Rolling with frequency characteristics in mind

Abstract

Rolling considering frequency characteristics. A metal strip (1) is fed into the rolling mill stand (2) of the rolling equipment at an inlet speed (v) by a conveying mechanism (3) arranged in front, and the metal strip is rolled in the rolling mill stand (2). A measuring mechanism (4) located between the conveying mechanism (3) and the rolling mill stand (2) detects the thickness value (d) of the metal strip (1) for each successive segment (9) of the metal strip (1) and conveys it to the control mechanism (6) of the rolling equipment. The control mechanism (6) obtains a corresponding temporary thickness deviation based on the deviation between the corresponding thickness value (d) and the target thickness for the corresponding segment (9) of the metal strip (1), and obtains the final thickness deviation based on the temporary thickness deviation. The control mechanism (6), taking into account the description of the inverse frequency characteristics of the rolling mill stand (2) and / or the conveying mechanism (3) and / or the measuring mechanism (4), obtains control values ​​(A2, A3) for the rolling mill stand (2) and / or the conveying mechanism (3) respectively for the corresponding segment (9) of the metal strip (1) and the final thickness deviation of multiple preceding and / or subsequent segments (9) of the metal strip (1). The control mechanism outputs the corresponding control values ​​(A2, A3) to the rolling mill stand (2) and / or the conveying mechanism (3) in a timely and accurate manner.

Description

Technical Field

[0001] This invention relates to a method for operating rolling mill equipment.

[0002] -The metal strip is fed to the rolling mill stand of the rolling equipment at an inlet speed by a conveying mechanism arranged in front of the rolling mill stand.

[0003] -The metal strip is rolled in the rolling mill stand.

[0004] -Thickness values ​​for the thickness of the metal strip are detected for each successive segment of the metal strip by means of a measuring mechanism arranged between the conveying mechanism and the rolling mill stand.

[0005] The detected thickness value is then transmitted to the control mechanism of the rolling equipment.

[0006] -The control mechanism obtains a corresponding temporary thickness deviation based on the deviation between the corresponding thickness value and the target thickness for the corresponding segment of the metal strip.

[0007] - wherein the control mechanism obtains the final thickness deviation based on the temporary thickness deviation,

[0008] The control mechanism acquires control values ​​for the rolling mill stand and / or the conveying mechanism for each segment of the metal strip and outputs the corresponding control values ​​to the rolling mill stand and / or the conveying mechanism in a timely manner.

[0009] Furthermore, the present invention relates to a control program comprising machine code executable by a control mechanism for a rolling mill, wherein the machine code is executed by the control mechanism such that the control mechanism operates the rolling mill in a manner consistent with the operating method.

[0010] Furthermore, the present invention relates to a control mechanism for rolling mills, wherein the control mechanism is programmed with a control program such that the control mechanism operates the rolling mills in a manner that allows it to operate.

[0011] Furthermore, the present invention relates to a rolling mill for rolling metal strips.

[0012] -The rolling equipment wherein the rolling mill has at least one rolling mill stand, a conveying mechanism arranged in front of the rolling mill stand, a measuring mechanism and a control mechanism arranged between the conveying mechanism and the rolling mill stand.

[0013] - wherein the metal strip is conveyed to the rolling mill stand by the conveying mechanism at an inlet speed.

[0014] -The metal strip is rolled in a rolling mill stand.

[0015] -The thickness value for the thickness of the metal strip is detected by the measuring mechanism for each successive segment of the metal strip.

[0016] The detected thickness value is then transmitted to the control mechanism.

[0017] -The control mechanism thereunder operates the rolling equipment according to such an operating method.

[0018] In the manufacture of metal strip, the slab is first hot-rolled after casting to produce hot-rolled strip. The thickness of the hot-rolled strip is typically in the range of a few millimeters, sometimes slightly higher or lower depending on the manufacturing method; for example, between 1.0 mm and 20 mm for standard hot-rolling equipment and between 0.6 mm and 6 mm for so-called ESP equipment. In some cases, the hot-rolled strip is further processed without further thickness reduction. In other cases, the strip thickness is further reduced in a cold-rolling mill after hot rolling. The purpose of cold rolling is to produce cold-rolled metal strip whose final thickness matches the target thickness as well as possible and with the smallest possible deviation.

[0019] Typically, finished hot-rolled strip, i.e., after hot rolling but before being rolled in a cold rolling mill, has thickness variations. These thickness variations often have not only periodic components but also random components. If these variations are not compensated for, the metal strip will also have such variations after cold rolling. Although the absolute degree of the variation is less than that in the case of hot-rolled strip, the relative variation remains unchanged.

[0020] Therefore, if, for example, the metal strip has a thickness of 3.0 mm and a thickness deviation in the range of 30 μm before cold rolling, and furthermore, the metal strip has a thickness of 1.0 mm after cold rolling, then the metal strip has a thickness deviation in the range of 10 μm after cold rolling without compensation for the thickness deviation. Background Technology

[0021] To compensate for such deviations, different processing methods are known in the prior art.

[0022] Therefore, as is known, for example from EP 0 435 595 A2, the thickness of the rolled metal strip is detected at the exit side of the rolling mill stand, and the thickness of the rolling mill stand is adjusted. Furthermore, tension fluctuations are compensated for, as these fluctuations also affect the thickness of the rolled metal strip. To achieve relatively high dynamic adjustment of the tension, rollers or similar elements are present, on the one hand, between the conveying mechanism and the rolling mill stand, and on the other hand, between the rolling mill stand and the receiving mechanism located after the rolling mill stand. These rollers or similar elements allow the metal strip to be shifted within the rolling mill stand before and / or after rolling. The processing method of EP 0 435 595 A2 is based on the idea that adjustment performed by the conveying mechanism itself and the receiving mechanism itself is very slow, and that the dynamics of adjustment can be improved by the additional rollers. EP 0 435 595 A2 also describes a processing method in which the thickness and speed of the metal strip are detected at the entrance side of the rolling mill stand and used within the range of adjustments to the rolling mill stand.

[0023] As is also known from EP 3 332 883 A1, the thickness of the rolled metal strip is detected at the exit side of the rolling mill stand, and the thickness of the rolling mill stand is adjusted accordingly. Periodic deviations must be distinguished from random deviations. Periodic deviations are considered to be caused by the eccentricity of the rolls passing through the rolling mill stand. Accordingly, the adjustment of the rolling mill stand is corrected.

