EP2603332A1

EP2603332A1 - Method for determining control variables of a rolling train comprising a plurality of roll stands for rolling a metal strip

Info

Publication number: EP2603332A1
Application number: EP11763870.0A
Authority: EP
Inventors: Matthias Kurz; Klaus Weinzierl
Original assignee: Siemens AG
Current assignee: Primetals Technologies Germany GmbH
Priority date: 2010-09-16
Filing date: 2011-09-02
Publication date: 2013-06-19
Also published as: WO2012034875A2; CN103402661A; CN103402661B; EP2431105A1

Description

description

Determination method for control variables of a rolling mill with several rolling mills for rolling a metal strip

The present invention relates to a determination method for control variables of a rolling mill having a plurality of rolling mills for rolling a metal strip. The present invention further relates to a Computerpro ^¬ program comprising machine code that can be executed directly by a computer and whose execution caused by the computer that the computer executes such a discovery ^{information model.}

The present invention further relates to a computer which is designed such that it carries out such Ermitt ^¬ treatment method. Furthermore, the present invention relates to a rolling mill with several rolling mills for rolling metal strip, which is equipped with such a computer.

In rolling mills - for example in the finishing train of a hot strip mill - mainly the band geometry, ie Di ^¬ bridge, profile and flatness, set. The rolling train consists of at least one scaffold, usually of several scaffolding. Usually, the rolling mill is provided with two automation ^¬ sierungsteilen:

In the basic automation is the technological control, which works in real time and usually with Abtastra ^¬ th in the range of a few milliseconds allows the processing of very fast control operations. Due to the low computation power and the short sampling time, models used there are generally so simplified compared to reality that only local statements are possible. In the basic automation ^¬ tion, the strip thickness control is executed. You rea- yaws mainly in retrospect, that is used mainly kei ^¬ nerlei knowledge, the band-temperature measurements were available, for example because of pre ^¬. One exception is the FeedForward AGC, which is at least partially piloted.

In the higher-level process automation, the setup for the tape head is calculated. Furthermore, the band-to-band adaptation is carried out there. The programmed there models are nonlinear, provide information on the global behavior of tape and rolling mill and are more complex than the models of the basic automation usually a vielfa ^¬ ches. It is also common practice to derive sensitivities from process automation that can be used to build local models in basic automation. In the prior art, it is not common to run the process automation models in real time, except, of course, that a tape head setup is calculated before the tape head enters the mill train. Process automation initiali ^¬ thus Siert the thickness control of the basic automation on the tape head. Process automation are readings of game, discloses the pre-strip at ^¬ with which an anticipatory thickness control would be possible.

In exceptional cases, it is also calculated in real time in the process computer, for example in the regulation of the final rolling temperature or in the run-in correction. In the latter information is used to rolling force in already traversed by the tape head scaffolding to perform more quickly, a correction for the next Ge ^¬ Rüst before this is achieved. Even during the recalculation, it may happen that parts of the invoices are done in real time. However, the recalculation is basically not used to control the current band but to provide model corrections for subsequent bands.

The aforementioned separation of tasks does not allow in the prior art, complex rules in real time under consideration to execute the entire system status. In particular, it is not possible by means of the strip thickness adjustment of all Ge ^¬ equip it with a predictive control algorithm to variie ^¬ reindeer, the total takes into account the state of the plant. The problem is that the transport of a material section through the rolling train typically lasts 20s, in the Ver ^¬ use of the entire system state in the Basisautomati ^¬ tion with the usual there very small clock cycle but the complexity of the scheme is so large that they no longer in Real time can be edited. If, for example, one wanted to control the positions of all scaffolds with a model-predictive controller, one would have to take into account the entire condition of the road, generate a prognosis at each time interval and select the prediction horizon at least as large as the throughput time of the material through the rolling train.

In the prior art, the basic automation for the control, the process computer for the setup on the tape head and the adaptation is responsible. The process computer usually uses band segments. These are software representations of sections of rolled material of given length. A regulation in the prior art in the process computer only with respect to the final rolling temperature. Here, a model predictive temperature control is known in the art, see for example DE 103 21 791 AI.

The basic automation essentially takes over the calculation of the process computer for the tape head and then executes a control in real time along the tape. Vorzugswei- is se first carried out a correction of the calculated exit thickness with a so-called AGC, by the gemes ^¬ sene rolling force with the expected rolling force is compared and is carried Hooke than the spring constant of the framework according to the (linearized overbased) law correction of the exit thickness. For example, if the measured rolling force larger than expected, the framework is more pressed than expect ^¬ tet, ie the thickness must be calculated exit ^¬ yaws upward Corridor. Since you do not want that, by reducing the setting of the framework of the roll gap immediately reduced by a corresponding amount, so that again the desired exit thickness arises. In addition, the basic automation is responsible for the adjustment of given trains and positions of the

Ski lifter (Looper). For this purpose, the trains are set with the loopers and the engine speeds of the scaffold engines constantly adjusted so that the Looper come to rest in a certain position.

The thickness control is also in the basic automation. Once a measured tape thickness is available (i.e., the tape head has exited the road), one can compare the measured thickness to the expected tape thickness and apply the zero point correction error to the zero point of the job. Since usually only behind the last framework the thickness is measured, the error must be distributed over several scaffolding usually. A relatively simple control corrects the settings of the scaffolds starting from the last scaffolding and, if necessary, also suitably adapts the scaffold speeds.

The prior art is still a Smith predictor for the thickness control. Here, starting from the current thickness deviation, the first thing that is deducted is what was activated earlier. The previously switched values are taken from a memory. The adjusted deviation is fed to a moving averaging, bounded, once again slightly smoothed and placed directly on the last stand. Upstream scaffolds also receive an intrusion, but attenuated with a respective time constant corresponding to the duration of a point from the respective band Ge ^¬ Rüst until the thickness gauge. Thus, upstream scaffolds receive a corresponding correction.

The last frame is usually the guide for the belt speed. Interference in the employment of scaffold The effect of a change in thickness is that the speeds of the upstream stands must also be changed. Of course, in the prior art, the speed of the scaffold motors of upstream stands is suitably precontrolled while the strip thickness is changed.

