Roll stand with comprehensive compensation of eccentricities
Patent Information
- Application Number
- PCT/EP2025/055016
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-03-06
- Filing Date
- 2025-02-25
- Publication Date
- 2025-10-02
AI Technical Summary
Conventional thickness control systems in rolling stands fail to effectively compensate for periodic deviations in strip thickness caused by roll eccentricities and out-of-roundness, particularly when there are deviations between the operating and drive sides, leading to periodic disturbances in rolling force and tension.
A control method that compensates for both symmetrical and asymmetrical roll eccentricities by using a control device that determines target values for adjustment devices based on symmetrical and asymmetrical measured variables, with separate processing for each type of eccentricity, and employs observers to model the rolling stand's behavior and compensate for phase shifts.
The method effectively compensates for both symmetrical and asymmetrical roll eccentricities, improving strip thickness accuracy and reducing periodic disturbances in rolling force and tension, while being compatible with existing symmetrical compensation systems and applicable to both cold and hot rolling.
Smart Images
Figure EP2025055016_02102025_PF_FP_ABST
Abstract
Description
[0001] Description
[0002] Title of the invention
[0003] Rolling stand with comprehensive compensation of eccentricities
[0004] field of technology
[0005] The present invention is based on an operating method for a control device of a rolling stand in which a flat metal rolling stock is rolled at a rolling speed,
[0006] - wherein the control device of the rolling stand receives a symmetrical target roll gap, wherein the symmetrical target roll gap is a uniform value which is symmetrical over the barrel length of the rolls of the rolling stand,
[0007] - wherein the control device receives a measured variable detected by a measuring device and dependent on a symmetrical eccentricity occurring during rolling of the rolling stock in the rolling stand,
[0008] - wherein the control device for an operator-side and a drive-side adjustment device of the rolling stand, by means of which the roll gap is adjusted on the drive side and operator side of the rolling stand, determines respective target values taking into account the target roll gap and a symmetrical compensation value and controls the adjustment devices accordingly,
[0009] - whereby the control device determines the symmetrical compensation value using the symmetrical measured value,
[0010] - the control device takes into account the target roll gap and the symmetrical compensation value with the same sign when determining the target values for the adjusting devices.
[0011] The present invention further relates to a control program for a software-programmable control device for a rolling stand, wherein the control program comprises machine code which can be directly processed by the control device, wherein the processing of the machine code by the control device causes the control device to carry out such an operating method.
[0012] The present invention further relates to a control device for a rolling stand in which a flat metal rolling stock is rolled at a rolling speed, wherein the control device is programmed with such a control program so that the control device executes such an operating method. The present invention further relates to a rolling stand in which a flat metal rolling stock is rolled at a rolling speed,
[0013] - wherein the rolling stand has an operator-side and a drive-side adjustment device, by means of which the roll gap in which the rolling stock is rolled can be adjusted on the drive side and operator side of the rolling stand,
[0014] - wherein the rolling stand has a measuring device by means of which a measured variable is recorded which is dependent on a symmetrical eccentricity occurring during rolling of the rolling stock in the rolling stand,
[0015] - wherein the measuring device for supplying the measured variable is connected to a control device of the rolling stand and the control device for controlling the adjusting devices is connected to the adjusting devices,
[0016] - wherein the control device is designed as such a control device which controls the rolling stand according to such an operating method.
[0017] State of the art
[0018] The above-mentioned subject matters are generally known. Reference may be made purely by way of example to EP 0 170 016 B1, DE 10 2006 008 574 A1, and EP 0 698 427 A1. Furthermore, reference may be made to DE 10 2019 207497 A1, EP 0 424 709 A2, and AT 407 015 B. EP 3 419 771 B1, EP 0 407 628 A1, and DE 10 2004 039 829 B3 may also be mentioned in this context.
[0019] A similar procedure is also known from DE 38 44 202 A1. In DE 38 44 202 A1, the rolling forces are recorded separately for the operating and drive sides, and the respective eccentricity compensation values are determined separately for the operating and drive sides using separate processing and taken into account when controlling the adjustment devices. JP S62127113 A and JP S63199012 A also demonstrate similar procedures.
[0020] Summary of the invention
[0021] In both cold and hot rolling, the accuracy with which the flat rolled stock is rolled to a desired thickness is an important quality criterion. Therefore, the strip thickness is controlled to the target value within the basic automation system by adjusting the roll gap, the stand speed, the inlet and outlet tensions, and, if necessary, other adjustable parameters.
[0022] Roll eccentricities and out-of-roundness can cause periodic deviations in strip thickness. Such periodically occurring errors cannot be compensated for by conventional thickness control systems. Furthermore, roll eccentricities can cause periodic disturbances in the rolling force and in the tension in the rolled material, particularly in the inlet tension. Such disturbances must be suppressed by other control systems (e.g., an AGC (automatic gage control) or a strip tension control system (e.g., using filters).
[0023] To compensate for such errors, it is known in the state of the art to compensate for the eccentricity (see the above-mentioned documents).
[0024] As long as the roll eccentricity is uniform across the rolling direction (i.e., across the barrel length of the rolls of the rolling stand), the state-of-the-art procedures lead to good results. When applying the teaching of DE 38 44 202 A1, eccentricities are well compensated even when there are deviations between the situation on the operating side and the situation on the drive side. However, the teaching of DE 38 44 202 A1 requires a completely separate procedure for the operating side and the drive side and, above all, is incompatible with the usual compensation of symmetrical eccentricity. The procedure of DE 38 44 202 A1 is therefore often impractical.
