EP3539777A1 - Method and device for correcting the printing position of a printing unit and printing machine - Google Patents

Abstract

The invention relates to a method for correcting a printing position of a printing unit (14) of a printing machine (10) having at least one sensor unit (12), comprising the steps of: detecting an actual position of at least one by means of the printing unit (14) on a printing material (16 ) printed print mark (18) by means of the sensor unit (12), - determination of a deviation of the detected actual position from a desired position, - determination of at least one difference value characterizing the deviation, - determination of a buffer value from a number of in a memory (20 22), - determining a correction value by subtracting the buffer value from the difference value, - storing the determined correction value as one of the support values (22) in the memory (20), - correcting the printing position based on the determined correction value; Furthermore, the invention relates to a device and a printing press (10).

Description

The invention relates to a method for correcting a printing position of a printing unit according to claim 1. Further, the invention relates to a method according to claim 11 and a printing press according to claim 12.

A printed product, such as a magazine, a newspaper or even a packaging, is usually produced by means of a printing press. Printing presses are also used in functional printing and decor printing. The printing press has at least one printing unit, which thus represents an assembly or component of the printing press. In general, the printing unit comprises a rubber or printing cylinder and often, for example, a color and / or dampening unit. If the printed product or the printed copy is to be multicolored, the so-called multi-color printing is used as the printing method. In this case, a color separation takes place in the individual printing inks, since the printed image, which is to be printed on a substrate to create the printed product, is broken down into primary colors and spot colors. Printing machines for multi-color printing thus generally have their own printing unit for each printing ink. Thus, for example, in a four-color printing, a decomposition into the colors cyan, magenta, yellow and black takes place. The colors of the printed image or the color of the printed image is made by the composition of the printing inks. In current printing machines often finds a series construction application, that is, the individual printing units are arranged one behind the other.

In order to obtain the highest possible quality of the printed product, the individual colors must pass through the printing units in a fixed predetermined, in particular exact, position are printed one above the other, so that the printed image is perceived particularly sharp. Print marks are used to measure the position of the inks. In this way, a so-called Passer, which is also referred to as register or color register and describes a matching of the individual colors in multi-color printing, are controlled. Thus, in the case of said multi-color printing, the process colors or printing inks are printed successively and one above the other and / or next to one another, thus producing the complete printed image. If the inks are not printed exactly one above the other and / or next to each other, the printed image will appear blurry or out of focus or show color shifts, which will reduce the quality.

It is said that the printed image is in the register when the printing inks are exactly on top of each other. If this is not the case, one can speak of a so-called register or registration error. This can, for example, if the press is designed as a sheet-fed press, caused by an inaccurate transport of the sheet of the printing material by the printing press. Passer problems can also occur, for example, in a web press. In order to bring the printed image into the register, a register control is used, for each of which a respective additional information, the print mark, is applied to the printing material. With the help of the respective print mark, the position of the individual print images printed in the respective color can be determined. Thus, a so-called brand field is printed per printed copy containing the individual print marks. In each brand field, the offset of the individual print marks in the longitudinal and / or lateral direction is measured by a sensor unit. As a rule, an error is calculated from a desired position and an actual position of the print mark, which is corrected. Due to the arrangement of the individual printing units, for example, sequenced in the printing press For example, the distance of the sensor unit to a respective printing unit can be particularly large, in particular if only one sensor unit is arranged after the last printing unit. This is called a so-called end-of-line arrangement, with particularly large distances between the printing unit, ie the control point, and the sensor unit, ie the measuring point, arise. Until a copy of the printed product, which was printed in the printing unit, reaches the sensor unit, a number of further printed copies were usually already printed by the printing unit, which are then measured in the order in which they were printed by the sensor unit. That is, the correction carried out is not yet visible in the or the next measurements of the position of the print mark by the sensor unit.

Now, if no additional measures are taken to perform the register control, the already corrected error would be corrected again, which leads to overshoot or instabilities in the scheme.