In heavy plate mills, a method is known in which

- points are recorded along the sheet before the sheet enters the rolling stand,

calculating expected rolling forces and exit thicknesses for all points by means of a rolling force model,

the expected exit thicknesses are compared with the target thicknesses assigned to the points,

calculate scaffold positions assigned to the comparisons for each point,

- The scaffolding adjustments are connected before the shrinkage of the sheet to a curve such that the Anstel ^¬ ment is given as a function of sheet metal progress, and - the sheet is rolled in accordance with the predetermined adjustment curve of Ge ^¬ equipment.

Thus, with thick plate a course of the thickness can be adjusted along the sheet.

Heavy plate is, seen in the direction of the plate, relatively short. The plate is therefore divided in the prior art in a relatively small number of points, for example, seven Punk ^¬ te. Due to the small number of points, it is possible for heavy plate to execute the entire calculation before entering the heavy plate into the framework. Furthermore, there are no situations which simultaneously detects more than one scaffold, the heavy plate, and thereby couples the production process at rough ^¬ sheet.

The method last described is now also applied to Steckel rolling mills. Steckel mills have - also as heavy plate mills - usually only a single rolling stand on.

The object of the present invention is to provide possibilities by means of which as far as possible a forward-looking control of the strip thickness in the rolling mill ^¬ road is possible, with known disturbances and plant limits can be considered.

The object is achieved by a preliminary investigation with the features of claim 1. Advantageous embodiments of the determination method according to the invention are the subject of the dependent claims 2 to 14.

According to the invention, it is provided to design a determination method for control variables of a rolling train with a plurality of rolling stands for rolling a metal strip,

- that prior to the arrival of band points of the metal strip in a front stand of the rolling mill in each case initial parameters of the respective band point are detected, which enters the respective band point in the front rolling mill, and the respective band point is taken over in the active Auswer ^¬ tion,

- that at least one expected in the front roll stand during the rolling of each strip point frame size is determined by means of a rolling force model for the front Walzge ^¬ Rüst,

- that further downstream for the prede ^¬ ren rear roll stand rolling mills each parameter of the respective tape point are determined by means of the rolling-force model, to which the respective strip point enters into the respective rear roll stand,

- in that further at least an expected in the respective rear roll stand during the rolling of each strip point ^¬ weiligen scaffold size is determined by means of the rolling force model for the rear ^¬ ren roll stands, - that the expected frame sizes of the rolling stands with jewei ^¬ time assigned to the respective band set point scaffold quantities are compared, and for the front roll stand and the rear roll stands by comparing each min- least a respective Einstellgerüstgröße is determined,

- That at least the determined Einstellgerüstgrößen are assigned to the respective band point before the arrival of the respective band point in the front rolling stand and tracked with the respective band point through the rolling mill,

that trajectories for the setting frame sizes can be calculated on the basis of the adjusting frame sizes assigned to the band points, at least for the rear rolling stands,

- that the trajectories for each rear rolling mill are each at least from the time at which the respective

Runs tape point in the front roll stand, extend to the time at which the respective band point enters the respective rear roll stand. As Einstellgerüstgrößen usually the rolling force or the roll stand position are used. In principle, other Einstellgerüstgrößen are sorted ^¬ but used, either in addition, whether an alternative to the above sizes. Examples of other Einstellgerüstgrößen are the rolling moment, the trains (inlet and outlet side), the back bending and even ^¬ TULLE an intermediate stand cooling.

The parameters of the band points can be temperature parameters, for example the temperature or the enthalpy, if appropriate including phase components. Alternatively or additionally, the parameters of the band points can be geometry parameters such as, for example, the strip thickness, the profile and the flatness. As a scaffold sizes particular outlet side Geometry ^¬ rieparameter and the rolling force come into question. The Einstellgerüstgrößen are the approach control variables of the rolling mill. The term Einstellgerüstgrößen is narrower than the term control variables, since the ^{Einstellgerüst ¬} sizes are always based on a particular of the rolling stands.

It is possible that the trajectories can also be determined for the front rolling mill, provided that the determination of the setting stand sizes is completed in good time prior to the arrival of the respective hinge point in the front roll stand. But it is sufficient if the trajectories of setting ^¬ scaffolding sizes are calculated only for the rear roll stands.

Furthermore, it is sufficient in principle if the calculability exists. An actual computing is not zwin ^¬ quietly required.

Already by the inventive procedure, it is game as possible ^¬ to expect in the process computer not only for the tape head, but to consider the entire band. As a result, holistic strategies can be applied, such as minimizing energy consumption or maximizing throughput. Also, material can be produced with temporally variab ^¬ len setpoints, z. B. material with a thickness wedge.

As a rule, however, the trajectories are calculated. In this case, there are two ways to use the trajectories.

Firstly, it is possible that the trajectories are at least partially transmitted to a basic automation, so that they are the basic control as the course of desired regulated ^¬ sizes for the rear roll stands are available. DA by the controller of the basic automation can regulate in advance and, for example, setpoint changes better vorsteu ^¬ ren. As a rule, the trajectories are calculated with a modeling period. Preferably, the trajectories are transmitted to ^¬ least for a control period of the basic automation, which is greater than the modeling period. DA by a continuous control by the base car ^¬ automation also possible if the parent process computer executing investigation according to the invention, due to special circumstances comes out of step (the process computer, so to speak "hat times").

On the other hand, it is possible for the trajectories to be evaluated before triggering the rolling stand for which the respective trajectory is valid and for it to be decided as a function of the evaluation whether and, if so, how the setting frame sizes are varied. As a result, for example, loads which only arise from the synopsis of the rolling stands can be determined and evaluated. An example of such a load is the total power requirement of the mill train. Also can be detected ahead if performance limits of the system - for example, the maximum possible torque of a rolling mill drive - be ^¬ reached. Furthermore, it is also possible to determine loads which only arise from the time course of the trajectories, for example the thermal loading of the rolling stand drives.

As already mentioned, the trajectories are usually calculated with a modeling period. Takes place, however, the evaluation of Tra ^¬ trajectories with an evaluation horizon that is a multiple of the modeling period. By "multiples" is not necessarily an integral multiple ge ^¬ means, but a considerably larger value In particular, the ratio of evaluation horizon Modellierungsperio ^¬ de usually at least 10. 1,.