[0025] The object of the present invention is to create possibilities by means of which the advantages of the teaching of DE 38 44 202 A1 - i.e. a compensation of eccentricities even when there are deviations between the situation on the operating side and the situation on the drive side - can be achieved, without having to accept the disadvantages of the teaching of DE 38 44 202 A1 - i.e. completely separate procedures for the operating side and the drive side, which are not compatible with the compensation of a symmetrical eccentricity.
[0026] The object is achieved by an operating method having the features of claim 1. Advantageous embodiments of the operating method according to the invention are the subject of dependent claims 2 to 10.
[0027] According to the invention, an operating method of the type mentioned at the outset is designed in that
[0028] - that the control device, in addition to the measured value dependent on the symmetrical eccentricity, also receives a measured value dependent on an asymmetrical eccentricity occurring during rolling of the rolling stock in the rolling stand,
[0029] - that the received measured variables comprise a drive-side and an operator-side measured variable and the control device determines the symmetrical measured variable by summing or averaging the received measured variables and the asymmetrical measured variable by subtracting the received measured variables,
[0030] - that the control device for the operator-side and drive-side adjustment device of the rolling stand also takes into account a target roll gap wedge and an asymmetric compensation value when determining the target values, whereby the target roll gap wedge is given implicitly with the value 0 or explicitly, - that the control device determines the asymmetric compensation value using the asymmetric measured value and
[0031] - that the control device takes into account the target roll gap wedge and the asymmetric compensation value with opposite signs when determining the target values for the adjustment devices.
[0032] The advantage of the inventive teaching becomes particularly apparent when the inventive teaching is to be retrofitted to a rolling stand that already compensates for symmetrical eccentricity. This is because, within the scope of the inventive teaching, the existing compensation for symmetrical eccentricity can be retained unchanged, so that only the compensation for asymmetrical eccentricity needs to be retrofitted. In contrast, DE 38 44 202 A1 requires a fundamentally new implementation.
[0033] The rolling of the rolled stock in the rolling stand can, as in the prior art, alternatively be hot rolling or cold rolling. The rolled stock is usually a strip. In individual cases, however, it can also be heavy plate. The rolled stock is often made of steel. However, it can also be made of another metal, for example, copper, brass, or aluminum. The adjusting devices are usually designed as hydraulic cylinder units. The target roll gap wedge can be explicitly specified. In this case, the target roll gap wedge can alternatively have the value 0 or - for example, when rolling a slit slab - a value other than 0. Alternatively, the target roll gap wedge can be implicitly specified. In this case, the target roll gap wedge has the value 0.
[0034] Preferably, separate processing takes place after determining the symmetrical measured variable and the asymmetrical measured variable. Thus, the control device uses the symmetrical measured variable, but not the asymmetrical measured variable, to determine the symmetrical compensation value, and the asymmetrical measured variable, but not the symmetrical measured variable, to determine the asymmetrical compensation value.
[0035] Preferably, the control device determines the setpoint value for the operator-side adjustment device by adding the symmetrical setpoint roll gap, the symmetrical compensation value, half of the setpoint roll gap wedge and half of the asymmetrical compensation value and determines the setpoint value for the drive-side adjustment device by adding the symmetrical setpoint roll gap and the symmetrical compensation value and subtracting half of the setpoint roll gap wedge and half of the asymmetrical compensation value.
[0036] If s* is the symmetrical target roll gap, k* the target roll gap wedge, ös the symmetrical compensation value and ök the asymmetrical compensation value, the relationships pDS = / (.s * +J + ( * + ) / 2) and pOS = / ( * + - ( * +^) / 2) preferably apply to the target value pOS on the operator side and the target value pDS on the drive side.
[0037] There are various options for determining the compensation values. Preferably,
[0038] - that the control device implements a first and a second observer, each comprising a plant model of the rolling stand, a number of periodic disturbance models and a state feedback,
[0039] - that the control device supplies the first observer with a speed variable characteristic of the rolling speed, the symmetrical target roll gap and the symmetrical measured variable and supplies the second observer with the speed variable, the target roll gap wedge and the asymmetrical measured variable and
[0040] - that the control device determines the symmetrical compensation value by means of the first observer and the asymmetrical compensation value by means of the second observer.
[0041] The speed variable can be the peripheral speed of a roll in the rolling stand. Based on the peripheral speed, in conjunction with the radius or diameter of the corresponding roll, the rotational speed of the respective roll can be easily determined. Alternatively, the speed variable can be the inlet or outlet rolling speed, which, in conjunction with the lead or retardation, can be converted into the peripheral speed of the rolls. The radii or diameters of the rolls can be readily known to the control device. The lead and retardation can also be readily known to the control device. Alternatively, the speed variable can be the rotational speed itself. In this case, no further conversion is necessary.
[0042] The plant models can be implemented in various ways. Currently, it is preferred that the respective plant model comprise three elements connected in series. The front element takes into account the reaction time of the adjustment devices. The middle element takes into account the sensitivity of the measured variable supplied to the respective plant model to the occurring eccentricities. The rear element takes into account the reaction time of the measuring devices. This design is simple, robust, and delivers good results.