[0024] In JP 58 068 414 A, the thickness of the unrolled metal strip is detected at the entrance side of the rolling mill stand and an average value is taken within a certain length unit. This average value is used to control the adjustment of the rolling mill stand. Summary of the Invention

[0025] Existing technologies can already compensate for deviations in inlet thickness to some extent. However, these technologies can still be improved.

[0026] The objective of this invention is to provide a feasible solution by means of which excellent compensation for thickness deviations on the entry side of the metal strip can be achieved.

[0027] This task is solved by an operating method for rolling mills. An advantageous design of this operating method is the subject of this discussion.

[0028] According to the present invention, the type of operation method mentioned at the beginning is designed in such a way that the control mechanism obtains the corresponding control value based on the final thickness deviation of the corresponding segment of the metal strip and the final thickness deviation of multiple segments of the metal strip before and / or after the corresponding segment of the metal strip, taking into account the description of the inverse frequency characteristics of the rolling mill stand and / or the conveying mechanism and / or the measuring mechanism.

[0029] The inventors have recognized that the degree to which a particular thickness deviation is corrected depends not only on the thickness deviation itself but also on its frequency spectrum. Generally, higher frequency thickness deviations are compensated only within a smaller range and with a greater phase delay than those with lower frequencies. In order to compensate for higher frequency thickness deviations across the entire range without phase delay, the frequency characteristics of the adjusted mechanism must be taken into account, typically the adjustment of the roll gap size of the rolling mill stand and the magnitude of the conveying mechanism and / or the tension on the inlet side for the inlet speed. If necessary, the measured values ​​also have frequency characteristics, which can also be considered in this case. This consideration is based on a description of the inverse frequency characteristics of the rolling mill stand and / or the conveying mechanism and / or the measuring mechanism.

[0030] There are different feasible solutions for the aforementioned considerations and methods. Currently, the preferred approach is:

[0031] - A description of the inverse frequency characteristics of the control mechanism for the rolling mill stand and / or conveying mechanism and / or measuring mechanism is pre-defined using an inverse model.

[0032] - The final thickness deviation of the metal strip segments is fed to the inverse model respectively, and

[0033] The control mechanism, by means of the inverse model, tracks the internal state of the inverse model and obtains the corresponding control values, while utilizing the corresponding final thickness deviation.

[0034] This approach is associated with minimal computational overhead.

[0035] Alternatively, the control mechanism may be given a pre-defined description of the inverse frequency characteristics of the rolling mill stand and / or conveying mechanism and / or measuring mechanism as the frequency response, and the control mechanism obtains the corresponding control value by transforming the final thickness deviation variation curve into a frequency range, then multiplying the transformed variation curve of the final thickness deviation by the frequency response, and then returning it to a time range. This approach results in particularly high-quality results.

[0036] It is generally known that multiplication in the frequency range corresponds to convolution in the time range. Therefore, as an alternative, and entirely equivalent, it is possible to pre-define the inverse frequency characteristics of the rolling mill stand and / or conveying mechanism and / or measuring mechanism as the convolution kernel, and the control mechanism obtains the corresponding control value by convolving the curve of the final thickness deviation with the convolution kernel.

[0037] The detection of the frequency response, and the acquisition or parameterization of the inverse model, or the acquisition of the gain for each frequency range, or the acquisition of the convolution kernel, can be performed automatically. In particular, defined small disturbances can be introduced into the roll gap target value of the rolling mill during continuous operation of the rolling mill. These disturbances are reflected in the corresponding fluctuations in the thickness of the strip at the exit side of the rolling mill. If a measuring mechanism is arranged behind the rolling mill to detect the thickness at the exit side, the frequency response can be obtained automatically through a combined evaluation of the introduced disturbances on one hand and the thickness fluctuations at the exit side on the other. This is known in principle to those skilled in the art.

[0038] Preferably, the control mechanism utilizes not only the final thickness deviation of the metal strip in the segments preceding the corresponding section of the metal strip, but also the final thickness deviation of the metal strip in the segments following the corresponding section of the metal strip, in order to obtain the respective control values. This ensures particularly reliable acquisition of the corresponding control values. This is especially applicable if the number of segments of the metal strip preceding the corresponding section (where the control mechanism uses the final thickness deviation of the segments to obtain the corresponding control values) is substantially equal to the number of segments of the metal strip following the corresponding section, in which case the control mechanism uses the final thickness deviation of the segments to obtain the corresponding control values.

[0039] In its simplest case, the control mechanism accepts the temporary thickness deviation 1:1 as the final thickness deviation. However, it is preferable that the control mechanism extracts the final thickness deviation from the temporary thickness deviation using zero-phase filtering. This approach results in more stable and robust operation of the rolling mill stand and / or conveying mechanism. This is particularly applicable if the temporary thickness deviation is low-pass filtered using zero-phase filtering.

[0040] Furthermore, the task is accomplished through a control program. According to the invention, the execution of the computer program causes the control mechanism to operate the rolling mill according to the operating mode of the invention.

[0041] Furthermore, the task is accomplished by a control mechanism. According to the invention, the control mechanism is programmed with a control program according to the invention, causing the control mechanism to operate the rolling mill according to the operating method according to the invention.

[0042] Furthermore, the task is accomplished using a rolling mill. According to the invention, the control mechanism operates the rolling mill according to the operating method of the invention. Attached Figure Description

[0043] The features, characteristics, and advantages of the present invention described above, and the ways and methods of achieving these features, characteristics, and advantages, will become clearer and more readily understood in conjunction with the following description of embodiments, which will be explained in detail with reference to the accompanying drawings. Here, in the schematic diagrams:

[0044] Figure 1 The rolling equipment is shown.

[0045] Figure 2 A flowchart is shown.

[0046] Figure 3 A metal strip is shown.

[0047] Figure 4 A flowchart is shown.

[0048] Figure 5 This illustrates one structural configuration of the control mechanism.

[0049] Figure 6 Another structural configuration of the control mechanism is shown.

[0050] Figure 7 The frequency response is shown.

[0051] Figure 8 Another structural configuration of the control mechanism is shown.