From the prior art, two ways are known how to solve the problem. In the first variant or the first type, no correction is carried out until the printed copy with the correction has arrived at the measuring point or the sensor unit. In the second mode, a gain of the controller is canceled, so that a correction value which controls the printing position of the print image in the printing unit is calculated, which is smaller than the actual register error. However, the two measures mentioned in the prior art have the disadvantage that a particularly sluggish control behavior occurs, which in particular becomes all the more sluggish, the greater a distance between the printing unit and the sensor unit, ie between the pressure point and the measuring point. becomes.

It is therefore an object of the present invention to provide a method and a device by means of which a register control is particularly advantageous feasible and thus print images can be realized with very high quality.

This object is achieved by a method having the features of the independent claims. Advantageous embodiments with expedient developments of the invention are specified in the remaining claims.

A first aspect of the invention relates to a method for correcting a printing position of a printing unit. In this case, the printing unit is a component or a part of a printing machine having at least one sensor unit. The printing press can for example produce a printed product by means of flexographic printing, gravure printing, screen printing, digital printing or offset printing. The printing machine can be, for example, both a printing machine for a roll printing and a printing machine for a sheet printing. In principle, the printing press can be any printing press.

In order to be able to correct the printing position of a printing unit, that is to say the position of a printed image or printing motif, on a printing material for producing a printed copy or a printed product in a particularly advantageous manner, the method comprises a plurality of steps:
In a first step or method step, an actual position of at least one print mark printed on the printing substrate by means of the printing unit is detected by means of the sensor unit. That is, the sensor unit is formed, for example by means of a sensor to detect the actual position of the printed print mark on the substrate. In a second step of the method, a deviation of the detected actual position from a desired position is determined. In other words, the deviation corresponds to one Offset, in particular in the longitudinal direction and / or transverse direction of the actual position of the print mark to the, in particular for a particularly high quality of the printed product, required position. In a third step of the method, at least one difference value characterizing the deviation is determined. This difference value can be, for example, a distance between the desired position and the actual position. In addition to the distance, ie the length of the distance, the difference value can have, for example, a direction value. For example, the difference value can also be represented as a vector, the amount of which is the distance between the setpoint and the actual position. In a fourth step or method step, a buffer value is determined from a number of support values previously stored in a memory.

The memory is, for example, a memory area of an electronic computing device in which a plurality of values, in particular backup values, can be stored or stored. For example, in a first implementation of the method, that is, when it is performed for the first time, the support values held in the memory are individually determined to be zero. The first implementation of the method can be performed, for example, at a start of a pressure or the like. The memory is preferably designed as a buffer memory, that is, it serves for a caching of the support values and not for a permanent storage of the support values. In a fifth step of the method, a correction value is determined by subtracting the buffer value from the difference value. In a sixth step, the determined correction value is stored as one of the support values in the memory, that is to say the correction value is stored in the memory so that it can be used when the method for establishing the buffer value is carried out again. Finally, in a seventh step of the method, the printing position is corrected on the basis of the determined correction value. That is, the printing unit is manipulated so that the printing position of the printed image to be printed on the Printing material is changed, in particular such that the actual position has been changed by at least the correction value at a future pressure. As a result, a control difference is changed, which represents the difference between the desired position and the currently measured actual position.

In other words, it can be achieved by means of the described method that each individual measured value of the respective actual position detected by the sensor unit can be used for the correction and thus a respective register correction can be made for each printed copy. For this purpose, the current correction value determined in each case when carrying out the method is entered into the memory. If the method is carried out again in the determination or calculation of the new, more recent correction value, the current system deviation as well as the correction already in the route can be taken into account. It should be understood as the route the path between the printing unit and the sensor unit, which is to be understood as the "in-route correction" the least one stored in the memory as a backup value former correction value.

The respective base value is formed from a previous correction value determined in a previous execution of the method. That is to say, if the method according to the invention is carried out, for example, at least a second time, the correction value of the first execution becomes the support value of the second implementation. If the process is repeated so many times that the number of repetitions corresponds to the number of supporting values stored in the memory, the memory is completely filled with reference values, which are each formed from a correction value of one of the preceding implementations of the method. An initialization of the support values, each with a predetermined value, for example in the case of a new print job, serves to carry out the method until the support values can each be formed from a correction value.

