JPH05229106A - Method for monitor and control of printing process - Google Patents

Abstract

PURPOSE: To achieve a target quality by an actual cause in such a way that an assignable vector to a class set by classification in a wide sense or a cause of the printing difficulty is assigned (learning phase), and assignability of a class is determined and indicated through a learning phase from features obtained by measurement technology of printing matters. CONSTITUTION: Point corresponding to a measured value in a characteristic space of e.g. full tone ink concentration DV and screen tone ink concentration (40%) DR40 is obtained by printing monitoring stripes printed together on sheets by concentration measurement, for each ink. Assingable N class (assingable vector) and N numerical values are assigned for each point. For example, in four classes (k1 to k4 ), an assignable vector is provided with four components, and the numerical size of this component indicates how strong point is assignable to characteristic space of a required class. In a printing process, to what class a printing matter is assignable is detected by the position of point in characteristic space obtained by a learning phase regarding a value detected by measurement technology.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for monitoring and controlling a printing process in, for example, an offset printing machine, which detects a characteristic representing the quality of a printed matter by a measuring technique and controls the printing process. Relatedly, the present invention relates to a printing process monitoring and controlling method for adjusting and operating the printing process so that the printed matter approaches a target quality.

[0002]

2. Description of the Related Art The production of printed matter on an offset printing machine, for example a sheet-fed offset printing machine, is a complicated process because of the large number of interdependent control variables. In the case of offset printing, the quality of the printed matter is affected by variables such as ink supply, wetting agent supply, type of printing material, type of ink and rubber bracket, and the like. Quality is to be understood here as the maintenance of a certain minimum demand on the printed matter. This is, for example, color reproducibility, sharpness and contrast. Due to the complexity of the offset printing process, complete control is theoretically possible only at the high cost of the process analysis. However, at present, no specific form of such multivariate control is known.

Instead of multivariate control, so-called partial control circuits, such as control circuits for ink supply, control circuits for wetting agent supply, control circuits for register control, etc., are currently implemented. In that case, for example, one control circuit is used as the main control circuit, and the other sub-control circuits are used only for holding the other control variables at the target amount once found (DE 3717904 A1). The interdependence of the control variables is eliminated by always controlling the quality error occurring in the printed matter by the main control circuit.

It has been known for a long time in practice to supply the ink in the main control circuit. That is, the ink application (inking) of the printed matter is detected by a sensor, and the ink supply is controlled via a controller and a corresponding adjusting element so that the target ink inking is achieved. At that time, the detection of the printing ink application by the sensor is generally performed photoelectrically. That is, by optical ink density (densitometer), or by detection of color location (colorimeter), or using video techniques including image processing. Apparatus and method for inking detection and ink flow control are described in DE2.
728738A1, EP228347 B1, DE34
40706 A1, DD274786A1 and DD2
74787 A1. The target value on which the control of the ink supply is carried out can be taken from the original, for example by measuring techniques, or can be found experimentally, or can be taken from the handbook for standardization of the printing process (BVD-Fogra). An example of a sub-control circuit for the wetting agent supply is given in DE 3636507 A1.

However, the above-mentioned printing process control has a drawback in that the main control is performed by ink supply and any error in the quality of the printed matter is controlled by ink supply. In offset printing, for example, the increase in tone value is an important variable that determines the reproduction quality of printing. The optical hiding power achieved with printing is significantly greater or lesser in multiple tone value steps, which impairs the contrast range for monochromatic printing and color reproducibility for multicolored printing. .. Means for controlling the ink supply mainly in relation to the degree of surface hiding are also known (DE 3543444).
A1, DE 3440706 A1). Here again, however, any changes in the degree of surface hiding achieved in printing are sought to be obtained by changes in the ink supply. For example, in general, increasing the ink supply can likewise increase the optically active surface hiding degree (increasing the halftone dots), which likewise reduces the wetting agent supply, or plate cylinders and rubber brackets, for example. It is known that it can also be obtained by increasing the pressure applied between the cylinders. For example, it would be pointless to try to reduce the excessively large surface hiding by reducing the ink supply, even though the color density of the full-tone measuring field is in the approximate target range. In this case, rather, the wetting agent supply or the previously mentioned cylinder pressure should be changed. The type of material to be printed, the type of printing bracket, the properties of the printing ink, the type and amount of so-called printing aids, the machine temperature and the subject matter are also important factors. Even if the control of the tone value rise in one print is achieved, for example, particularly well by a change in the cylinder pressure, another print (another subject matter,
For paper etc.) the tone value rise can be best controlled by varying the wetting agent supply. The decision of what is the optimum correction means is currently left to the experienced printer.
With his knowledge and the ability of the existing printing process to make decisions, the printer finds the control variables that, if changed, can most effectively limit the tone value rise.

[0006]

SUMMARY OF THE INVENTION An object of the present invention is to provide a method for monitoring and controlling a printing process according to the preamble of claim 1, which allows the quality deviation of printed matter to be prevented. The optimal control variable, for example, the actual cause, can be controlled to reach the target quality.

[0007]

According to the present invention, according to the present invention, a point in a feature space formed by a measurement technical feature is determined according to a format of a broad sense from a printing trial result evaluated by a process technique. By assigning a attribution vector to a cause for preset grade and / or printing difficulty (quality defect) (learning phase), and monitoring the printing process, the characteristics obtained by the measurement technique on the printed matter, And the corresponding attribution vector of the point in the feature space, the attribution of the printed matter to a predetermined grade or cause is determined, and the attribution to the grade is indicated based on this, and / or the printing process is Thus, it is configured and solved so as to perform an adjusting operation (execution phase) through the cause thus detected.