In the context of the present invention, it is sufficient if the trajectories of the setting frame sizes can be determined. Preferably, however, it is provided that the respective Bandpunkt also the expected framework sizes are assigned and that also the trajectories for the expected framework sizes can be calculated. Preferably, the determination method according to the invention is carried out cyclically and in real time. This is particularly desirable for the period during which at least ei ^¬ nes of the roll stands of the rolling train is engaged, that is, when at least one point of the metal strip is rolled.

Preferably, determination rules by which the expected framework sizes, the parameters of the band points and the Einstellgerüstgrößen are determined, continuous maps. This results in - at least as a rule - relatively smooth tra ectories.

It is preferably provided that at least a definitive ^¬ term parameters of the respective tape point is detected after the expiry of the respective tape point from the rolling mill that the rolling-force model based on the detected final Parame ^¬ ters and behind the last roll stand of the rolling mill he ^¬ waited parameter in real time is adapted and that the expected parameters of those band points that have already run into the rolling mill, be tracked in real time.

The corresponding procedure is known for the parameter "Tem ^¬ temperature" from DE 101 56 008 Al. The parameter "geometry" is the corresponding procedure in the patent tentanmeldung "real-time detection method of temperature and geometry of a metal hot-rolled strip in a Fertigstrasse ", which is filed at the same time as this patent application, which patent application bears the applicant's internal reference 201015307.

Preferably it is further provided that the erfindungsge ^¬ Permitted investigation by the adaptation of the rolling-force model and the tracking of the expected parameters for those those strip points whose parameters were tracked, is carried out again and that for the renewed execution of the investigation process that rolling mill is considered as a front rolling ^¬ scaffold, in which the respective band point whose parameters were tracked next enters. This ensures that the trajectories are continuously tracked and updated, even with respect to those band points that have already run into the rolling mill. Preferably, the time of detection of the initial parameters of the respective tape guiding principle a point schwindigkeitsverlauf known, with which the respective band ^¬ point to pass through the rolling line at the latest. One way to determine a Leitgeschwindigkeitsverlaufs is described in particular in the older, on the filing date of the present invention not yet disclosed European Patent Application 10162135.7 from 06.05.2010.

The object is further achieved by a computer program of the type mentioned. In this case, the Ab ^¬ processing causes the machine code by the computer, that the computer executes an inventive investigation.

The object is further achieved by a computer which is configured such that it performs an inventive Determined ^¬ averaging method.

The object is further achieved by a rolling mill with several rolling stands for rolling metal strip, which is equipped with a computer according to the invention.

Further advantages and details will become apparent from the following description of exemplary embodiments in conjunction with the drawings. In a schematic representation:

1 shows schematically a rolling train,

2 shows schematically a metal band,

3 shows a flow chart, 4 tractors,

5 shows a flow chart,

6 and 7 each have a trajectory,

8 and 9 are diagrams and

10 to 12 flowcharts.

According to FIG. 1, a rolling train has a plurality of rolling stands 1. Shown in FIG 1 are four such rolling stands 1. This number is usually the minimum. Often even five to eight rolling stands 1 are present, for example six or seven rolling stands 1. In the rolling train, a metal strip 2 is rolled. The metal strip 2 may alternatively be cold rolled in a roughing or finishing train or cold rolled in a tandem mill. The metal strip 2 may for example consist of steel, aluminum or copper.

The rolling train is equipped with a computer 3. The computer 3 is programmed with a computer program 4. The computer program 4 has been supplied to the computer 3, for example via a conventional storage medium 5, for example a USB memory stick. Due to the programming with the computer program 4, the computer 3 is designed such that it carries out a determination process, which will be explained in more detail below.

The computer program 4 comprises machine code 6. The machine code 6 can be processed directly by the computer 3. The processing of the machine code 6 by the computer 3 effects the execution of the determination method according to the invention.

The behavior of the metal strip 2 in the rolling train is modeled in the computer 3 by means of a rolling force model 7. ^Appropriate models are well known to ^those skilled in the art. For the modeling of the band 2, the band 2 of Figure 2 is represented in ^¬ nerhalb of the computer 3 by a plurality of band-times 8. Each relating to a band of items 8 getrof ^¬ fene statement is self-evident for a corresponding Length range of the metal strip 2, wherein dividing lines between the band points 8 are usually in the middle between the band points 8. The number of band points 8 is usually several 100 to more than 1000 band points 8. A distance a of the band points 8 from each other can be defined for example by a fixed timing T. The time clock T is in this case usually at 100 ... 500 ms, for example at 150 ... 300 ms. Alternatively, the distance a can be defined by a geometric distance of, for example, 20... 100 cm. Preferably, the distance a between the band points 8 is defined at ^¬ metal ribbon 5 by the located between in each case two belt points 8 mass. In this case, the distance a can be, for example, between 10 and 50 kg, in particular between 15 and 30 kg.

As already mentioned, the metal band 2 is divided into a plurality of band points 8 with respect to its representation in the computer 3. The corresponding sections of the real

Metal strip 2 run successively into a front rolling stand 1 of the rolling train, for example in the first rolling stand 1 of the rolling train. Even before the running in of the sections of the real metal strip 2, the respectively corresponding strip points 8 are taken over into the active evaluation.

Applying to the active evaluation is connected to a detection of initial parameters P of the corresponding band points 8. As a rule, the initial parameters P are recorded one after the other for one of the band points 8 in each case. However, it is alternatively possible, capturing the initial parameter P - take groups vorzu ^¬ - possibly including the acquisition of the corresponding points on the strip 8 in the active evaluation. It is even possible to carry out the detection of the initial parameters P for all band points 8 at once. The

However, via acquisition of the corresponding points on the strip 8 in the active From ^¬ evaluation individually or in groups, not for all the strip points 8 of the metal strip 2 at a time. It is possible to carry out the detection of the parameters P metrologically. Alternatively, another detection is possible. By way of example, the parameters P can be supplied to the computer 3 by a superordinate computing device or by another control computer which controls a system arranged upstream of the rolling train. Examples of such pre ^¬ arranged plants are an oven, a roughing and a

Continuous casting plant.