[0052] Figure 9 The convolution kernel is shown, and

[0053] Figure 10 Another structural configuration of the control mechanism is shown. Detailed Implementation

[0054] according to Figure 1 The rolling mill for rolling the metal strip 1 has a rolling mill stand 2. The metal strip 1 is rolled in the rolling mill stand 2. The rolling mill stand 2 is particularly capable of being a cold rolling stand, in which the metal strip 1 is therefore cold rolled. Figure 1 Only the work rolls of the rolling mill stand 2 are shown. Typically, the rolling mill stand 2 additionally includes at least two support rolls (a four-roll stand), and in some cases, more rolls. For example, the rolling mill stand 2 can be configured as a six-roll stand (two work rolls, two intermediate rolls, and two support rolls), a 12-roll stand, or a 20-roll stand. The metal strip 1 can be made of steel, aluminum, or other metals, such as copper or brass.

[0055] Furthermore, the rolling mill has a conveying mechanism 3. The conveying mechanism 3 is arranged in front of the rolling mill stand 2. The metal strip 1 is conveyed to the rolling mill stand 2 by the conveying mechanism 3 at an inlet speed v. Figure 1 The conveying mechanism 3 is configured as a coiler. However, it can also be configured differently, for example, as a drive or other rolling mill stand different from the rolling mill stand 2. The conveying mechanism 3 can also be configured as a so-called S-roll, i.e., multiple rolls through which the metal strip 1 is guided in an S-shape.

[0056] To be precise, the speed at which the metal strip 1 enters the rolling mill stand 2 and the speed at which the metal strip 1 is output by the conveying mechanism 3 (e.g., when it is uncoiled) must be different from each other. The speed at which the metal strip 1 enters the rolling mill stand 2 is determined by the circumferential speed of the work rolls of the rolling mill stand 2 and the backslip within the rolling mill stand 2. The speed at which the metal strip 1 is output by the conveying mechanism 3 is determined, for example, by the rotational speed of the coil and the current, time-varying diameter of the coil, in the case of a coiler. For a short period, there may be a slight difference between these two speeds. If such a short-term difference exists, then the tension in the metal strip 1 between the conveying mechanism 3 and the rolling mill stand 2 changes. However, the following discussion only refers to the entry speed v. Unless explicitly stated otherwise, the following discussion refers to the speed at which the metal strip 1 is output by the conveying mechanism 3 in cases of doubt.

[0057] A measuring mechanism 4 is arranged between the conveying mechanism 3 and the rolling mill stand 2. The thickness value d of the metal strip 1 is repeatedly and consistently detected by means of the measuring mechanism 4. Furthermore, another measuring mechanism 5 can be present, by means of which the measurement value of the inlet velocity v is also consistently detected.

[0058] The detected thickness value d, and if necessary, the detected inlet velocity value v, are transmitted to the control mechanism 6, which is also a component of the rolling mill. The control mechanism 6 repeatedly acquires control values ​​A2 and A3 for the rolling mill stand 2 and / or the conveying mechanism 3. Typically, the control mechanism 6 acquires two control values, A2 and A3.

[0059] The control value A2 for the rolling mill stand 2 typically affects at least the adjustment of the rolling mill stand 2, i.e., the adjustment of the roll gap. For example, the corresponding control value A2 can be output to a so-called HGC (Hydraulic gap Control). Alternatively, the control value A2 can affect the main drive of the rolling mill stand 2, i.e., change the rolling torque or rolling speed. The control value A2 often affects not only the adjustment of the rolling mill stand 2 but also its main drive. In this case, the control value A2 for the rolling mill stand 2 can be considered a vector parameter. The control value has components for the adjustment of the rolling mill stand 2 and components for the main drive of the rolling mill stand 2.

[0060] The control value A3 is fed to the speed or torque adjustment mechanism of the conveying mechanism 3 and affects the inlet speed v and / or the tension present in the metal strip 1 at the inlet side of the rolling mill stand 2. Whenever necessary, another mechanism arranged before the conveying mechanism 3 must also be controlled within the scope of control of the conveying mechanism 3. While considering such an additional mechanism complicates the calculation of the control value A3, it does not alter the principle of the invention.

[0061] The control mechanism 6 is programmed using a control program 7. The control program 7 includes machine code 8, which can be executed by the control mechanism 6. The result of programming the control mechanism 6 with the control program 7 or executing the machine code 8 through the control mechanism 6 is that the control mechanism 6 operates the rolling mill according to an operating method, which is explained in detail below. Here, we first refer to… Figure 2 And then refer to Figure 3 and 4 .

[0062] according to Figure 2 In step S1, the control mechanism 6 receives the detected thickness value d and, if necessary, the detected value for the inlet velocity v. In step S2, the control mechanism 6 obtains the deviation δ between the detected thickness value d and the target thickness d*, hereinafter referred to as thickness deviation δd.

[0063] The thickness deviation δd obtained in step S2 is only a temporary thickness deviation δd. In step S3, the control mechanism 6 obtains the corresponding final thickness deviation δd' based on the temporary thickness deviation δd. In the simplest case, step S3 is negligible. In this case, the control mechanism 6 receives the temporary thickness deviation δd 1:1 as the final thickness deviation δd'.

[0064] However, it is preferable to perform real-time acquisition, thereby obtaining the final thickness deviation δd' from the temporary thickness deviation δd using acquisition rules that are not insignificant. For example, in step S3, the control mechanism 6 performs zero-phase filtering to obtain the final thickness deviation δd'. With the aid of zero-phase filtering, a filtered curve of the numerical value (here, the time-varying curve of the temporary thickness deviation δd) is obtained from the original curve of the numerical value (here, the time-varying curve of the temporary thickness deviation δd), wherein no systematic phase shift occurs between the original curve and the filtered curve. Typically, and within the scope of this invention, the zero-phase filtering is a low-pass filtering, thus filtering out high-frequency fluctuations. In particular, the low-pass filtering significantly improves the stability of the inverse modeling of the rolling mill stand 2, the conveying mechanism 3, and / or the measuring mechanism 4.

[0065] For zero-phase filtering, it is also necessary to know the temporary thickness deviation δd of segment 9 following such segment 9 (the final thickness deviation δd' of this segment should be obtained). Therefore, for zero-phase filtering, step S3 is performed for another segment 9, whose temporary thickness deviation δd has been detected.