By means of the method described, any actual position detected by the sensor unit can be used for the register control and thus for the correction. As a result, the register correction, that is, the correction of the printing position, can take place at each print copy, that is to say for each individual printed product which is printed by the printing unit and conveyed or moved through the printing press to the sensor.

Due to the distance between the printing unit and the sensor unit occurs a so-called dead time, which is also called running or transport time, between the pressure of the print mark and the detection selbiger on. Since the dead time depends on the known distance between the printing unit and the sensor unit, it is also possible to speak of a dead path. Due to the dead path, an already performed correction in the currently detected actual position is not yet visible. However, this correction or the associated correction value is entered in the memory. If the method is now carried out again, when the new correction value is determined, consideration is given to the at least one current correction value of the correction carried out. For example, the correction may be performed in a certain area so that the correction does not corrupt the print marks.

By means of the method according to the invention, a particularly advantageous register control for correcting the printing position of the printing unit can be realized, in particular for large distances between printing unit and sensor unit. The register control can work very dynamically and stably, which can be avoided, for example, overshoot.

In an advantageous embodiment of the invention, the reference value stored longest in the memory relative to the support values stored in the memory is deleted from the memory and replaced by the determined correction value. As already mentioned, the memory can be used for a first execution of the Method have each initialized to zero support values. The memory locations in the memory, that is to say the areas in which a backup value is or can be stored in each case, furthermore have a fixed reference with respect to a time sequence. Thus, the support values, which are formed in particular from respectively an old correction value or correspond to it, can be read from the memory in such a way that the time sequence of their entry into the memory can be recognized. If a first correction value is entered into the memory as a new base value, then the correction value, which was entered last before the first correction value, moves from a first memory location to a second memory location. A correction value, which lies in the last memory location, is deleted from the memory or discharged and occupied by the support value of the penultimate memory location, which thus becomes the last support value of the memory when the new base value is entered. By deleting or from the memory, pushing the most recently stored in the memory and the addition of the currently determined correction value as a new base value, in the memory, a current number of support values, which in particular from respective correction values, each in an implementation of the invention Are kept so that the correction of the printing position can be carried out particularly efficiently and / or with particular precision. If the procedure has been repeated so many times that the number of repetitions corresponds to the number of memory locations in the memory, the memory is completely filled with the most recent support values.

In an advantageous embodiment of the invention, the memory, which can also be referred to as a buffer, designed as a shift register. In a shift register, the number of available memory locations, in each of which a backup value can be stored, is constant. Each time the shift register is written, a superscript value is shifted one space further, so that the first written one or more stored single value leaves the store first. Pushing in and out or reading in and out are generally synchronized. By using the shift register as a memory when carrying out the method, the method can be carried out in a particularly advantageous manner, since the respective correction values entered as support values in the memory or shift register can be used particularly simply to determine the respective current correction value.

In an advantageous embodiment of the invention, the number of support values stored in the memory is determined by a distance between a printing location of the print mark in the printing unit and a read-out position of the sensor unit. Furthermore, this number is determined by a format length of the printing material. The number is advantageously the rounded quotient of the distance and the format length. The format length is the length which the printing material has along the direction along which the printing material can be transported or transported through the printing press and in particular through the printing unit. In other words, the number of supporting values stored in the memory corresponds to the number of the following printed copies: the printed copy whose print mark is detected by the sensor unit at one time, the printed copy which is currently being printed by the printing unit, and the printed copy (s) which or which are or are currently in the printing press and at the time already printed by means of the printing unit, but has not yet been detected by the sensor unit or have been. By the number of printed copies, which corresponds to the number of supporting values stored in the memory, the respectively current correction value can be calculated particularly dynamically. Furthermore, for example, an overshoot can be avoided because the control difference, the difference between the desired position and actual position due the dead time or the dead travel can be compensated for by the number of stored support values, or the support values are taken into account in the correction of the printing unit. In particular, the number of printed copies or copies which have already been printed, but whose respective print mark has not yet been recorded, is to be understood as that which corresponds to the number of printed copies in a continuous operation of the printing press, which is just from the printing unit, for example by means of a conveyor, on the way to the sensor unit.