The parameter P, for example, the temperature of the band corresponding to point 8, a phase fraction or contained ^¬ pie may be. Also combinations of these are possible. It is also possible that the parameter P describes the geometry of the respective band point 8, for example its thickness, its profile or its flatness. A combination of several geometry parameters is possible. It is also possible to record both temperature parameters and geometry parameters.

In the case of a temperature, a single value may be given, for example the value immediately at the center of the band segment defined by the band point 8. Example ^¬ example, the temperature at the top or bottom or - as seen in the strip thickness direction - are used in the center of the corresponding location. Alternatively, it may be a temperature distribution. The temperature distribution can be spatially resolved in this case in the band thickness direction, in the band width direction and / or possibly also in the band longitudinal direction. Averaging in one or more of the named directions is also possible. Also combinations are possible. For example, the temperature at the top can be used spatially resolved in the bandwidth direction. Wei ^¬ terhin can be used instead of the temperature and the enthalpy. Possibly. Furthermore, phase proportions and parameters for material properties, such as, for example, the grain size and / or the solidification, can be taken into account.

In an analogous manner, even in the case of a geometry, a single value may be given, for example, the strip thickness directly telbar at the center of the defined by the band point 8 band segment. Alternatively, it may be a Geometrievertei ^¬ development. The geometry distribution can be spatially resolved in the case of the band thickness in the bandwidth direction and possibly also in the band longitudinal direction. In the case of the band profile and / or the band flatness, the geometry distribution may possibly be spatially resolved in the band longitudinal direction. Analogous to the temperature distribution, an averaging in the band longitudinal direction, in the case of the band thickness alternatively or additionally also in the band width direction, is possible.

The band points 8 are taken over by the computer 4 in a step Sl in the active evaluation. If necessary, the computer 3 also accepts target frame sizes G * in step S 1 for the corresponding strip points 8.

The target frame sizes G * are sizes that should occur during rolling of the corresponding band point 8 in the rolling stands 1. They are broken down for each Walzgerüs ^¬ te 1 defined. It may be * act at the target frame sizes G to manipulated variables of the respective roll stand 1, such as the roll force F, the appointing s or the rolling ^¬ moment M. Preferably, it is the geometry parameters such as the outlet-side target thickness, the outlet-side target profile and / or the outlet side Sollplanheit. It may also be relative sizes, for example, a certain proportion of the total, applied by all rolling mill drives power. In a step S2, the computer 3 assigns the initial parameters P and the set values G * to the corresponding band points 8.

Preferably, there is a step S3, in which the computer 3 determines or accepts a characteristic course for those band points 8 which it has taken over into the evaluation in step S1. The Leitgrößenverlauf indicates with wel ^¬ chem speed profile (for example as a function of time) the band 8 points should pass through the rolling mill. A suitable method of determining the course of the process variable is For example, in the aforementioned European Patent ^¬ application 10162135.7 from 06.05.2010 described.

The Leitgrößenverlauf - of course - always on the entire metal band 2. Furthermore, the Leitgrößenverlauf for band points 8, which were already taken into the evaluation, already determined for each time. Therefore, in step S3, only the addition of the Leitgrößenverlaufs is answered ^¬ or determined, ie the Leitgrößenverlauf for the period in which the respective strip points of the step Sl 8 leaving the rolling train.

In a step S4, the computer 3 selects the front rolling mill 1, for example the first rolling mill 1 of the rolling mill. In a step S5, the computer 3 determines for each band point 8 of the step Sl for the selected stand 1 at least one expected stand size G, which is expected during rolling of the corresponding band point 8 in the selected stand 1. The determination takes place by means of the rolling force model 7.

As possible framework sizes G, the computer 3 can determine in particular at least one of the following variables: the rolling force F, the inlet-soaking train in the metal strip 2, the run-out train in the metal strip 2, the stand position s, the frame speed or the roller peripheral speed, the rolling moment M, the Adjustment of a loop lifter arranged on the inlet or outlet side, nominal values for coolant actuating elements acting on the rolls of the respective rolling stand 1 and / or the metal strip 2, auslaufseifige geometry pa ^¬ parameters of the corresponding band points 8, temperature-related parameters of the corresponding band points 8 during leakage, etc. The computer 3 assigns the determined quantities to the corresponding band points 8 in a step S6.

In a step S7, the computer 3 checks whether the selected rolling stand 1 is the last rolling stand 1 of the rolling train. If this is not the case, the computer 3 goes in a step S8, in which he selects the next rolling stand 1.

In a step S9, the computer 3 determines for the corresponding band points 8 by means of the rolling force model 7 respectively parameters P of the band points 8, with which the respective band points 8 einlau ^¬ fen in the now selected mill 1. As far as the geometry parameters of the band points 8 are concerned, the step S9 is trivial, since the outlet geometry, with which a certain band point 8 runs out of the previously selected rolling stand 1, is identical to the inlet geometry with which this band point 8 enters the newly selected rolling stand 1 , As far as temperature parameters are concerned, for example, based on the distance of the rolling stands 1 vonein ^¬ other in conjunction with the Leitgeschwindigkeitsverlauf (or a fixed predetermined speed) the Temperaturent ^¬ winding the band points 8 are updated. If necessary, cooling by means of a corresponding intermediate cooling device can be taken into account.

The quantities determined in step S9 are also assigned to the respective band points 8, see a step S10 in FIG. 3. From step S10, the computer 3 returns to step S5.

Due to the nature of the equations of the rolling-force model 7 is guaranteed in most cases that the determination of rules, by means of which the expected frame sizes G and the parame- ters P of the band points are determined 8, are continuous Abbil ^¬ applications. Thus, if the input variables of the determination regulations change slightly, there are no abrupt changes in the determined output variables. If the computer 3 determines in step S7 that it has arrived at the last rolling stand 1, it proceeds to a step Sil. In step S, the computer 3 compares the expected framework sizes G determined by it with the setpoint size. ßen G * of the step Sl. Furthermore, the computer 3 determines in step Sil for each band point 8 of the step Sl on the basis of the comparison in each case at least one Einstellgerüstgröße A *. Usually - though not necessarily - determined the accounting ner 3 in step Sil for each rolling stand 1 a Sollanstel ^¬ lung s * or set rolling force F *.