[0043] The response time of the control devices can be taken into account in the respective plant model, for example, by a PTn element (with n = 1, 2, etc.). Similarly, the response time of the measuring devices can also be taken into account in the respective plant model, for example, by a PTn element (with n = 1, 2, etc.). As a rule, it is sufficient to model them as PT1 elements (i.e., with n = 1). This type of modeling corresponds to a low-pass behavior of the respective modeled element.
[0044] The sensitivity is considered as a factor. The factor can be specified as required. For example, in the case where the measured variables are local thicknesses of the rolling stock at the exit side of the rolling stand, the sensitivity can be specified as the inverse of a quotient increased by 1. The quotient contains the spring constant of the rolling stock as the numerator and the spring constant of the rolling stand as the denominator. If the spring constant of the rolling stock is denoted by cm and the spring constant of the rolling stand by cg, the (in this case dimensionless) sensitivity V is thus obtained as
[0045] Likewise, in the case where the measured values for rolling are local forces exerted on the rolling stock, the sensitivity can be specified as the product of the spring constant of the rolling stock and the spring constant of the rolling stand divided by the sum of the spring constant of the rolling stock and the spring constant of the rolling stand. In this case, the sensitivity V is given by
[0046] In this case, the sensitivity has the dimension N / m.
[0047] Furthermore, if the measured variables are tensions occurring on the inlet or outlet side during rolling of the rolled stock, a value determined during testing can be specified as the sensitivity. In this case, the sensitivity also has the dimension N / m.
[0048] Any DC component determined within the framework of the respective system model can alternatively be compensated by means of a respective integrator or filtered out by means of a high-pass filter arranged downstream of the measuring device.
[0049] Furthermore, it is possible for the respective plant model to include an additional element which is arranged downstream of the rear element and is designed as a high-pass filter, in particular as a DT1 element. This allows the plant model to model the behavior at low frequencies more precisely. In particular, the influence of external controls at low frequencies can be simulated. The plant model can be universal. It can therefore be used regardless of which of the aforementioned variables (thickness, tension, rolling forces) are recorded as measured variables. It is only necessary to specify the type of measured variables to be recorded during commissioning and to parameterize the plant models accordingly (e.g. the time constants for the two PTn elements and the value for the sensitivity).
[0050] There are also various options for the design of the periodic disturbance models. Preferably, the respective disturbance model comprises a front multiplier, a front integrator, a rear multiplier, and a rear integrator. In this case, the control device
[0051] - the front multiplier is supplied with the output signal of the rear integrator and an angular frequency of the respective disturbance model,
[0052] - the front integrator receives the output signal of the front multiplier,
[0053] - the output signal of the front integrator and the angular frequency of the respective disturbance model are sent to the rear multiplier,
[0054] - the output signal of the rear multiplier to the rear integrator,
[0055] - the output signal of the front integrator to the respective system model and
[0056] - the output signals of both integrators and the angular frequency of the respective disturbance model to a state feedback block assigned to the respective disturbance model.
[0057] This design is simple, robust and delivers good results.
[0058] The number of disturbance models for each observer can be determined as needed. For example, in a conventional four-high mill stand comprising two work rolls and two backup rolls, four disturbance models for the fundamental vibration and three harmonics of the respective backup roll can be available for each of the two backup rolls, and two disturbance models for the fundamental vibration and the first harmonic of the work rolls can be available for each of the two work rolls. The assignment of the respective disturbance model to a specific roll and a specific vibration is determined - regardless of the design of the roll stand - by the respective angular frequency. The respective angular frequency w is determined in a known manner by the relationship
[0059] CO = 2 - 7 - f , where f is the speed of the respective roller.
[0060] Often, the significant portion of the eccentricity is found in the fundamental vibration of the respective roller, while the amplitude of the eccentricity is significantly lower in the harmonics. In the simplest case, the control device adds the output signal of the front integrator to the output signal of the first term of the respective plant model. The output signals of the front integrators are added independently of each other—separately for the two observers. As a result, the output signals of the front integrators of the disturbance models of the respective observer are added to the output signal of the first term of the respective observer's plant model.
[0061] Preferably, the feedback block assigned to the respective disturbance model determines a respective component for the respective compensation value by utilizing the angular frequency of the respective disturbance model, the output signal of the front integrator of the respective disturbance model, the output signal of the rear integrator of the respective disturbance model, the reaction time of the adjustment devices, and a resulting delay time. This allows for the compensation of a phase shift resulting from the unavoidable delay times. Here, too, processing is performed separately for the individual disturbance models. The output signals of the feedback blocks are added – separately for the two observers.
[0062] The compensation of the phase shift is particularly simple and provides particularly good results if the feedback block assigned to the respective disturbance model
[0063] - a first and second modified output signal is determined based on the relationships zl'= zl + 0 • z2- TPC and z2'= T2 - a> • zl • TPC, where zT is the first modified output signal, z1 is the output signal of the front integrator, w is the respective angular frequency, z2 is the output signal of the rear integrator, TPC is the reaction time of the adjusting devices and z2' is the second modified output signal, and
[0064] - using the relationship z = zl'- cos(® • TG) -z2'- sin(o • T ) the respective share for the respective compensation value is determined, where z is the respective share for the respective compensation value and TG is the resulting delay time.