[0066] Zero-phase filtering is well known to those skilled in the art. The so-called IIR (infinite impulse response) can be mentioned purely exemplarily. Another feasible approach to implement zero-phase filtering is to convolve the temporary thickness deviation δd with the symmetric impulse response of an FIR filter (FIR = finite impulse response).

[0067] In step S4, the control mechanism 6 acquires control values ​​A2 and A3 using the final thickness deviation δd'. In step S5, the control mechanism 6 outputs control values ​​A2 and A3 to the rolling mill stand 2 and / or the conveying mechanism 3. Then, the control mechanism 6 returns to step S1.

[0068] Therefore, the above combination Figure 2 The explained processing method is executed repeatedly and periodically. In most cases, the processing method is even executed strictly according to clock pulses, that is, at a fixed periodic time T of, for example, 8 ms.

[0069] The following is combined with Figure 3 and 4 The processing method according to the present invention will be explained in more detail again.

[0070] Figure 3The metal strip 1 is shown from above. The metal strip 1 is virtually divided into segments 9. Some segments of segment 9 are... Figure 3 In addition to the reference numeral 9 in the attached figure, lowercase letters (such as a, b, etc.) are added to allow for individual descriptions of these sections.

[0071] In each cycle, that is, during each execution of step Sl, the thickness value d of a specific segment 9, such as segment 9a, is detected and transmitted to the control mechanism 6. Therefore, the detected thickness value d, the corresponding temporary thickness deviation δd, and the corresponding final thickness deviation δd' are associated with this segment 9a.

[0072] During the same cycle, another segment 9, such as segment 9b, is rolled in the rolling mill 2. The geometric spacing between segments 9a and 9b on the metal strip 1 corresponds to the geometric spacing a of the measuring mechanism 4 away from the rolling mill 2.

[0073] A specific time interval T' is required to transport segment 9a from measuring mechanism 4 to rolling mill stand 2. This time interval T' is usually referred to as the transport time. The time interval is determined by the inlet speed v of the metal strip 1 and the distance a of the measuring mechanism 4 from the rolling mill stand 2. With a constant inlet speed v, the relationship T' = a / v applies.

[0074] The time interval T' is typically significantly longer than the period time T. Therefore, there are multiple additional segments 9, such as segment 9c, between segments 9a and 9b. For these segments 9, the corresponding thickness value d has been detected before segment 9b is rolled. Furthermore, the metal strip 1 has segments 9, such as segment 9d, that have already been rolled in the rolling mill 2.

[0075] Since the time interval T' is required to transport the corresponding segment 9 from the measuring mechanism 4 to the rolling mill stand 2, it is possible that, in a specific cycle, although the thickness value d for segment 9a is detected in step S1 and a temporary thickness deviation δd is obtained for this segment 9a, control values ​​A2 and A3 for, for example, segment 9c are obtained in step S4, and furthermore, in step S5, the control values ​​A2 and A3 obtained in step S4 are output to the rolling mill stand 2 and / or the conveying mechanism 3. If necessary, the control values ​​A2 and A3 already obtained for segment 9 between segment 9c and segment 9b can also be output to the rolling mill stand 2 and / or the conveying mechanism 3. In the last case, it is only necessary to assign the thickness value d detected in the corresponding cycle, the thickness deviations δd and δd' obtained in the corresponding cycle, and the control values ​​A2 and A3 obtained in the corresponding cycle to the corresponding segment 9 and perform path tracking for segment 9. The corresponding processing method is as follows: Figure 4The following method is used to delineate the corresponding segments 9a, 9b, and 9c in steps S1 to S5, where steps S1 to S5 are performed for each segment. The control values ​​A2 and A3 output by the control mechanism 6 are not related to segment 9b, but to segment 9c or the segment between segment 9b and segment 9c. This is because a certain dead time of the rolling mill stand 2 and / or the conveying mechanism 3 must be taken into account.

[0076] The implementation of path tracking is well known to those skilled in the art. This makes it possible to correctly output the control values ​​A2 and A3 to the rolling mill stand 2 and / or the conveying mechanism 3 at the correct time. Within the scope of this invention, "correct time" means that the control values ​​A2 and A3 output to the rolling mill stand 2 and / or the conveying mechanism 3 affect the metal strip 1 at a given moment, at which moment the corresponding segment 9 of the metal strip 1 is rolled in the rolling mill stand 2. Here, the time interval T' can also be taken into account as needed, and, if necessary, the reaction time (dead time) of the rolling mill stand 2 and / or the conveying mechanism 3 can also be taken into account. The reaction time is the time required for the rolling mill stand 2 and / or the conveying mechanism 3 to react to the newly conveyed control values ​​A2 and A3. Furthermore, dead times occurring in communication between different mechanisms or in automation can also be taken into account. The acquisition of the control values ​​A2 and A3 must, of course, be completed before output.

[0077] Since the corresponding thickness value d has been detected for the segment 9 between segment 9a and segment 9c, and therefore the corresponding temporary thickness deviation δd is also known, and furthermore, the final thickness deviation δd' is also known at least for the segment 9 adjacent to segment 9c in the direction toward segment 9a, it is possible, for example, in order to obtain the control values ​​A2, A3 for segment 9c, not only to consider the final thickness deviation δd' of the segment 9c, but also to consider any of the other final thickness deviations δd', provided they have already been obtained. For example, in addition to the thickness deviation δd' of the segment 9c, the control mechanism 6 can also consider the final thickness deviations δd' of multiple adjacent segments 9 toward segment 9a. As an alternative or supplementary solution, in addition to the thickness deviation δd' of the segment 9c, the control mechanism 6 can consider the final thickness deviations δd' of multiple adjacent segments 9 toward segment 9b, and, if necessary, can also consider the final thickness deviations outside of segment 9b.