Regardless of which setting frame sizes A * the computer 3 determines, the setting frame sizes A * are variables different from the expected and the set frame sizes G, G *. For example, if the rolling force as he waited ^¬ and target frame size G, G * is used, the rolling force can not be used as Einstellgerüstgröße A * at the same time.

The determined Einstellgerüstgrößen A * assigns the computer 3 in a step S12 the respective band point 8. In a step S13, the computer 3 adds the band points 8 to which it has assigned the setting frame sizes A * in step S12 to the totality of the band points 8 managed by it. The entirety of the managed by the computer 3 band points 8 reprä ^¬ sent those sections of the metal strip 2, which were ^¬ already taken in the active evaluation and have not yet leaked out of the mill. The step S13 and with it, the steps Sl to S12 will therefore be ^¬ leads before the corresponding points on the strip 8 enter into the front roll stand. 1

Analogously to the determination of the gantry sizes G and the parameters P of the belt points 8, as a rule, the determination rules by means of which the adjustment gantry sizes A * are determined are also continuous maps.

Advantageously, a target function is set up in the calculations into which at least the deviation between an expected stand size G behind the respective rolling stand 1 and subsequent rolling stands 1, for example an expected thickness of the associated set stand size, is received. equation constraints formulated for the plant limits. By minimizing the objective function, the mathematical minimization algorithm, for example. For example, a SQP method or a Gauss-Newton method, the required settings for the rolling stands 1 are calculated.

In a step S14, the computer 3 determines the trajectories and evaluates them. In a step S15, the computer 3 performs path tracking with respect to all band points 8 of the totality of band points 8. The computer 3 thus determined for each managed band 8 point where the respective band ^¬ point 8 is currently located.

In a step S16, the computer 3 is at least the ermit telten ^¬ and the tape points 8 * associated time and location correctly to a base automation 9 from Einstellgerüstgrößen A. The base automation 9 receives the Einstellgerüstgrö ^¬ Shen A * and controls using the adjustment frame sizes A * the rolling mill. The basic automation 9 can use the values transmitted from the computer 3 as pilot control and possibly even execute a local regulation that deviate ^¬ tions minimized by the given trajectories. Such a method is known. This prediction can be used at ^¬ play as for determining the energy consumption, to Progno ^¬ se of future electricity consumption for pilot control of other control circuits, for the calculation of engine loads and so on. In particular, such calculations can lead to a correction of the plant limits (for example a reduction of the engine power in 5s due to thermal overload, which is only permissible for a short time), which can be included in the next time clock of the computer 3. If appropriate, the computer 3 can additionally output the expected framework sizes G to the basic automation 9 in step S16. In this case, the basic automation 9 can regulate the rolling train with additional consideration of the expected gantry sizes G. In a step S17, the computer 3 checks whether just a band ^¬ point 8 from leaking out of the rolling line. If this is the case, the computer 3 removes the corresponding band point 8 from the entirety of the band points 8 it manages in a step S 18.

In a step S19, the computer 3 checks whether straight (Minim ^¬ least) a tape point to be re-taken into the evaluation. 8 Depending on the result of the examination of step S19, the computer 3 returns to step S1 or to step S14.

The steps S17 and S19 are always required, ie not only when the distance a of the band points 8 from each other is determined as a geometric distance or as between the band points 8 mass. The steps S17 and S19 are also required if the distance a of the band points 8 from each other is determined by the timing T, the method of FIG 3 is thus executed not only cyclically, but even clocked. For if the first band points 8, so the tape head, run into the rolling mill 8, yet no tape points 8 run out of the rolling mill. Conversely, the last strip points 8, ie the strip foot, only run out of the rolling mill long after the last strip points 8 have entered the rolling train.

Due to the fact that in the steps Sl to S12 the Einstellgerüstgrößen A * of the respective band points 8 are determined for all rolling stands 1 and the Einstellgerüst ^¬ sizes A * for those band points 8, which pass through the rolling mill before the newly added band points 8 due of the operating principle of FIG. 3 have already been determined previously, trajectories for the setting frame sizes A * (ie the future course of the setting frame sizes A *, at least for the rear rolling stands 1 - possibly also for the front rolling stand 1 ) calculable. If a guide velocity or a corresponding course is given, the trajectories can be determined as a function of the time t. Because in this case can be determined on the basis of Leitge ^¬ speed or the corresponding course, when which band point 8 is rolled in which rolling mill 1. Even if the conduction velocity is not given, the trajectories as a function of the mass flow are mediate ^¬ mediate. The mass flow m 'can be given for each roll stand 1, for example, by the strip mass that has run past the respective roll stand 1. Optionally, a roll stand-specific offset can be taken into account in each roll stand 1 in order to standardize the abscissa of the individual trajectories relative to one another. 4 shows purely by way of example the set rolling force F * at the second and the rolling moment M at the fourth rolling stand 1 of the rolling train as possible adjusting stand sizes A *. FIG. 4 furthermore shows that the abscissa can alternatively be the time t or the mass flow m '-. From FIG 4 is further ersicht- lent that the trajectories for each rear roll stand 1 in each case at least from the time at which the respective band point 8 enters the front roll stand 1, extend to the time to which the respective band point 8 in the respective rear roll stand 1 enters.

In the following, a possible type of evaluation of the determined trajectories by the computer 3 will be explained in conjunction with FIG. According to FIG. 5, the computer 3 calculates the trajectories in a step S21. In a step S22, the computer 3 evaluates the determined trajectories. In a step S23, the computer 3 checks whether the evaluation has revealed that the adjustment frame sizes A * must be changed. If necessary, the computer 3 makes the necessary changes in a step S24. For example, the computer 3 in steps S22 to S24 check whether plant limits of the rolling mill are met. For example, the computer 3 can check whether the rolling torque M and the rolling force F remain within the permissible framework for each rolling stand 1. If this is not the case for one of the rear roll stands 1, a Lastum ^¬ distribution can for example be made.

Of course, other limits may be taken into account accordingly, for example, a maximum possible ^¬ Liche roller peripheral velocity. Also couplings - at ^¬ play, the speed dependency of the maximum possible rolling moment M - can be considered. It is also possible to carry out the evaluation of the trajectories across rolling stands. For example - see FIG 6 - the sum of the powers P 'of the rolling mill drives are ^¬ telt ermit and with a maximum permissible limit power are (dashed lines in FIG 6 lines) compared.