[0065] The delay time can have a relatively small or a relatively large value. The delay time has a relatively small value if the measured variables for rolling are forces exerted on the rolled stock or tensions occurring during rolling of the rolled stock. In this case, the resulting delay time can be set equal to the response time of the measuring devices. If, on the other hand, the measured variables are thicknesses of the rolled stock on the outlet side of the roll stand, in addition to the response time of the measuring devices, a transport time must also be taken into account which elapses for conveying a certain section of the rolled stock from the roll stand to the measuring devices. In this case, the resulting delay time is the sum of the response time of the measuring devices and the transport time.The transport time is essentially determined by the speed of the rolling stock at the exit side of the rolling stand and the distance of the measuring devices from the rolling stand.
[0066] The object is further achieved by a control program having the features of claim 11. According to the invention, the execution of the control program causes the control device to execute an operating method according to the invention.
[0067] The object is further achieved by a control device having the features of claim 12. According to the invention, the control device is programmed with a control program according to the invention, so that the control device executes an operating method according to the invention.
[0068] The object is further achieved by a rolling stand having the features of claim 13. According to the invention, a rolling stand of the type mentioned at the outset is designed in that
[0069] - that the rolling stand has, in addition to the measuring device, a further measuring device by means of which a further measured variable is recorded which depends on an asymmetric eccentricity occurring during rolling of the rolling stock in the rolling stand,
[0070] - wherein the measuring devices are designed to record a drive-side measured variable and an operator-side measured variable,
[0071] - that the further measuring device for supplying the further measured variable is also connected to a control device of the rolling stand and
[0072] - that the control device is designed as a control device according to the invention which controls the rolling stand according to an operating method according to the invention.
[0073] Short description of the drawings
[0074] The above-described properties, features, and advantages of this invention, as well as the manner in which they are achieved, will become clearer and more readily understood in connection with the following description of an embodiment, which is explained in more detail in conjunction with the drawings.
[0075] FIG 1 a rolling arrangement,
[0076] FIG 2 a rolling stand from above,
[0077] FIG 3 a section through a rolling stand,
[0078] FIG 4 is a perspective view of a rolling stand, FIG 5 is a control scheme,
[0079] FIG 6 a flow chart,
[0080] FIG 7 a determination method for measured quantities,
[0081] FIG 8 shows an embodiment of the control scheme of FIG 5,
[0082] FIG 9 a disturbance model and
[0083] FIG 10 a feedback block.
[0084] Description of the embodiments
[0085] According to FIG. 1, a rolling stock 2 is rolled in a rolling stand 1 at a rolling speed v. The rolling stock 2 consists of a metal, for example, steel. According to the illustrations in FIGS. 2 and 3, it is a flat rolling stock. The rolling stand 1 has rolls 3, for example, work rolls and backup rolls. However, the rolling stand 1 can also have more or fewer rolls 3 than the number shown.
[0086] According to FIG. 3, the rolling stand 1 has an operator-side adjustment device 4 and a drive-side adjustment device 5. The adjustment devices 4, 5 are typically designed as hydraulic cylinder units, as shown in FIGS. 3 and 4. The roll gap can be adjusted on the drive side and operator side of the rolling stand 1 using the adjustment devices 4, 5. The rolling stock 2 is rolled in the roll gap.
[0087] According to FIG 1, the rolling stand 1 further comprises measuring devices 6, 7 or 8. By means of the measuring devices 6, 7 or 8, measured variables MDS, MOS are recorded which occur during the rolling of the rolled stock 2. The measured variables MDS, MOS are determined in such a way that they depend on a symmetrical eccentricity occurring during the rolling of the rolled stock 2 in the rolling stand 1 and an asymmetrical eccentricity occurring during the rolling of the rolled stock 2 in the rolling stand 1. Corresponding measured variables MDS, MOS are generally known to those skilled in the art. For example, as shown in FIG 5, the thicknesses of the rolled stock 2 on the outlet side of the rolling stand 1 can be recorded using the measuring devices 6. The thicknesses are recorded at least on the operator side and the drive side, optionally also in the center. Likewise, the respective rolling forces can be recorded using the measuring devices 7. The rolling forces are usually recorded on the operator side and the drive side.Tensions occurring in the rolling stock 2 can also be detected using the measuring devices 8. Tensions can be detected on the inlet or outlet side of the rolling stand 1, with detection on the inlet side being preferred. Tensions are detected at least on the operator side and the drive side, and if necessary, also centrally. Other measurement variables are also conceivable.
[0088] In the following, it is assumed that, as shown in FIG. 5, only one drive-side measured variable MDS and one operator-side measured variable MOS are recorded as measured variables MDS and MOS by the measuring devices 6, 7, or 8. According to FIG. 1, the rolling stand 1 is controlled by a control device 9. The control device 9 is connected to the measuring devices 6, 7, or 8 so that the measuring devices 6, 7, or 8 can feed the recorded measured variables MDS and MOS to the control device 9, and the control device 9 can receive the recorded measured variables MDS and MOS. The control device 9 is also connected to the adjusting devices 4, 5 so that the control device 9 can control the adjusting devices 4, 5.
[0089] The control device 9, as indicated in FIG. 1 by the designation "pP" within the control device 9, is designed as a software-programmable control device. According to FIG. 1, the control device 9 is programmed with a control program 10. The control program 10 comprises machine code 11, which can be directly processed by the control device 9. The programming of the control device 9 with the control program 10 or the processing of the machine code 11 by the control device 9 causes the control device 9 to execute an operating method, which is explained in more detail below—initially in conjunction with FIG. 6.