In an advantageous embodiment of the invention, a plurality of actual positions of respective printed by means of the printing on respective substrates print marks are detected by the sensor unit, wherein an actual position average of the plurality of actual positions is formed and the difference value by means of the actual position Average value is determined. In other words, the actual position is smoothed or averaged. Alternatively, a plurality of actual positions of respective, printed by means of the printing on respective substrates print marks are detected by means of the sensor unit. In this case, the respective deviations of the detected actual positions from those of the desired positions are determined, whereby respective deviation values characterizing the respective deviations are determined. The difference value, which is used to form the correction value, can be determined from the actual position mean value or from the mean value of the deviation values. In this case, a number of printed copies, from which the actual position average value or the mean value of the deviation values is formed, can be referred to as the so-called filter depth. Thus, the filter depth corresponds to the number of actual positions used to form the actual position average. By averaging or averaging the actual positions or the deviation values, for example, a, in particular metrological, fluctuation in detecting the actual position can be compensated.

In an advantageous embodiment of the invention, the number of supporting values stored in the memory corresponds to the sum of a first number and a second number. The first number is determined by the distance between the printing location and the read-out position and / or the format length of the printing substrate. The first number is advantageously the rounded quotient of the distance and the format length. The second number corresponds to the number of actual positions used to form the actual position average and thus the filter depth. In other words, the number of supporting values stored in the memory is the sum of a first number and a second number. In this case, the first number is the number of the following print copies: The printed copy whose print mark is detected by the sensor unit at a time, the print copy which is currently printed by the print unit, and the print copy or copies at the time are located in the printing press or are and at the time already printed by means of the printing unit, but has not yet been detected by the sensor unit or were. The second number of the sum, wherein the sum describes the number of memory locations, corresponds to the number of actual positions used for forming the actual position average, that is to say the filter depth mentioned. This is based on that the dead time or the dead path is extended in the case of averaging. If the correction of the printing position of the printing unit is regulated with an averaged or smoothed actual position or a predetermined filter depth and the extended dead path is not taken into account, this can lead to an erroneous calculation of the correction and thus to an unstable behavior of the control. By extending the number of memory locations by the second number, the number of actual positions used to form the actual position average, in the memory or shift register, the correction is made more precise or the filter depth is taken into account in the correction, so that an unstable behavior of the scheme can be excluded.

In an advantageous embodiment of the invention, the buffer value is determined as the sum of the supporting values. That is to say, the support values stored or stored in the memory, which in particular represent or are the correction values determined at an earlier time, that is to say in an earlier execution of the method, are added up, whereby the buffer value is formed or determined. By forming the sum and subtracting this sum as a buffer value from the difference value, a respective current correction value can be determined particularly quickly and, for example, an override of the correction of the printing position of the printing unit can be kept particularly low.

In an advantageous embodiment of the invention, the sum of at least one of the supporting values is weighted when determining the sum. That is, at least one of the support values is in particular multiplied by a scalar, so that this support value has a particularly large or a particularly small influence on the determination of the correction value in comparison to the other support values. Thus, for example when using the above-mentioned filter depth and thus the extension of the memory locations of the memory by the second number, the last supporting values can be weighted in each case. In this case, the last support values are the support values that have already been passed through the memory, so that they lie in the last memory locations before leaving the memory, so that the last mentioned support values fall from the memory when a second number of new correction values is entered. In other words, if the method is repeated so often that the number of repetitions corresponds to the second number, then the abovementioned last reference values are removed from the memory, with only the reference value located in the last memory location being carried out from the memory for each execution of the method falls. For example, is the filter depth three, the last support value can be weighted with 0, the penultimate support value with 1/3 and the penultimate support value with 2/3. As a result, the mean value binding determined by the filter depth can be taken into account in the correction in a particularly advantageous manner.