It is also possible to predict state variables of the rear rolling stands 1 on the basis of the trajectories. In this case, the predicted state variables can be evaluated. Depending on the evaluation of the predicted state variables, it is likewise possible to decide whether and, if appropriate, in what manner the adjustment frame sizes A * are varied. For example, according to FIG. 7, the performance curve of one of the rear rolling stands 1 can be used to predict the resulting temperature profile of the corresponding rolling stand drive. In this example ^¬ illustrative case varying the Einstellgerüstgrößen A * may be required if a maximum permissible limit tempera ^¬ ture is exceeded. The trajectories are usually calculated with a modeling period t '. The modeling period t 'generally corresponds to the distance a of the band points 8 from each other or an integer multiple thereof. The evaluation of According to FIGS. 4, 6 and 7, jectors are performed with an evaluation horizon H, which is a multiple of the modeling period t '. In general, the ratio of Auswertungsho ^¬ rizont H is to modeling period t 'is 10: 1 or above. Even values of 20: 1 or 30: 1 and even higher ratios are possible.

From the above explanations regarding FIG. 5, it can be seen that the adjustment frame sizes A * may possibly be changed on the basis of the steps S21 to S24. The corresponding evaluations must therefore before driving the roll stand 1, for which the particular trajectory is valid, ^¬ follow. Preferably, the evaluation is carried out even before the driving of the front rolling mill 1.

The determination method according to the invention is usually carried out cyclically and / or in real time. In the case of the cyclic design, the last-mentioned statement ("before the control of the roll stand 1...") Refers only to the strip points 8 which were actually added in the last cycle.

As part of the procedure described above in connection with FIG 3 of the base automation ^¬ tion 9 "correctly timed" in step S16, the A * Einstellgerüstgrößen vorgege- ben. It is possible that the computer 3, the corresponding

Setting frame sizes A * transmitted directly to the base automation 9. Again, however, an alternative utilization of the trajectories is possible. This will also be explained below in connection with FIG. 5. However, the corresponding utilization is independent of the steps S22 to S24 of FIG 5 feasible. However, step S21 is present.

If the trajectories are also utilized in the context of the specification of Einstellgerüstgrößen A * to the base automation 9, the step S16 of FIG 3 is replaced by a step S26 shown in FIG 5. In step S26, the computer 3 transmits the trajectories of the Einstellgerüstgrößen A * at least partially to the base automation 9. For example, the Calculator 3 for each rear roll stand 1 determine which band point 8 is currently in the sphere of influence of the respective roll stand 1. For each rolling stand 1, the computer 3 then transmits to the basic automation 9 the current setting stand size A * for the respective rolling stand 1 and a - usually low - number of subsequent Einstellge ^{¬ set} sizes A * for the respective rolling mill 1 (for example, the next 2, 3 or 5 adjustment frame sizes A *). The base automation 9 are therefore provided at least for the rear rolling stands 1 corresponding courses of Einstellgerüstgrößen A * as a setpoint control variables.

As already mentioned, the trajectories (in step S21) are calculated with a modeling period t ', wherein the modeling period t' generally corresponds to the distance a of the band points 8 from each other or an integral multiple thereof. Due to the partial transmission of the trajectories of the base automation 9 of the base automation 9 future Einstellgerüstgrößen A * for a re gelzeitraum R known, for example for the above-mentioned 2, 3 or 5 strip points 8. Preferably, the Re ^¬ gelzeitraum R is greater than the modeling period t '. Purely by way of example shown in FIG 4, where as Modellie ^¬ approximately period T 'of the (simple, not multiple) distance a of the strip points 8 is assumed to each other and the control period R is t three times the modeling period' is.

Also shown in FIG 4, the control cycle Z is the basic automation 9. The control cycle Z of the basic automation 9 is considerably smaller than the modeling period t '. The control cycle Z is usually a few milliseconds ^¬ customers, usually at 20 milliseconds or less.

In the above, it was explained that the belt points 8 are assigned to the A ^¬ alternate scaffold sizes A *. It is also possible to additionally assign the expected framework sizes G to the band points 8. In this case, it is of course also possible to calculate the trajectories for the expected framework sizes G. For example, in the band point 8 as Target variable, a temperature profile as a function of time t before ^{¬ be} given - see FIG 8. In conjunction with the mass flow m 'or the Leitgeschwindigkeitsverlauf therefore a related to a particular rolling mill 1 temperature profile can be determined, with which the metal strip 2 of this rolling mill passes through or the band points 8 gradually pass through this Walzge ^¬ rüst 1, see FIG. 9

The investigation, as explained so far, is already working very well. It can be further improved by the embodiment, which is explained below in connection with FIG. FIG. 10 essentially shows measures which, in addition to removing a band point 8 from the totality of the managed band points 8, can be picked up by the computer 3 when a band point 8 has run out of the rolling train.

According to FIG. 10, in a step S31, after the respective strip point 8 leaves the rolling train, a final parameter PE of the respective band point 8 is detected. The Erfas ^¬ solution is generally carried out at a respective measuring station 10. The sensed parameters PE example, the Tempe ^¬ temperature or the geometry of the respective tape point 8 may be. It is also possible to record several PE parameters - for example, temperature and final thickness or final thickness and flatness.

In a step S32, the rolling-force model 7 is adapted based on the detected final parameter PE and the korrespondieren-, expect ^¬ th behind the last stand 1 of the rolling mill parameter P. The adaptation takes place in the step

532 in real time, so while other band points 8 are in the mill train. Furthermore, in one step

533 the expected parameters P of those band points 8, which have already run into the rolling train (but have not yet run out of the rolling mill), nachge ^¬ leads in real time. The procedure of steps S31 to S33 allows the adaptation in real time and is therefore compared to an adaptation in the context of a recalculation of great advantage.

For the parameter "temperature", the procedure of steps S31 to S32 is known from DE 101 56 008 AI For the parameter "geometry", the procedure of steps S31 to S33 in the already mentioned European patent application "real-time determination method for temperature and Geometry of a Metal Hot Strip in a Finishing Line ", Appl.'S Mark 201015307, described in detail.