[0090] According to FIG. 6, the control device 9 receives a symmetrical target roll gap s* in a step S1. The symmetrical target roll gap s* is a uniform value that is symmetrical across the barrel length of the rolls 3 of the rolling stand 1.
[0091] In a step S2, a target roll gap wedge k* is detected by the control device 9. As shown in FIG. 1, it is possible for the target roll gap wedge k* to be explicitly specified to the control device 9, so that the control device 9 receives the target roll gap wedge k*. Alternatively, an implicit specification is possible, for example, if it is assumed that the target roll gap wedge k* should always have the value 0.
[0092] In a step S3, the control device 9 receives the measured variables MDS, MOS acquired by the measuring devices 6, 7, or 8. As already mentioned, it is assumed here that the measuring devices 6, 7, or 8 only acquire a drive-side measured variable MDS and an operator-side measured variable MOS. Therefore, the control device 9 can only receive these measured variables MDS, MOS in step S3.
[0093] In a step S4, the control device 9 determines a symmetrical measured variable MS and an asymmetrical measured variable MA based on the measured variables MDS, MOS. To determine the symmetrical measured variable MS, the control device 9 first determines the sum of the recorded measured variables MDS, MOS according to FIG 7. In some cases, the sum must be divided by 2 to calculate the mean value. In other cases, the sum can be used directly. The two options are indicated by dashed lines in FIG 7. To determine the asymmetrical measured variable MA, the control device 9 determines the difference between the recorded measured variables MDS, MOS according to FIG 7.
[0094] In a step S5, the control device 9 then determines a symmetrical compensation value ös using the symmetrical measured variable MS and an asymmetrical compensation value ök using the asymmetrical measured variable MA. When determining the symmetrical compensation value ös, the asymmetrical measured variable MA is not used. Likewise, when determining the asymmetrical compensation value ök, the symmetrical measured variable MS is not used. Step S5 will be explained in more detail later.
[0095] Furthermore, in a step S6, the control device 9 determines the setpoints pDS, pOS for the adjusting devices 4, 5. The setpoint pOS is the setpoint for the operator-side adjusting device 4, the setpoint pDS is the setpoint for the drive-side adjusting device 5. The control device 9 determines the setpoints pDS, pOS taking into account the setpoint roll gap s*, the setpoint roll gap wedge k*, the symmetrical compensation value ös and the asymmetrical compensation value ök. The setpoints pDS, pOS are generally position setpoints.
[0096] When determining the target values pDS, pOS, the control device 9 according to FIG 6 takes into account the target roll gap s* and the symmetrical compensation value ös with the same sign and the target roll gap wedge k* and the asymmetrical compensation value ök with the opposite sign. Specifically, the control device 9 adds the symmetrical target roll gap s*, the symmetrical compensation value ös, half of the target roll gap wedge k*, and half of the asymmetrical compensation value ök, thus determining the target value pOS for the operator-side adjustment device 4. In a fundamentally analogous manner, the control device 9 adds the symmetrical target roll gap s* and the symmetrical compensation value ös and subtracts half of the target roll gap wedge k* and half of the asymmetrical compensation value ök, thus determining the target value pDS for the drive-side adjustment device 5.The determination is therefore carried out according to the relationships pDS = f(s * +ö + k * l 2 + 5k l 2)' and pOS = f(s * +ö -k * l 2 -8k l 2) .
[0097] In a step S7, the control device 9 controls the adjusting devices 4, 5 according to the determined target values pDS, pOS. The control device 9 then returns to step S3 or (if a new average target roll gap s* or a new target roll gap wedge k* is to be specified) to one of the steps S1 and S2. The control device 9 repeatedly executes (at least) steps S3 to S7 iteratively. The execution takes place with a cycle time in the range of a few milliseconds, for example between 1 ms and 20 ms. The cycle time (in milliseconds) is often a power of 2, i.e., 2 ms, 4 ms, 8 ms, or 16 ms. The cycle time also determines the timing of the procedures explained below in connection with FIGS. 7 to 10.
[0098] To determine the symmetrical compensation value ös, the control device 9 implements an observer 12 according to FIG 5, hereinafter referred to as the first observer. According to FIG 5, the first observer 12 comprises a system model 13 of the rolling stand 1, a number of periodic disturbance models 14 and a state feedback 15. The control device 9 feeds the first observer 12 a speed variable v' characteristic of the rolling speed v, the symmetrical target roll gap s* and the symmetrical measured variable MS. The speed variable v' can, for example, be the rolling speed v itself. Using the first observer 12, the control device 9 determines the symmetrical compensation value ös.
[0099] According to FIG. 5, only a single disturbance model 14 is present. Accordingly, the state feedback 15 also has only a single feedback block 16. However, according to FIG. 8, multiple disturbance models 14 can also be present. In this case, the state feedback 15 has a separate feedback block 16 for each disturbance model 14. The number of four disturbance models 14 shown in FIG. 8 is purely exemplary. More or fewer disturbance models 14 can also be present. The disturbance models 14 can be assigned, as required, to specific rolls 3 or groups of rolls 3 of the rolling stand 1 and to specific fundamental vibrations or harmonics of the fundamental vibrations of the rolls in question.