[0078] Within the range of acquiring the corresponding control values ​​A2 and A3, the control mechanism 6 further considers the description of the inverse frequency characteristics for the rolling mill stand 2 and / or the conveying mechanism 3 and / or the measuring mechanism 4. Therefore, the control mechanism 6 is given a pre-defined description that directly characterizes the corresponding frequency characteristics themselves. In other words, the frequency characteristics can be acquired based on the mentioned description. The feasible schemes for pre-defined descriptions of the frequency characteristics will be explained in detail below. The control mechanism 6 thus acquires the corresponding control values ​​A2 and A3 not only in one manner and method, but also by means of the corresponding inverse frequency characteristics. More precisely, the control mechanism 6 explicitly knows the corresponding inverse frequency characteristics themselves. Therefore, the control mechanism 6 knows the following characteristic parameters that define the inverse frequency characteristics. This will be explained in detail below in conjunction with the rolling mill stand 2. Similar explanations apply to the conveying mechanism 3 and, if necessary, to the measuring mechanism 4.

[0079] The rolling mill stand 2 can be modeled in different ways and methods. In the simplest case, the rolling mill stand 2 is modeled as a PTL element. Alternatively, higher-order modeling is considered. This modeling describes the rolling mill stand 2 itself, including its control mechanisms if necessary. The transport time, i.e., the time period T', is not part of the modeling.

[0080] The frequency characteristics of the rolling mill 2 can be described, for example, by a transfer function. If—as is usually the case—G denotes the transfer function itself and the letter s denotes the Laplace operator, then the transfer function G(s) can be written as:

[0081]

[0082] Here, b i (where i = 1, 2, ..., m) and C j (where j = 1, 2, ..., n) are constant coefficients. The degree m of the numerator polynomial is at most as large as the degree n of the denominator polynomial. If the rolling mill stand 2 is modeled as a PTL element, then for example as

[0083]

[0084] The transfer function G(s) is derived from this, where T2 is the time constant representing the rolling mill stand 2.

[0085] For the corresponding inverse transfer function G -1 (s) generally applies

[0086]

[0087] The inverse transport property G- 1(s) is thus explicitly defined. If the rolling mill stand 2 is modeled as a PT1 element, then precisely...

[0088]

[0089] To derive the corresponding inverse transfer function G -1 (s).

[0090] If the inverse transfer function G -1 (s) If the modeled characteristics of the rolling mill 2 are not accurately represented, the modeled characteristics of the rolling mill 2 will generally become unstable. In some cases, the actual characteristics of the rolling mill 2 may even become unstable. For example, the reciprocal of the PT1 element generates the PD element. The PD element greatly enhances the high frequency. However, it is practically impossible to achieve the theoretically attainable output signal of the PD element. This is due to the adjustment limitations of the actuator, specifically the adjustment limitations of the rolling mill 2. To ensure stability and feasibility, the inverse transfer function G is... -1 The denominator polynomial of (s) extends the inverse transfer function G. -1 The component of s in the molecule of (s) is proportional to the highest power of s. This is known from the outset to those skilled in the art. In this regard, reference can be made to Thomas Frenz's professional book "Stabile Neuronale Online Identification and Compensation und Kompensation statischer Nichtlinearitäten" (Static Nonlinear Stable Neuron Online Identification and Compensation). Therefore, the inverse modeling actually used for the frequency characteristics of the rolling mill 2 is achieved through the modified inverse transfer function G. -1 (s) describes the inverse transfer function as follows:

[0091]

[0092] TC is a relatively small time, that is, a time much smaller than the time constant T2 representing the rolling mill stand 2. The smaller the time TC can be chosen, the better the modeling of the inverse frequency characteristics of the rolling mill stand 2 will be. In practice, the time TC is chosen to be equal to or approximately equal to the period time T.

[0093] As already mentioned, a similar interpretation applies to the conveying mechanism 3. If the conveying mechanism 3 is modeled using PTl elements similar to the rolling mill stand 2, then the modified inverse transfer function G... -1 (s) describes the inverse transfer function G used for the conveying mechanism 3. -1 (s), the modified inverse transfer function has the following form:

[0094]

[0095] Where T3 is the time constant representing the conveying mechanism 3.

[0096] Based on the above facts, it is possible to... Figure 5 The illustration shows the corresponding inverse model 10 of the rolling mill stand 2 pre-defined by the control mechanism 6. As explained above, the inverse model 10 describes the inverse frequency characteristics of the rolling mill stand 2, and, if necessary, includes the inverse frequency characteristics of the measuring mechanism 4. Constant dead time, etc., can be considered within or outside the inverse model 10 as needed, within a reserved time T2'. For example, the inverse model 10 can be based on... Figure 5 The diagram in the image achieves the following form

[0097]

[0098] inverse transfer function G -1 (s).

[0099] The final thickness deviation δd' of segment 9 of the metal strip 1 is transmitted to the inverse model 10 in clock pulses over a period of time T. The control mechanism 6, with the aid of the inverse model 10 and taking into account its internal state Z2, obtains the corresponding control value A2 for the rolling mill 2 and outputs the control value A2 to the rolling mill 2. Furthermore, the control mechanism 6 tracks the internal state Z2 using the corresponding final thickness deviation δd' and the previous internal state Z2 of the inverse model 10. It is necessary to take into account and track the internal state Z2, because otherwise the inverse model 10 of the rolling mill 2 could not store an understanding of the change curve of the final thickness deviation δd' to date, and thus could not model the frequency characteristics, but only the purely proportional characteristics.

[0100] A transport model 11 is arranged in front of the inverse model 10. The corresponding final thickness deviation δd' and inlet speed v are pulsed to the transport model 11 at periodic intervals T. The transport model 11 models the path tracking of the corresponding segment 9, to which the corresponding final thickness deviation δd' belongs. Furthermore, a reservation time T2' is supplied to the transport model 11. The transport model 11 outputs the corresponding final thickness deviation δd' with a delay relative to the moment it is supplied to the transport model 11. As already mentioned, this time delay is chosen such that the control value A2 output for a specific segment 9 takes effect at the moment when the corresponding segment 9 of the metal strip 1 is rolled in the rolling mill stand 2.

[0101] The modeling and implementation of path tracing are well known to those skilled in the art. Therefore, there is no need to explain them in detail.

[0102] Furthermore, the corresponding final thickness deviation δd' is not typically directly fed to the inverse model 10 of the rolling mill 2 by the transport model 11, but is instead multiplied beforehand by a static gain factor V2 in the multiplier 12. The multiplier 12 converts the corresponding final thickness deviation δd' into, for example, an additional target value for the roll gap of the rolling mill 2 or the main drive unit of the rolling mill 2. However, in principle, the gain factor V2 can also be integrated into the inverse model 10 of the rolling mill 2.