In an advantageous embodiment of the invention, the sensor unit has a camera sensor and / or a fiber optic. By means of the camera sensor, for example, the printed mark can be detected in a particularly advantageous manner and read in particular for subsequent processing. Furthermore or alternatively, light which is reflected by the print mark can be guided in a particularly advantageous manner to a suitable sensor of the sensor unit by a fiber optic, so that this light can be detected. Furthermore, by using a fiber optic, the sensor unit can be arranged relatively freely on or in the printing press. Alternatively or additionally, the respective print mark can be detected with other techniques, such as a contrast scanner. The sensor unit should be designed such that it can provide sensor data from which the actual position can be determined or detected by means of the method according to the invention.

In an advantageous embodiment of the invention is used as the substrate paper and / or cardboard and / or plastic and / or metal and / or wood and / or glass. In this case, the printing material may be formed, for example, as a film, that is, that of the substrate at least partially forming material, in particular, for example, plastic and / or metal, is designed to be particularly thin. In addition, the method can be applied to other materials which are suitable for printing, so that the method for a particularly large variety of printing machines is feasible and thus for a particularly large number of different printed products can be realized. Thus, in a variety of printed products, the inventive Applied method and thus the correction of the printing position are carried out particularly efficiently.

A second aspect of the invention relates to a device for correcting a printing position of a printing unit of a printing machine, comprising at least one electronic computing device, which is adapted to an actual position of at least one printed by means of the printing unit on a printing material print mark based on sensor data from a sensor unit to be received. Furthermore, the electronic computing device is configured to detect a deviation of the detected actual position from a desired position, at least one difference value characterizing the deviation, a buffer value from a number of support values previously stored in a memory, and a correction value, by subtracting the buffer value from the value Determine difference value. Furthermore, the electronic computing device is designed to store the correction value as one of the support values in a memory and to provide a signal for the correction of the printing position on the basis of the determined correction value, wherein the printing position of the printing unit can be corrected by means of the signal.

A third aspect of the invention relates to a printing press comprising a device according to the invention for correcting a printing position of a printing unit.

Advantages and advantageous embodiments of the first aspect of the invention are to be regarded as advantages and advantageous embodiments of the second and / or third aspect of the invention and vice versa.

• FIG 1 eine schematische Ansicht einer Druckmaschine, welche eine Sensoreinheit und wenigstens ein Druckwerk aufweist; und
• FIG 2 ein schematisches Ablaufdiagramm eines Ausführungsbeispiels eines Verfahrens zum Korrigieren einer Druckposition wenigstens eines der Druckwerke der Druckmaschine.

Embodiments of the invention are explained in more detail below with reference to schematic drawings. It shows:

• FIG. 1 a schematic view of a printing machine, which has a sensor unit and at least one printing unit; and
• FIG. 2 a schematic flow diagram of an embodiment of a method for correcting a printing position of at least one of the printing units of the printing press.

FIG. 1 shows a schematic view of a printing press 10, which has a sensor unit 12 and at least one printing unit 14. In particular, in multi-color printing, in which a printed image is decomposed into primary colors, wherein each printing unit 14 prints one of the primary colors on a printing material 16, so-called registration errors can occur. In a register error, the individual inks are not exactly on top of each other, so that the finished printed copy or printed product, for example, appears blurred. By means of the method presented, the register error can be compensated particularly dynamically and thus efficiently, whereby the printed image and thus the finished printed copy or its individual colors are in the register.

The method for correcting the printing position of the, in particular respective, printing unit 14 of the at least the sensor unit 12 having printing machine 10 comprises a plurality of steps:
In a first step of the method, an actual position of at least one printed mark 18 printed on the printing substrate 16 by means of the printing unit 14 is detected by means of the sensor unit 12. In a second step, a deviation of the detected actual position from a desired position is determined. In a third step, at least one difference value characterizing the deviation is determined. In a fourth step, a buffer value is determined from a number or plurality of support values 22 previously stored in a memory 20. In a fifth step, a correction value is determined by subtracting the buffer value from the difference value. In a sixth step, the correction value determined in the fifth step is stored as one of the support values 22 in the memory. In a seventh Step, the printing position of the printing unit 14 is corrected on the basis of the determined correction value.