[0100] According to FIG. 8, the system model 13 comprises three elements 17, 18, and 19 connected in series—this applies regardless of the number of disturbance models 14. The front element 17 takes into account a reaction time TPC of the control devices 4, 5. The front element 17 can be configured, for example, as a PT1 element (generally: as a PTn element). The rear element 19 takes into account a reaction time TM of the measuring devices 6, 7, or 8. The rear element 19 can also be configured, for example, as a PT1 element (generally: as a PTn element).
[0101] The middle element 18 is designed as a multiplier. By means of the middle element 18, a sensitivity V of the symmetrical measured variable MS to the occurring eccentricities is taken into account. The sensitivity V can, if the thicknesses of the rolling stock 2 are measured as measured variables MOS, MDS, to the value where cm and cg are the spring constants of the rolling stock 2 and the rolling stand 1. If the measured variables MOS and MDS are forces, the sensitivity V can be set to the value y _ cm -cg cm + cg. If the measured variables MOS and MDS are tensions, the sensitivity V can be set to a value determined during tests.
[0102] The system model 13 of FIG. 8 is designed as a linear system model. This configuration is preferred, but not mandatory. A non-linear system model can also be used. Furthermore, FIG. 8 shows an integrator within the system model 13. The integrator is used to adjust the operating point of the system model 13. Alternatively, the integrator could be omitted and the DC component of the symmetrical measured variable MS could be filtered out instead. In this case, however, the filter characteristic in the state feedback 15 must be taken into account. Finally, the system model 13 can be expanded by an additional high-pass filter (DT1 element). In this case, the high-pass filter is arranged downstream of element 19.
[0103] FIG 9 shows the preferred structure of a single disturbance model 14. Such disturbance models 14 are known as such and are explained, for example, in EP 0 170 016 B1.
[0104] According to FIG 9, the disturbance model 14 has a front multiplier 20, a front integrator 21, a rear multiplier 22 and a rear integrator 23.
[0105] The signal flow is as shown in FIG 9:
[0106] The control device 9 supplies the output signal z2 of the rear integrator 23 and an angular frequency w to the front multiplier 20. The angular frequency w is specified specifically for the respective disturbance model 14. It is given by
[0107] CO = 2 - 7 - f , where f is an integer multiple of the instantaneous speed of a considered roll 3 or (with the same diameters) of a considered group of rolls 3 of the rolling stand 1. The instantaneous speed of the considered roll 3 or the considered group of rolls 3 is determined by the speed variable v' in conjunction with the diameter of the considered roll 3 or the considered group of rolls 3 of the rolling stand 1. If the rolling speed v changes (and thus also the speed variable v'), the angular frequency w is also adjusted and tracked accordingly.
[0108] The front multiplier 20 multiplies the signals supplied to it, thus forming the product of the output signal z2 of the rear integrator 23 and the angular frequency w.
[0109] Furthermore, the control device 9 feeds the output signal of the front multiplier 20 to the front integrator 21. The front integrator 21 forms the integral of the input signal fed to it and thus determines its output signal z1.
[0110] Likewise, control device 9 feeds the output signal z1 of the front integrator 21 and the angular frequency w to the rear multiplier 22. The rear multiplier 22 multiplies the signals fed to it, thus forming the product of the output signal z1 of the front integrator 21 and the angular frequency w. Furthermore, control device 9 feeds the output signal of the rear multiplier 22 to the rear integrator 23. The rear integrator 23 forms the integral of the input signal fed to it and thus determines its output signal z2.
[0111] The control device 9 feeds the output signal z1 of the front integrator 21 to the system model 13. There, it is added by the control device 9 to the output signal of the first element 17 of the system model 13 (see FIG. 8). If, as shown in FIG. 8, several disturbance models 14 are present, the output signals z1 of the front integrators 21 are added together, as indicated by a summation symbol in FIG. 8.
[0112] According to FIG. 10, the control device 9 further feeds the output signals z1, z2 of both integrators 21, 13 and the angular frequency w of the disturbance model 14 to the feedback block 16 of the state feedback 15. As already mentioned, the respective feedback block 16 is available individually for the respective disturbance model 14. The respective feedback block 16 compensates for the dynamic behavior of the adjusting devices 4, 5 and the other dead times and delay times of the rolling arrangement.
[0113] The feedback block 16 determines, based on the associated disturbance model 14, the component z for the compensation value ös. The feedback block 16 performs this determination using the angular frequency w of the disturbance model 14, the output signals z1, z2 of the integrators 21, 23, the reaction time TPC of the adjusting devices 4, 5, and a resulting delay time TG. The resulting delay time TG can, for example, take into account delays that elapse due to the signal communication from the measuring devices 6, 7, or 8 to the control device 9 or from the control device 9 to the adjusting devices 4, 5, and a cycle time of the control device 9.If thicknesses of the rolling stock 2 are used as measured variables MDS, MOS, the transport time required for the transport of a specific section of the rolling stock 2 from the roll gap of the rolling stand 1 to the measuring devices 6 can also be taken into account within the resulting delay time TG.
[0114] For example, the respective feedback block 16 can, as shown in FIG. 10, first determine a first modified output signal z1' and a second modified output signal z2' based on the relationships zl'= zl + 0 • z2- TPC and z2'= T2 - a> • zl • TPC and then determine the component z based on the relationship z = zl'- cos(® • TG) -z2'- sin(o • TG).
[0115] If only a single disturbance model 14 is present, the symmetrical compensation value ös corresponds to the component z. If multiple disturbance models 14 are present, the symmetrical compensation value ös corresponds to the sum of the components z of the disturbance models 14.