[0103] As already mentioned, the inverse frequency characteristics of the conveying mechanism 3 can also be modeled in a completely similar manner, including, if necessary, modeling the inverse frequency characteristics of the measuring mechanism 4. According to Figure 5 The illustration in the diagram produces a completely similar structure with inverse model 13, conveying model 14, and multiplier 15. T3' is the reserved time for the conveying mechanism, and V3 is the gain factor. By means of multiplier 15, the corresponding final thickness deviation δd' is converted into an additional target value for the inlet speed v of the metal strip 1. If the conveying mechanism 3 does not adjust the inlet speed v, but instead adjusts the tension present in the metal strip 1 at the inlet side of the rolling mill stand 2, the moment of inertia of the conveying mechanism 3 must be considered together with it if necessary.

[0104] If the control value A2 is a vector parameter having components for adjusting the rolling mill stand 2 and components for the main drive mechanism of the rolling mill stand 2, then the modeling for the rolling mill stand 2 explained above must be performed separately for each component of the vector parameter. That is, multiple inverse sub-models exist for the rolling mill stand 2 if necessary. However, this does not change the principle.

[0105] The rolling mill 2 is controlled by control value A2 to minimize fluctuations in the thickness of the metal strip 2 at the exit side of the rolling mill 2. The conveying mechanism 3 is controlled by control value A3 to keep the inlet speed v and / or the tension at the inlet side of the metal strip 1 as constant as possible. In particular, the tension affects the pass thinning in the rolling mill 2. To prevent variations in the tension in the metal strip 1 from undesirably affecting the pass thinning, the inlet speed v must be adjusted in sync with changes in the adjustment of the rolling mill 2 and changes in the circumferential speed of the work rolls of the rolling mill 2.

[0106] As already mentioned, the final thickness deviation δd is obtained by zero-phase filtering of the temporary thickness deviation δd. Therefore, corresponding zero-phase filters 16 and 17 can be arranged before or after the transport models 11 and 14. The zero-phase filtering can also be integrated into the corresponding transport models 11 and 14.

[0107] The transport models 11 and 14 are essentially of the same type. Therefore, it is possible to... Figure 6 The structure to Figure 5 The structure of control mechanism 6 was modified. As a result, in accordance with... Figure 6 In this design, one of the transport models 11 and 14 can be eliminated. Instead, a delay element 18 is provided to compensate for the difference between the reservation times T2' and T3'. Typically, the reservation time T3' will be greater than the reservation time T2'. Figure 6 In the case shown, the transport model 14 is omitted and the delay element 18 is arranged in the path for controlling value A2.

[0108] The above combination of the control mechanism 6 Figure 5 and 6 The explained structure is constructed as a software block in control mechanism 6. That is, the software block is formed based on programming performed using control program 7 and execution of machine code 8.

[0109] An alternative design of the present invention is that the description of the inverse frequency characteristics of the rolling mill 2—if necessary, a combined description of the inverse frequency response of the measuring mechanism 4—is given in advance as the frequency response FG. According to Figure 7 The diagram shows that the frequency response FG represents different frequency ranges FBk (where k = 1, 2, ...), indicating the corresponding complex gain V. This complex gain V must be used to amplify the time-variable thickness deviation (which has a frequency in the corresponding frequency range FB) in order to be fully compensated.

[0110] The frequency response FG is based on the following consideration: if a metal strip 1 of constant thickness is fed to the rolling mill 2 and the control value A2 fed to the rolling mill 2 is changed with a specific amplitude and a specific frequency, it indicates that, with the same amplitude (=input parameter) of the control value A2, the degree to which the thickness variation (=output parameter) is added to the metal strip 1 by the rolling mill 2 at the exit side depends on the frequency. Specifically, both the amplitude and phase of the thickness variation at the exit side change. In particular, it has been shown in practice that as the frequency increases, the amplitude of the thickness deviation at the exit side decreases and the phase delay increases. Therefore, according to the frequency of the final thickness deviation δd' at the inlet side, the correction parameters "position change of the rolling mill 2" and / or "torque change of the work roll" or "rotation speed change of the work roll" must be dynamically adjusted in terms of amplitude and phase to generate an optimal correction signal. In order to compensate for the final thickness deviation δd' that occurs at a higher frequency at the inlet side of the rolling mill 2, the rolling mill 2 must be controlled more forcefully.

[0111] The amplitude and phase of the response of the rolling mill 2 to the corresponding control value A2 can be combined into a complex factor for the corresponding frequency. The reciprocal of the corresponding complex factor corresponds to a complex gain factor V for the corresponding frequency, which must be used to scale the thickness deviation at the corresponding frequency so that the thickness deviation is fully compensated at the exit side of the rolling mill 2. The entirety of these gain factors V, i.e., the gain factors V for different frequencies or frequency ranges FB, forms the frequency response FG pre-given by the control mechanism 6.

[0112] If the frequency response FG is a pre-defined description of the frequency characteristics of the control mechanism 6, then it is possible to determine the frequency response based on the frequency response FG. Figure 8 The diagram illustrates the processing to obtain the corresponding control value A2. If necessary, each component of the corresponding control value A2 must be processed individually. Figure 8 The handling method.

[0113] according to Figure 8For the corresponding segment 9 and multiple other segments 9, a corresponding final thickness deviation δd' is pre-defined for the control mechanism 8. The final thickness deviation δd' forms a time-varying curve. The control mechanism 6 transforms the time-varying curve into a frequency range in the transformation block 19. For example, the control mechanism 6 can implement a Fourier transform (FT), especially a short-time Fourier transform (STFT), in the transformation block 19. The Fourier transform can be continuous or discrete as needed. Similarly, it can be analog or digital as required. Furthermore, other transforms, such as the discrete cosine transform, can be considered as alternatives to the Fourier transform.