The printing machine 10 can print, for example, by means of a so-called flexographic printing, for example, packages. In addition, the printing press 10 may also be a printing press, which is designed for offset printing. Furthermore, the printing material 16 or the printing materials 16 can be printed by means of, for example, a web-fed printing or a sheet-fed printing. In order to be able to print on the printing material 16 in a particularly advantageous manner, the respective printing group 14 advantageously has at least one printing cylinder 24 in each case. Depending on the nature of the printing process or the printing machine 10, the respective printing unit 14 instead of the printing cylinder 24, for example, have a print head, as it may be the case for example in digital printing.

By means of the respective printing cylinder 24 of the respective printing unit 14, the respective printing ink can be applied to the printing material 16 in a particularly advantageous manner. In this case, the printing material 16 is advantageously paper and / or cardboard and / or plastic and / or metal, whereby a particularly large variety of different printed products by means of the printing press 10 can be produced. In the exemplary embodiment shown, the printing press 10 has two printing units 14 arranged one after the other. One form of multi-color printing is the so-called four-color printing, in which in particular the colors cyan, magenta, yellow and black are used to produce colored printed products. In a four-color printing thus four printing units 14 would be provided, for the sake of simplicity are in the FIG. 1 however, only two printing units 14 are shown. In the method, the number of printing units 14 of the printing press 10 can be freely selected. Advantageously, the sensor unit 12 has a camera sensor and / or fiber optics, as a result of which the respective print mark 18 can be detected or detected in a particularly simple manner by the sensor unit 12.

Further shows FIG. 1 in that a distance between the printing units 14 and the sensor unit 12 is a certain length. This length can be, for example, as in FIG. 1 shown, the length of two substrates 16 amount. This leads to a dead time, that is, the print mark 18, which is detected by the sensor unit 12 at a first time, has already been printed at an earlier, second time. Conversely, the print mark 18 printed at the first time is only detected by the sensor unit 12 at a later third time. The distance between the time and the second or third time is called dead time. This can also be interpreted or understood as a dead path. So run in the printing press 10 of FIG. 1 the substrates to be printed 16 lined up from the left first on the first printing unit 14 and then on the second printing unit 14 and finally meet one behind the other and thus successively in the sensor unit 12 a. In the embodiment shown, a printed copy 26 between the last printing printing unit 14 (the right printing unit 14 of FIG. 1 ) and the sensor unit 12 are arranged in the printing machine 10. That is, until the sensor unit 12 gets to see a printed copy 26 or the printing material 16, which were printed by the last printing unit 14, two further copies 26 have already been printed by this printing unit 14.

The respective print mark 18 can, as shown in the example, be formed from individual rectangles or triangles, wherein in particular at least one rectangle or at least one triangle is provided for each color to be printed. The method is fundamentally independent of the characteristics of the print marks 18. For the procedure, measured values need only be available at a defined time. In the exemplary embodiment, the last printing unit 14 is designed to be in the FIG. 1 left triangle of print mark 18 to print. Alternatively or additionally, in particular, depending on the type of sensor unit 12, the print marks For example, be dots in the respective ink. Thus, the rectangular and triangular configuration of the print marks 18 or the mark fields has established itself for a fiber-optical detection of the print mark, so that they are designed as wedge or block marks, whereas often find so-called dot marks for camera-based sensor units 12.

FIG. 2 1 shows a schematic flowchart of an embodiment of the method for correcting the printing position of, in particular, the respective printing unit 14 of the printing press 10. Advantageously, the memory 20 is designed as a shift register, so that a repository value written into the memory 20 when re-adding another support value 22 through the memory or shift register is passed through and leaves it after reaching a last memory position of the memory 20. By using a shift register, a fixed time reference of the reference value 22 can be produced in a particularly simple manner at the respective detection time, for example the first time, whereby a time sequence of the reference values 22 in the memory 20 can be realized. Thus, in a particularly advantageous manner, the reference value 22 stored longest in the memory 20, relative to the support values 22 stored in the memory 20, can be deleted from the memory and replaced by the correction value determined by the current repetition of the method as new or current base value 22 , In particular, by forming the memory 20 as a shift register, an association between one of the memory positions and a time sequence of the repetitions of the method is given.