[0116] The procedure for determining and evaluating the symmetrical compensation value ös was explained in detail above. To determine the asymmetrical compensation value ök, the control device 9 implements a further observer 24, hereinafter referred to as the second observer, as shown in FIG. 5. The second observer 24 is structurally designed in the same way as the first observer 12. It also operates in the same way. The only difference is that the control device 9 feeds other variables to the second observer 24, namely the speed variable v', the desired roll gap wedge k*, and the asymmetrical measured variable MA, and that the control device 9 determines the asymmetrical compensation value ök using the second observer 24. Otherwise, the above statements made with regard to the first observer 12 and its components 13 to 23 also apply 1:1 to the second observer 24.
[0117] The present invention offers many advantages. In particular, it compensates not only for symmetrical roll eccentricities, but also for asymmetrical roll eccentricities. Furthermore, retrofitting or expanding an existing symmetrical compensation system with asymmetrical compensation is easily possible. Thicknesses, tensions, or forces can be used as measured variables MOS and MDS as needed. The same structure can be used for cold and hot rolling. Due to the use of observers 12, 24, the determination of the roll eccentricities remains active even during rolling.
[0118] Although the invention has been illustrated and described in detail by the preferred embodiments, the invention is not limited by the disclosed examples and other variations may be derived therefrom by those skilled in the art without departing from the scope of the invention.
[0119] List of reference symbols
[0120] 1 rolling stand
[0121] 2 Rolled goods
[0122] 3 reels
[0123] 4, 5 Adjustment devices
[0124] 6 to 8 measuring devices
[0125] 9 Control device
[0126] 10 Control program
[0127] 11 Machine code
[0128] 12, 24 observers
[0129] 13 Route model
[0130] 14 Disturbance model
[0131] 15 State feedback
[0132] 16 Return block
[0133] 17, 18, 19 limbs
[0134] 20, 22 multipliers
[0135] 21 , 23 Integrators k* Target roll gap wedge
[0136] MA, MS Measured values MDS, MOS Measured values pDS, pOS Set values s* Average set roll gap
[0137] S1 to S7 steps
[0138] TG resulting delay time
[0139] TM response time of the measuring devices
[0140] TPC reaction time of the adjusting device
[0141] V Sensitivity v Rolling speed v' Speed value z1 , z2 Output signals zT, z2' Modified output signals z Share ök Asymmetric compensation value ös Symmetric compensation value
[0142] Cü angular frequency
Claims
Claims 1. Operating method for a control device (9) of a rolling stand (1) in which a flat rolled product (2) made of metal is rolled at a rolling speed (v), - wherein the control device (9) of the rolling stand (1) receives a symmetrical target roll gap (s*), wherein the symmetrical target roll gap (s*) is a uniform value which is symmetrical over the barrel length of the rolls (3) of the rolling stand (1), - wherein the control device (9) receives measured variables (MOS, MDS) detected by means of measuring devices (6 to 8) and dependent on a symmetrical eccentricity occurring in the rolling stand (1) during rolling of the rolling stock (2) and an asymmetrical eccentricity occurring in the rolling stand (1) during rolling of the rolling stock (2), - wherein the received measured variables (MOS, MDS) comprise a drive-side and an operator-side measured variable (MDS, MOS), and the control device (9) determines the symmetrical measured variable (MS) by summing or averaging the received measured variables (MOS, MDS) and the asymmetrical measured variable (MA) by subtracting the received measured variables (MOS, MDS), - wherein the control device (9) for an operator-side and a drive-side adjustment device (4, 5) of the roll stand (1), by means of which the roll gap is adjusted on the drive side and operator side of the roll stand (1), determines respective target values (pOS, PDS) taking into account the target roll gap (s*), a target roll gap wedge (k*), a symmetrical compensation value (ös) and an asymmetrical compensation value (ök) and controls the adjustment devices (4, 5) accordingly, - where the target roll gap wedge (k*) is given implicitly with the value 0 or explicitly, - wherein the control device (9) determines the symmetrical compensation value (ös) using the symmetrical measured variable (MS) and the asymmetrical compensation value (ök) using the asymmetrical measured variable (MA), - wherein the control device (9) takes into account the desired roll gap (s*) and the symmetrical compensation value (ös) with the same sign and the desired roll gap wedge (k*) and the asymmetrical compensation value (ök) with the opposite sign when determining the desired values (pOS, pDS) for the adjusting devices (4, 5).
2. Operating method according to claim 1, characterized in that the control device (9) uses the symmetrical measured variable (MS) but not the asymmetrical measured variable (MA) to determine the symmetrical compensation value (ös) and uses the asymmetrical measured variable (MA) but not the symmetrical measured variable (MS) to determine the asymmetrical compensation value (ök).
3. Operating method according to claim 1 or 2, characterized in that that the control device (9) determines the setpoint value (pOS) for the operator-side adjustment device (4) by adding the symmetrical setpoint roll gap (s*), the symmetrical compensation value (ös), half of the setpoint roll gap wedge (k*) and half of the asymmetrical compensation value (ök) and determines the setpoint value (pDS) for the drive-side adjustment device (5) by adding the symmetrical setpoint roll gap (s*) and the symmetrical compensation value (ös) and subtracting half of the setpoint roll gap wedge (k*) and half of the asymmetrical compensation value (ök).