[0114] Regardless of the specific processing method, the control mechanism 6 acquires the frequency component FA of the variation curve via transformation block 19. In the subsequent acquisition block 20, the corresponding frequency component FA is multiplied individually for each frequency range FB by the gain factor V for the corresponding frequency range FB. Thus, the variation curve to be transformed is multiplied by the frequency response FG. Due to the multiplication over a complex frequency range, the result is not only scaled in amplitude but also shifted in phase. Through this multiplication, a corrected spectrum of the final thickness deviation δd' is generated in the frequency region, which optimally compensates for the frequency-dependent transmission characteristics of the rolling mill 2.

[0115] In another transformation block 21, the control mechanism 6 transforms the output signal of acquisition block 20, i.e., the frequency variation curve scaled by frequency, back into a time range. The transformation of transformation block 21 is the opposite of the transformation of transformation block 19. Finally, the control mechanism 6 selects the output signal that has been acquired for the corresponding segment 9 from the output signal of transformation block 21.

[0116] In accordance with Figure 8 The final thickness deviation δd' utilized within the range of processing methods can be determined as needed. It is particularly suitable to choose this number such that it is equal to a power of two, because the subsequent Fourier transform can then be implemented as a fast Fourier transform.

[0117] It is generally known to those skilled in the art that multiplication in the frequency range corresponds to convolution in the time range. Therefore, as a predefined alternative to the frequency response FG, it is possible to... Figure 9 The illustration shows that the control mechanism 6 pre-defines the convolution kernel FK. The convolution kernel FK can, for example, by... Figure 7 and 8 The frequency response FG is obtained by separately transforming it into a time range.

[0118] If the frequency characteristics of the control mechanism 6 are pre-defined as such a convolution kernel FK, then the corresponding control value A2 can be obtained according to... Figure 10 The diagram processing in the image is as follows:

[0119] As in Figure 8 Similarly, for the corresponding section 9 and multiple other sections 9, a corresponding final thickness deviation δd' must be pre-defined for the control mechanism 8. The final thickness deviation δd' is as follows: Figure 8 Similarly, a time-varying curve is formed. In acquisition block 22, the control mechanism 6 performs convolution on this curve using a convolution kernel FK. The control mechanism 6 selects the control value already acquired for the corresponding segment 9 from the output signal of acquisition block 22 as the control value A2.

[0120] In accordance with Figure 8 and 10 Within the scope of the processing method, only a single final thickness deviation δd' is explicitly re-transmitted to the control mechanism 6. Other required final thickness deviations δd' have already been transmitted to the control mechanism 6 within the execution range of previous cycles. Therefore, they only need to be stored in the middle and then recalled and used again.

[0121] The above is combined with the acquisition of the control value A2 for the rolling mill stand 2. Figure 8 and 10 The processing method has been explained. A completely similar processing method is possible to obtain the control value A3 for the conveying mechanism 3. In both cases, in addition to the frequency characteristics of the corresponding mechanisms 2 and 3, the frequency characteristics of the measuring mechanism 4 can also be taken into account as needed.

[0122] The upper part of the control mechanism 6 is combined Figure 8 and 10 The explained structure is like following Figure 5 and 6 The control mechanism 6 is structured similarly to other control mechanisms, being constructed as software blocks. In other words, they are formed based on programming using the control program 7 and the execution of machine code 8.

[0123] In each design of the invention, it is possible that the additional segments 9 (whose final thickness deviation δd' is considered within the range of the corresponding control values ​​A2, A3) are simply segments 9 preceding the corresponding segment 9 of the metal strip 1. Similarly, in Figure 8 and 10 In the design scheme, it is possible that the additional segment 9 is simply the segment 9 following the corresponding segment 9 of the metal strip 1. However, in Figure 8 and10 In the case of a design scheme, a mixed approach is generally adopted, i.e., if a portion of the other segment 9 precedes the corresponding segment 9 of the metal strip 1 and another portion of the other segment 9 follows the corresponding segment 9 of the metal strip 1, better results are usually obtained. For example, it is always possible to utilize segments 9 whose thickness d is detected in the corresponding cycle. Figure 3 The region represented by 23.

[0124] Figure 3 It also shows another advantageous design scheme. Because if the control values ​​A2 and A3 for segment 9c are to be obtained separately, then according to Figure 3 The number of segments 9 preceding the corresponding section 9 of the metal strip 1 shown in the diagram (the control mechanism 6 uses the final thickness deviation δd' of these segments to obtain the corresponding control value A2) is substantially equal to the number of segments 9 following the corresponding section 9 of the metal strip 1, the control mechanism 6 using the final thickness deviation δd' of these segments to obtain the corresponding control values ​​A2, A3. However, minor deviations (e.g., up to two segments 9 more or less) are generally not a problem. Furthermore, it is often appropriate to utilize a total of 2 n In this case, the number of segments 9 preceding the corresponding segment 9 of the metal strip 1 is preferably 1 greater or 1 less than the number of segments 9 following the corresponding segment 9 of the metal strip 1.

[0125] This invention offers numerous advantages. In particular, it provides a nearly complete correction of the thickness deviation δd on the inlet side in a simple manner. This is especially applicable if the control value A2 and control value A3 are obtained not only by means of the invention but also by means of the invention. Furthermore, existing rolling mills can be easily retrofitted according to this invention. This is because the hardware itself—namely, the rolling mill stand 2, the conveying mechanism 3, the measuring mechanisms 4 and 5, and the control mechanism 6—does not need to be changed. Only the control program 7 for the control mechanism 6 must be modified.

[0126] Although the invention has been illustrated and described in detail through preferred embodiments, it is not limited to the disclosed examples and other variations can be derived by those skilled in the art without departing from the scope of protection of the invention.

[0127] List of reference numerals in the attached diagram:

[0128] 1. Metal strip

[0129] 2 Rolling Mill Stand

[0130] 3 Conveying mechanism

[0131] 4, 5 Measuring mechanisms

[0132] 6. Control mechanism

[0133] 7 Control Procedure

[0134] 8 Machine Code

[0135] 9 sections

[0136] 10, 13 Inverse Model

[0137] 11, 14 Transportation Model

[0138] 12 and 15 multipliers

[0139] 16, 17 Zero-phase filters

[0140] 18 Delay Components

[0141] Transform blocks 19 and 21

[0142] Get blocks 20 and 22

[0143] 23 Scope

[0144] a Spacing

[0145] A2, A3 control values

[0146] d Thickness value

[0147] d * Target thickness

[0148] FA frequency components

[0149] FB frequency range

[0150] FG Frequency Response

[0151] FK convolution kernel

[0152] G transfer function

[0153] s Laplace operator

[0154] Steps S1 to S5

[0155] T is the cycle time.