In the schematic representation of the method, which in FIG. 2 is shown, the respective individual printed copies 26 or printing materials 16 coming from the right following one after the other printed by the printing unit 14 and then the print marks 18 through the sensor unit 12 recorded. In this case, the sawtooth-like line 27 indicates the respective position of a guide axis of the printing unit 14 or of the printing cylinder 24. This means that after a complete revolution of the printing cylinder 24, the leading axis has returned to its original position.

The method is performed the first time when the first printed copy 26 or its print mark 18 is detected by the sensor unit 12. In this case, the memory 20, whose memory size 28 in the example shown comprises a number of three memory locations, which is represented by the left curly bracket, corresponds to the number which is predetermined inter alia by the distance or the dead path between the printing unit 14 and the sensor unit 12. The number is determined by the number of copies 26, 30, 32: The printed copy 26, the print mark 18 is detected at a time by means of the sensor unit 12, the other printed copy 30, which is printed at the time by means of the printing unit 14 and the printed copy 32 or the printed copies 32, which are located at the time in the printing press 10 and at the time already printed by means of the printing unit 14, but has not yet been detected by the sensor unit 12 or were. In other words, the number of supporting values 22 stored in the memory is determined by a distance 42 between a printing location of the print mark 18 in the printing unit 14 and a readout position of the sensor unit 12. Further, this number is determined by a format length 40 of the printing material. Here, the number of rounded quotient of the distance 42 and the format length 40.

With respect to the line 27, which determines the position of the leading axis, a time axis 34 is defined. The position of the leading axis is shown in a second, perpendicular to the axis 36. Thus, the first print mark of the first print copy 26 is detected at the first time ZP1, at which time the support values 22 of the store are all 0 are fixed. This represents the first step of the method, which is carried out in the example shown at time ZP1. This is followed in the second step, the determination of the deviation of the detected measurement position of the print mark 18 from the target position. From this deviation, a difference value characterizing the deviation is determined in the third step. In the next fourth step, a buffer value is determined from a number of support values 22 previously stored in the memory 20. In the first implementation of the method in which the memory 20 is in the state V1, the buffer value is 0, since no correction has yet been carried out prior to the first execution of the method. From the buffer value, in a next fifth step of the method, the correction value is formed by subtracting the buffer value from the difference value. Then, in the sixth step of the method, the determined correction value is stored as one of the support values 22 in the memory 20.

At time ZP2, the first correction value KORR1 is stored in the memory and the third print mark 18, that is to say the print mark 18 of the printed copy 30, is printed, but not yet measured. If the method is carried out again at the time ZP3 or directly after the time ZP3, the steps of the method are performed again so that a new correction value KORR2 is new at the time ZP4 at which the fourth print mark 18 was printed on a further print copy Suppose value 22 has been written to the memory 20 or is written. At the time ZP5, the third print mark 18, that is to say the print mark of the print copy 30, is measured and the method is carried out a third time again, so that the correction value KORR3 is already entered into the memory 20 at time ZP6 at which a fifth print mark 18 is printed is. At time ZP7, the fourth printed mark 18 is detected and the process is performed again, so that at time ZP8 at which the sixth mark 18 is printed, the fourth correction value KORR4 is entered in the memory 20. Here, the memory 20 is filled for the first time with reference values 22 formed by the earlier implementations of the method, formed from the correction values KORR1 to KORR3, so that the correction value KORR1 written in the memory at the time ZP1 or time ZP2 is forced out of the memory area 28. At the times ZP9 and ZP11, the process is started again, since the first process step is performed at the respective time. In each case at the time ZP10 or time ZP12, the process is completed and each printed another print mark 18.

The buffer value is formed from the sum of the support values 22 stored in the memory 20 with the memory size 28, which are formed from respective correction values (KORRn to KORRn + 2) for a respective implementation of the method. This buffer value, which is updated before each execution of the method, is subtracted from the respective current difference value characterizing the deviation. Memory size 28 is also referred to as memory depth.