In addition to the steps S31 to S33, a step S34 may be further included. However, the step S34 is only optional and therefore shown in dashed lines in FIG. If step S34 is present, the determination method explained above in connection with FIGS. 3 and 5 is executed again for those band points 8 whose parameters P have been tracked in step S33. For the renewed execution of the investigation, ie the execution of the investigation in the context of step S34, but for each such band point 8 that rolling mill 1 is considered as a front mill stand 1, in which the respective band point 8 enters next. The step S34 will be explained in more detail below - first in conjunction with FIG. 11, and later also in conjunction with FIG. 12.

According to FIG. 11, in a step S41 the computer 3 selects the first rolling stand 1 arranged downstream of the front rolling stand 1 of FIG. 3, for example the second rolling stand 1 of the rolling train. In a step S42, the computer 3 selects those band points 8 which are now in front of the selected rolling stand 1, ie at the time of the execution of step S42. These strip points 8 thus have that rolling mill 1, which is the upstream of the selected rolling stand 1, already pass, but have not yet run into the selected mill stand 1. In a step S43, the computer 3 checks whether the selected rolling stand 1 is the last rolling stand 1 of the rolling train. If this is not the case, the computer 3 proceeds to a step S44 in which it performs a subroutine call. As parameters of the subroutine of the computer 3 delivers the se lected ^¬ rolling stand 1 and the selected points on the strip 8 (including the tape these points 8 associated values). The selected rolling stand 1 is passed under the Unterpro ^¬ program call as a value ( "call by value"), the selective oriented strip points 8 (including the associated values) as a variable ( "call by reference"). The called Unterpro ^¬ program will be explained later in connection with FIG 12th

In a step S45, the computer 3 selects the next roll stand 1. Then, the computer 3 proceeds to step S42 to ^¬ back.

If the computer 3 determines in the examination of step S43 that the last rolling stand 1 has already been selected, the computer 3 proceeds to a step S46. In step S46 calculates the computer 3 for each now selected band point 8 by means of the rolling force model 7 for the (now se ^¬ selected) last rolling stand 1 of the rolling mill at least one expected during rolling of the selected band points 8 in the last rolling stand 1 of the rolling mill frame size G. In one

Step S47 assigns the computer 3 to the ermittel in step S46 ^¬ th stand sizes G the corresponding tape times 8. In the context of this assignment, any previous values of the expected framework sizes G are overwritten.

In a step S48, the computer 3 compares the determined framework sizes G with the corresponding nominal framework sizes G *. Furthermore, in step S48 the computer 3 determines at least one adjustment frame size A * for the last rolling stand 1 based on the comparison. In a step S49, the computer 3 assigns the determined setting frame sizes A * to the corresponding strip points 8. The steps S46 and S47 correspond substantially to - based on the last stand 1 of the rolling mill - the crotch ^¬ th S5 and S6 of FIG 3. The steps S48 and S49 correspond to steps Sil substantially and S12 of Fig. 3

Subsequently, the 12 already explained in United ^¬ connection with step S44 of FIG 11 mentioned subroutine in conjunction with FIG. According to FIG 12, the subroutine first takes in a

Step S51, the value of the at calling the subroutine selec ^¬ oriented roll stand 1 and the selected points on the strip 8 (a ^¬ finally their associated values) counter. Due to the nature of the transfer of the value for the roll stand 1 (namely, call by value), changes in the value of the selected rolling stand 1 made within the subroutine of FIG. 12 have no influence on FIG. 11.

The band points 8 themselves are not changed within the subroutine of FIG. The changes in the band points 8 associated values ^¬ SUC gene within the subroutine, however, act due to the nature of the value transfer (namely call by reference) to FIG 11 back. In particular, the values of the band points 8 changed as part of the subroutine of FIG. 12, if necessary, enter into the results which are determined in the steps S46 and S48 of FIG.

In a step S52, the computer 3 for each selected strip point 8 determined by means of the rolling-force model 7 for the selected roll stand 1 is at least an expected when rolling the se ^¬ lected tape point 8 in the selected rolling mill 1 stand size G. In a step S53 assigns the computer 3, the calculated Framework size G to the corresponding band point 8 too. The steps S52 and S53 correspond to the steps S5 and S6 of FIG. 3 with respect to the rolling stand 1 selected in the execution of the step S52. In a step S54, the computer 3 checks whether the currently selected as part of the subroutine mill 1 is the last rolling stand 1 of the rolling mill. If this is not the case, the computer 3 proceeds to a step S55, in which the computer 3 - with respect to the subroutine of FIG 12 - selects the next rolling stand 1. In a step S56 ermit ^¬ the computer 3 for the now selected roll stand 1 telt by means of the rolling-force model 7 each parameter P is the selec ^¬ oriented strip points 8, with which the selected points on the strip entering the now selected roll stand 1 8. In a step S57, the computer 3 assigns the determined parameters P to the corresponding band points 8. The computer 3 then returns to step S52. The steps S55 to S57 entspre ^¬ chen essentially steps S8 to S10 of Fig. 3

If the computer 3 has selected the last rolling stand 1 of the rolling train in the subroutine of FIG. 12, the computer 3 proceeds from step S54 to a step S58. In step S58, the computer 3 compares the determined framework sizes G with the corresponding target framework sizes G *. Further, the computer 3 determines in step S58 using the equalization Ver ^¬ at least a Einstellgerüstgröße A * for the roll stands 1. In a step S59, the computer assigns the determined Einstellgerüstgrößen 3 A * the corresponding band to 8 points. The steps S58 and S59 substantially correspond to the steps S11 and S12 of FIG. 3. The step S58 is executed from the rolling stand 1, the value of which was taken in step S51, to the last rolling stand 1.

In step S60, the computer 3 returns to step S44 of FIG. 11, from which the subroutine of FIG. 12 has been called. The "duplicate" presence of steps S48 and S58 is only seemingly given, because steps S48 and S58 handle only the values of the currently selected band points 8, as an alternative to the presence of both steps S48 and S58 also of the step S58, it is possible to provide only the step S48, but not the step S58. In this case, in step S48, all band points 8 must be handled. In this case, of course, step S59 also disappears together with step S58.