4. Operating method according to claim 1, 2 or 3, characterized in that - that the control device (9) implements a first and a second observer (12, 24), each comprising a line model (13) of the rolling stand (1), a number of periodic disturbance models (14) and a state feedback (15), - that the control device (9) supplies the first observer (12) with a speed variable (v') characteristic of the rolling speed (v), the symmetrical nominal roll gap (s*) and the symmetrical measured variable (MS) and supplies the second observer (24) with the speed variable (v'), the nominal roll gap wedge (k*) and the asymmetrical measured variable (MA) and - that the control device (9) determines the symmetrical compensation value (ös) by means of the first observer (12) and the asymmetrical compensation value (ök) by means of the second observer (24).
5. Operating method according to claim 4, characterized in that - that the respective system model (13) comprises three elements (17 to 19) connected in series, - that a reaction time (TPC) of the adjusting devices (4, 5) is taken into account by means of the frontmost link (17), - that by means of the middle element (18) a sensitivity (V) of the measured variable (MS, MA) supplied to the respective system model (13) to the occurring eccentricities is taken into account and - that a reaction time (TM) of the measuring devices (6 to 8) is taken into account by means of the rear element (19).
6. Operating method according to claim 7, characterized in that the respective system model (13) additionally comprises a further element which is arranged downstream of the rear element (19) and is designed as a high-pass filter, in particular as a DT1 element.
7. Operating method according to claim 4, 5 or 6, characterized in that - that the respective disturbance model (14) has a front multiplier (20), a front integrator (21), a rear multiplier (22) and a rear integrator (23), - that the control device (9) supplies the front multiplier (20) with the output signal (z2) of the rear integrator (23) and an angular frequency (w) of the respective disturbance model (14), - that the control device (9) supplies the output signal of the front multiplier (20) to the front integrator (21), - that the control device (9) supplies the output signal (z1) of the front integrator (21) and the angular frequency (w) of the respective disturbance model (14) to the rear multiplier (22), - that the control device (9) supplies the output signal of the rear multiplier (22) to the rear integrator (23), - that the control device (9) feeds the output signal of the front integrator (z1) to the respective system model (13) and - that the control device (9) feeds the output signals (z1, z2) of both integrators (21, 23) and the angular frequency (w) of the respective disturbance model (14) to a feedback block (16) of the state feedback (15) assigned to the respective disturbance model (14).
8. Operating method according to claim 7, characterized in that the control device (9) adds the output signal (z1) of the front integrator (21) to the output signal of the first element (17) of the respective route model (13).
9. Operating method according to claim 7 or 8, characterized in that the feedback block (16) assigned to the respective disturbance model (14) determines a respective component (z) for the respective compensation value (ös, ök) by utilizing the angular frequency (w) of the respective disturbance model (14), the output signal (z1) of the front integrator (21) of the respective disturbance model (14), the output signal (z2) of the rear integrator (23) of the respective disturbance model (14), the reaction time (TPC) of the adjusting devices (4, 5) and a resulting delay time (TG).
10. Operating method according to claim 9, characterized in that - that the feedback block (16) assigned to the respective disturbance model (14) is based on the relationships zl'= zl + o • z2- TPC and z2'= v - co • zl • TPC a first and second modified output signal (z1', z2') is determined, where z1' is the first modified output signal, z1 is the output signal of the front integrator (21), w is the respective angular frequency, z2 is the output signal of the rear integrator (23), TPC is the reaction time of the adjusting devices (4, 5) and z2' is the second modified output signal, and - that the feedback block (16) assigned to the respective disturbance model (14) determines the respective component (z) for the respective compensation value (ös, ök) on the basis of the relationship z = zl'- cos(® • TG) -z2'- sin(o • ZG), where z is the respective component for the respective compensation value (ös, ök) and TG is the resulting delay time.
11. Control program for a software-programmable control device (9) for a rolling stand (1), wherein the control program comprises machine code (11) which can be directly processed by the control device (9), wherein the processing of the machine code (11) by the control device (9) causes the control device (9) to carry out an operating method according to one of the above claims.
12. Control device for a rolling stand (1) in which a flat rolled stock (2) made of metal is rolled at a rolling speed (v), wherein the control device is programmed with a control program (10) according to claim 11, so that the control device carries out an operating method according to one of claims 1 to 10.
13. Rolling stand in which a flat rolled product (2) made of metal is rolled at a rolling speed (v), - wherein the rolling stand has an operator-side and a drive-side adjusting device (4, 5), by means of which the roll gap in which the rolling stock (2) is rolled can be adjusted on the drive side and operator side of the rolling stand, - wherein the rolling stand comprises measuring devices (6 to 8) by means of which measured variables (MOS, MDS) are recorded which depend on a symmetrical eccentricity occurring in the rolling stand during rolling of the rolling stock (2) and an asymmetrical eccentricity occurring in the rolling stand during rolling of the rolling stock (2), - wherein the measuring devices (6 to 8) are designed to detect a drive-side measured variable (MDS) and an operator-side measured variable (MOS), - wherein the measuring devices (6 to 8) for supplying the measured variables (MOS, MDS) are connected to a control device (9) of the rolling stand and the control device (9) for controlling the adjusting devices (4, 5) is connected to the adjusting devices (4, 5), - wherein the control device (9) is designed as a control device according to claim 12, which controls the rolling stand according to an operating method according to one of claims 1 to 10.