The method can alternatively be carried out such that a plurality of actual positions of respective printed by means of the printing unit 14 printed on respective substrates 16 print marks 18 are detected by the sensor unit 12, wherein an actual position average of the plurality of actual positions is formed and the difference value is determined by means of the actual position mean value. Thus, for example, in the case of three different printed copies, for example the printed copies 26, 30 and 32, the actual positions are first detected and averaged and the deviation of the desired position from the averaged actual position or the actual position average is determined. This is referred to as a filter depth, which in the example corresponds to three, that is to say the number of printed marks 18 used and used for averaging. Thus, the filter depth is equal to the number of formation of the Actual position average used actual positions. If such a difference value or actual position mean value smoothed by the averaging is used in the method or if this method is used for the method, advantageously the memory size 28, that is to say the number of memory locations which can be stored in the memory 20, increases by the number of times to form of the actual position mean used actual positions. Advantageously, the additional support values 22 are weighted in the determination of the buffer value by the sum of the support values 22.

If smoothed deviation values are used, additional values 22, which are taken into account by the weighting in the formation of the buffer value, can additionally be provided by the extended memory size 28 of the memory 20 in the form of a shift register. Thus, in the correction of the printing position, a faulty deviation can be avoided, since the averaging of the actual position average value extends the dead time or the dead travel. Thus, by the method in a particularly efficient and / or dynamic manner, a regulation of print marks 18 by means of a, in particular as a shift register, formed memory 20 and thus lead to qualitatively very precisely printed copies 26, 30, 32, wherein a smoothing of the actual position can be carried out.

Claims (12)

Method for correcting a printing position of a printing unit (14) of a printing machine (10) having at least one sensor unit (12), comprising the steps of: - Detecting an actual position of at least one by means of the printing unit (14) on a substrate (16) printed print mark (18) by means of the sensor unit (12); Determining a deviation of the detected actual position from a desired position; Determining at least one difference value characterizing the deviation; - determining a buffer value from a number of support values (22) previously stored in a memory (20); - determining a correction value by subtracting the buffer value from the difference value; - storing the determined correction value as one of the support values (22) in the memory (20); - Correction of the printing position based on the determined correction value. Method according to Claim 1, characterized in that the reference value (22) stored longest in the memory (20), based on the reference values (22) stored in the memory (20), is deleted from the memory (20) and replaced by the determined correction value becomes. A method according to claim 1 or 2, characterized in that the memory (20) is designed as a shift register. Method according to one of the preceding claims, characterized in that the number of supporting values (22) stored in the memory (20) is determined by: - A distance (42) between a printing location of the print mark in the printing unit and a readout position of the sensor unit and / or - A format length (40) of the printing material (16). Method according to one of the preceding claims, characterized in that a plurality of actual positions of respective, by means of the printing unit (14) on respective substrates (16) printed print marks (18) are detected by means of the sensor unit (12), wherein an actual position Mean value is formed from the plurality of actual positions and the difference value is determined by means of the actual position average. Method according to claim 5, characterized in that the number of supporting values (22) stored in the memory (20) corresponds to the sum of a first number and a second number, the first number being determined by: the distance (42) between the printing location and the read-out position and / or the format length (40) of the printing material (16), and the second number corresponds to the number of actual positions used to form the actual position average. Method according to one of the preceding claims, characterized in that the buffer value is determined as the sum of the support values (22). A method according to claim 7, characterized in that, when determining the sum of at least one of the support values (22) is weighted. Method according to one of the preceding claims, characterized in that the sensor unit (12) has a camera sensor and / or a fiber optic. Method according to one of the preceding claims, characterized in that as the printing material (16) paper and / or cardboard and / or plastic and / or metal and / or wood and / or glass is used. Device for correcting a printing position of a printing couple (14) of a printing machine (10), comprising at least one printing device electronic computing device which is designed to: - An actual position of at least one by means of the printing unit (14) printed on a substrate (16) print mark (18) based on sensor data, which are received by a sensor unit (12) to detect; To determine a deviation of the detected actual position from a desired position; - to determine at least one difference value characterizing the deviation; To determine a buffer value from a number of support values (22) previously stored in a memory (20); To determine a correction value by subtracting the buffer value from the difference value; To store the correction value as one of the support values (22) in a memory (20); and To provide a signal for the correction of the printing position based on the determined correction value. Printing machine (10) with a device according to claim 11.

Images