The procedure of the invention has - even beyond the be ^¬ already mentioned in the introduction to properties beyond - many advantages:

- Plant limits can be included exactly. With the approach of achieving any belie ^¬-lived limit of the rolling mill is in advance exactly detected as the only torque ^¬ ment limit of a drive, and it reacts accordingly, for example with a load redistribution,

It is possible to communicate to a final rolling temperature control accurate forecasts of future limits of the plant and to make sure that the final rolling temperature control in turn intervenes in good time to avoid exceeding limits.

- The deviations of the strip thickness, which occur in the case of a model ^¬ error on the tape head are corrected faster. This prolongs the fillet, which is the high-quality strip part.

The above description is only for explanation of the present invention. The scope of the present invention, however, is intended to be determined solely by the appended claims.

Claims

claims

1. Determination method for control variables of a rolling mill with several rolling mills (1) for rolling a metal strip (2), - wherein prior to entry of band points (8) of the metal strip (2) in a front rolling stand (1) of the rolling mill depending ^¬ initial parameters (P) of the respective band point (8) are detected, with which the respective band point (8) enters the front rolling stand (1), and the respective band point (8) is taken over into the active evaluation,

- wherein by means of a rolling force model (7) for the front rolling stand (1) at least one expected during rolling of the respective band point (8) in the front rolling stand (1) framework size (G) is determined

- In addition by means of the rolling force model (7) for the front roll stand (1) downstream rear rolling stands (1) respectively parameters (P) of the respective band point (8) are determined with which the respective band point (8) in the respective rear rolling stand ( 1) enters,

- wherein by means of the rolling force model (7) further for the rear rolling stands (1) in each case at least one when Wal ^¬ zen of the respective band point (8) in the respective rear rolling stand (1) expected frame size (G) is determined

- Where the expected framework sizes (G) of the rolling stands (1) with respective the respective band point (8) associated target stand sizes (G *) are compared and for the front roll stand (1) and the rear rolling stands (1) based on the comparison in each case at least one respective set-up frame size (A *) is determined,

- At least the determined Einstellgerüstgrößen (A *) the respective band point (8) before the arrival of the respective band point (8) in the front roll stand (1) zugeord ^¬ net and tracked with the respective band point (8) through the rolling mill .

- Based on the band points (8) associated Einstellgerüstgrößen (A *) at least for the rear mills (1) trajectories can be calculated for the setting framework sizes (A *),

- Wherein the trajectories for each rear rolling stand (1) each extend at least from the time at which the respective band point (8) enters the front rolling mill (1) until the time at which the respective band ^¬ point (8) enters the respective rear rolling stand (1).

2. Investigation process according to claim 1,

characterized ,

that the trajectories are computed and that the Trajekto ^¬ rien be at least partially transmitted to a basic control (9), so that they are the basic control (9) as a flow of target manipulated variables for the rear roll stands (1) are available.

3. Investigation process according to claim 2,

characterized ,

that the trajectories are calculated with a modeling period (f) and that the trajectories are transmitted to the basic automation (9), which is greater than the modeling period (f), at least for a control period (R).

4. The investigative procedure according to claim 1, 2 or 3,

characterized ,

that the trajectories are calculated, that the trajectories before driving of the rolling mill (1) for which the respective trajectory is valid, are evaluated and in that in depen ^¬ dependence decided by the evaluation of whether and possibly how the Einstellgerüstgrößen ( A *) can be varied.

5. Investigation process according to claim 4,

characterized ,

that the trajectories are calculated with a modeling period (f) and that the evaluation of the trajectories takes place with an evaluation horizon (H) that is a multiple of the modeling period (f).

6. Investigative procedure according to claim 4 or 5,

characterized ,

in that the predicted state variables are evaluated on the basis of the trajectories of state variables of at least the rear rolling stands (1) and that it is decided as a function of the evaluation whether and, if so, how the setting framework sizes (A *) are varied.

7. Investigative procedure according to claim 4, 5 or 6,

characterized ,

that in the context of the evaluation of the trajectories, plant limits of the rolling train are taken into account.

8. Investigative procedure according to one of claims 4 to 7, d a d u c h e c e n e c e n e,

the evaluation of the trajectories takes place across the rolling mill.

9. Investigative procedure according to one of the above claims, d a d e r c h e c e n e c e n e,

in that the expected frame sizes (G) are also assigned to the respective band point (8) and that the trajectories for the expected frame sizes (G) can also be calculated.

10. Investigative procedure according to one of the above claims, d a c e c o v e c e n e c e n e,

that it is executed cyclically and in real time.

11. Investigative procedure according to one of the above claims, d a c e c o v e c e n e c e n e,

that determination rules by means of which the expected framework sizes (G), the parameters (P) of the band points (8) and the alignment framework sizes (A *) are determined are continuous replicas.

12. Investigative procedure according to one of the above claims, characterized that after the expiry of the respective band point (8) from the rolling train at least one final parameter (PE) of the respective band point (8) is detected, that the rolling force ^¬ model (7) based on the detected final parameter (PE) and behind the last Roll stand (1) of the rolling mill he ^wa waited parameter (P) is adapted in real time and that the expected parameters (P) of those band points (8), which have already run into the rolling mill, in real time after ^¬ led.

13. Investigation process according to claim 12,

characterized ,

that it points to the adaptation of the rolling-force model (7) and the tracking of the expected parameter (P) for those belt (8) whose parameters have been tracked (P) is carried out again, and that for the re-execution of the He ^¬ averaging procedure that of Roll stand (1) is considered as a front stand, in which the respective band point (8) whose parameters (P) were tracked next enters.

14. Investigative procedure according to one of the above claims, d a d e r c h e c e n e c e n e,

at the latest at the time of the detection of the initial parameters (P) of the respective band point (8), a master speed profile is known with which the respective band point (8) is to pass through the rolling train.

15. computer program, the machine code (6), which is directly abarbeitbar of a computer (3) and its processing by the computer (3) causes the computer (3) a determination process with all the steps of a determination method according to one of the above claims performs.

16. Calculator,

characterized ,

that the computer is designed such that it has an investigator procedure with all the steps of a preliminary investigation according to one of claims 1 to 14.

17. Rolling mill with several rolling mills (1) for rolling metal strip (2),

characterized ,

in that the rolling train is equipped with a computer (3) according to claim 16.