by the evaluation device (9) has the effect of performing the features of any one of claims 6 to 9 while the roll stand (1) is operating.
15. An evaluation device for a roller assembly, wherein the device is programmed to cooperate with the acquisition device (8) and the control device (14) of the roll stand (1) of the roller assembly in accordance with the operating method described in claim 1.
16. It is a roller assembly, The roller assembly has a roll stand (1), the roll stand (1) is equipped with flatness control elements (5, 6), and the roll stand (1) rolls a planar rolling material (2) of metal that extends in the width direction (y) over the width (b) of the rolling material, and after rolling, guides it out of the roll stand (1) in the transport direction (x). The roller assembly has an acquisition device (8), which operates non-contact without mechanical action on the planar rolling material (2), and the acquisition device (8) repeatedly acquires at least one two-dimensional data set (D) of the surface of the planar rolling material (2) on the output side of the roll stand (1), and the data value (DW) of the two-dimensional data set depends at least on the localized outer flatness at each corresponding location of the planar rolling material (2). The roller assembly has an evaluation device (9) as described in claim 15, the evaluation device (9) is connected to the acquisition device (8) for data transfer to receive repeated two-dimensional data sets (D) of the surface of the planar rolling material (2) acquired using the acquisition device (8), and determines a flatness error-dependent error value (PF) for each strip (13) of the planar rolling material extending in the transport direction (x), using the strip (12) of the two-dimensional data sets (D) corresponding to each strip (13), and supplies the determined error value (PF) to the control device (14) of the roller assembly. The control device (14) is a roller assembly that takes into account the determined error value (PF) in determining the control variable (S) for the flatness control elements (5, 6) of the roll stand (1).