[0156] T2 and T3 represent time constants

[0157] T2' and T3' retention times

[0158] v Entry speed

[0159] Gain factors of V, V2, and V3

[0160] Internal states of Z2 and Z3

[0161] δd, δd' thickness deviation

Claims

1. Operating methods for rolling equipment, -The metal strip (1) is conveyed to the rolling mill stand (2) of the rolling equipment by a conveying mechanism (3) arranged in front of the rolling mill stand (2) at an inlet speed (v). -The metal strip (1) is rolled in the rolling mill (2). -Thickness values ​​(d) for the thickness of the metal strip (1) are detected by means of a measuring mechanism (4) arranged between the conveying mechanism (3) and the rolling mill stand (2) for each successive segment (9) of the metal strip (1). -The detected thickness value (d) is transmitted to the control mechanism (6) of the rolling equipment. -The control mechanism (6) obtains a corresponding temporary thickness deviation (δd) based on the deviation between the corresponding thickness value (d) and the target thickness (d*) of the corresponding segment (9) of the metal strip (1). - wherein the control mechanism (6) obtains the final thickness deviation (δd') based on the temporary thickness deviation (δd). -The control mechanism (6) acquires control values ​​(A2, A3) for the sections (9) of the metal strip (1) for the rolling mill (2) and / or the conveying mechanism (3), and outputs the corresponding control values ​​(A2, A3) to the rolling mill (2) and / or the conveying mechanism (3) in a timely and accurate manner. Its features are, The control mechanism (6) obtains the corresponding control values ​​(A2, A3) based on the final thickness deviation (δd') of the corresponding segment (9) of the metal strip (1) and the final thickness deviation (δd') of multiple segments (9) of the metal strip (1) before and / or after the corresponding segment (9) of the metal strip (1), taking into account the description of the inverse frequency characteristics of the rolling mill (2) and / or the conveying mechanism (3) and / or the measuring mechanism (4) in advance, wherein the control mechanism (6) is given a description of the inverse frequency characteristics of the rolling mill (2) and / or the conveying mechanism (3) and / or the measuring mechanism (4) as a convolution kernel (FK), and the control mechanism (6) obtains the corresponding control values ​​(A2, A3) by convolving the curve of the change of the final thickness deviation (δd') with the convolution kernel (FK).

2. The operating method according to claim 1, Its features are, -A description of the inverse frequency characteristics of the rolling mill stand (2) and / or conveying mechanism (3) and / or measuring mechanism (4) is given in advance for the control mechanism (6) by using inverse models (10, 13). - The final thickness deviation (δd') of the segments of the metal strip (1) is fed to the inverse models (10, 13) respectively, and - The control mechanism (6) uses the inverse model (10, 13) to track the internal state (Z2, Z3) of the inverse model (10, 13) and obtain the corresponding control values ​​(A2, A3) by means of the inverse model (10, 13) in the case of the corresponding final thickness deviation (δd').

3. The operating method according to claim 1, Its features are, As a frequency response (FG), the control mechanism (6) is given a description of the inverse frequency characteristics of the rolling mill stand (2) and / or the conveying mechanism (3) and / or the measuring mechanism (4), and the control mechanism (6) obtains the corresponding control values ​​(A2, A3) by transforming the change curve of the final thickness deviation (δd') into a frequency range, then multiplying the change curve of the final thickness deviation (δd') transformation with the frequency response (FG) and then inversely transforming it into a time range.

4. The operating method according to claim 1 or 3, Its features are, In order to obtain the corresponding control values ​​(A2, A3), the control mechanism (6) utilizes not only the final thickness deviation (δd') of the metal strip (1) in the section (9) before the corresponding section (9) of the metal strip (1) but also the final thickness deviation (δd') of the metal strip (1) in the section (9) after the corresponding section (9) of the metal strip (1).

5. The operating method according to claim 4, Its features are, The number of segments (9) of the metal strip (1) before the corresponding segment (9) of the metal strip (1) is substantially equal to the number of segments (9) of the metal strip (1) after the corresponding segment (9) of the metal strip (1), wherein the control mechanism (6) uses the final thickness deviation (δd') of the preceding segment to obtain the corresponding control value (A2, A3), and the control mechanism (6) uses the final thickness deviation (δd') of the following segment to obtain the corresponding control value (A2, A3).

6. The operating method according to claim 1 or 2, Its features are, The control mechanism (6) receives the temporary thickness deviation (δd) 1:1 as the final thickness deviation (δd'), or the control mechanism (6) obtains the final thickness deviation (δd') from the temporary thickness deviation (δd) by means of zero-phase filtering.

7. The operating method according to claim 6, Its features are, The temporary thickness deviation (δd) is low-pass filtered using zero-phase filtering.

8. A control program comprising machine code (8) executable by a control mechanism (6) for a rolling mill, wherein the execution of the machine code (8) by the control mechanism (6) results in the control mechanism (6) operating the rolling mill according to the operating method according to any one of claims 1 to 7.

9. A control mechanism for a rolling mill, wherein the control mechanism is programmed with the control program (7) according to claim 8, such that the control mechanism operates the rolling mill according to the operating method according to any one of claims 1 to 7.

10. Rolling equipment for rolling metal strip (1), -The rolling equipment wherein the rolling mill has at least one rolling mill stand (2), a conveying mechanism (3) arranged in front of the rolling mill stand (2), a measuring mechanism (4) arranged between the conveying mechanism (3) and the rolling mill stand (2), and a control mechanism (6). -The metal strip (1) is conveyed to the rolling mill stand (2) by the conveying mechanism (3) at an inlet speed (v). -The metal strip (1) is rolled in the rolling mill stand (2), -The thickness value (d) for the thickness of the metal strip (1) is detected by the measuring mechanism (4) for each of the successive segments (9) of the metal strip (1). -The detected thickness value (d) is transmitted to the control mechanism (6). - wherein the control mechanism (6) operates the rolling equipment according to the operating method according to any one of claims 1 to 7.

Images