Embodiments of the invention are indicated in the dependent claims, the description and the drawings.

A broadly-known classification method known per se is applied, that is, a classifier in a broad sense is applied from the characteristics representing the quality condition of the existing printed matter, and which quality deviation is detected on the one hand, and which is the other. It is detected whether the quality deviation can be removed according to the control variable. Condition monitoring using a broad classification method is known from DD251534 A1.
In contrast to the means of controlling obstacles exclusively in the printing process by means of inking, the method according to the invention allows the controller associated with the broader classification method to eliminate the obstacles via a pseudo most probable cause. Has the advantage. This can be compared to some extent with human behavior. This is because a skilled printer removes obstacles in the printing process through what appear to him to be the most probable cause.

[0010]

Embodiments of the present invention will be described in detail with reference to the drawings.

The sheets are printed on a sheet offset printing press and subsequently subjected to a mechanical quality check. The densitometer measures the print monitoring stripe (its measurement field) printed together on the sheet (inspection sheet). Here, the print monitoring stripes include, for each printing ink, a measurement field for full-tone ink density DV, a screen field for screen tone ink density (80%) DR80 and a screen field for screen tone ink density (40%) DR40, and It also has a quality diagnostic field, eg a gray balance, and possibly further shift and blur fields. An example of this type of print monitoring stripe is the MAN Rand Druckmasch.
It is a 6-color CCI print monitoring stripe of inen AG. Therefore, by scanning the print monitoring stripe with the densitometer, the characteristics of each ink, the full-tone ink density D
V, screen tone ink density (40%) DR40, screen tone ink density (80%) DR80 are obtained. In this way, the feature space is formed. The components of this space are formed by full tone ink density DV, screen tone ink density (40%) DR40 and screen tone ink density (80%) DR80.
A densitometer measurement of the print monitoring stripes of the printed matter gives points for each ink in this feature space corresponding to the measured values. When the number of printing inks is large, a large number of points are obtained corresponding to the number of printing inks.

The skilled person, based on his experience, determines the grade (quality grade) of interest in the printing process. This includes, for example, class K1 “increase in tone”, class K2 “excessive increase in tone value over standard”, class K3 “excessive increase in tone value over standard”, class K4 “standard”. Characteristic curve ”. Number of grades (here K1 to K
Corresponding to 4), a so-called membership N-tuple is assigned to each point in the feature space. According to the 4 grades in the above example, 4 numerical values will be assigned to each point. This number can be standardized to be between 0 and 1. The membership N-tuple (in the case of N grade) can also be referred to as a so-called membership vector, and will be referred to as such. The numerical magnitude of the membership vector component indicates how “strongly” the points belong to the feature space of a given grade. In the case of the 4th grade, the membership vector has four components, the first component is the degree of membership in the class K1 of the point, and the second component is the class K2 of the point.
Indicates the degree of attribution to, and so on. The components of the attribution vector of the individual points in the feature space are then defined in a so-called learning phase by a process engineering evaluation of the results of a number of printing trials. When the component of the membership vector is standardized as a numerical value between 0 and 1 as described above, for example, a component having a numerical value close to 0 indicates that the membership to a corresponding class is very small, and a numerical value close to 1 is suitable. Then, it shows that the belonging to each class is very strong. If the points in the feature space are thus occupied by the attribution vector, it is possible to detect which class the print belongs to (execution phase) after scanning the print monitoring stripe with the densitometer.

In FIG. 1, components, full-tone ink density DV, screen tone ink density (40%) D
The plane formed by R40 is shown. Furthermore, on this plane, the classes K1 to K4 according to the above definition are lotted. Shown is a line of constant attribution for one class (association = 0.5).
Therefore, the attribution is greater than 0.5 within the line of the illustrated class. Surprisingly, in terms of process technology, the graded regions thus formed can overlap one another.
Also, the line with attribution = 0.5 does not have to surround a closed area.

The present invention further assigns another membership vector to a point in the feature space. This attribution vector represents the attribution of each point to the cause of negatively affecting the print quality. Fig. 2 shows coordinates and full-tone ink density D
V, a plane formed by a screen tone ink density (40%) DR40 is shown. Again, the line of attribution = 0.5 is shown for causes U1 to U3. Here, for example, U1 stands for "wrong ink supply", U2 for "wrong pressure application" and U3 for "wrong wetting agent supply". Also here, the assignment of the attribution vector of the feature space (in this example, the attribution vector has three components. This is because three causes U1 to U3 are set). It is carried out by process technology evaluation. The components of the membership vector can again be standardized between 0 and 1.

After measuring the printed matter with a densitometer, the positions of the characteristic points of this printed matter in the feature space are known on the basis of it. By occupying the points in the feature space made in the learning phase with the membership vector, based on the position of the points in the feature space detected by the measurement technique,
The attribute of the printed matter to a predetermined grade and the attribute of the printed matter to the cause of the quality failure are detected. This is obtained from the maximum value, ie the maximum value of the components of each membership vector. At that time, for example, class K1
In the case of to K4 attribution, the following examples can occur. .

Attribution to a class such as K3 is clear. This is because the attribution vector of this feature point in the third component is high, and the other components have very small values.

The membership vector already has a maximum value. That is, they have approximately the same large membership value in two or more components. This means that the printed sheet or its quality cannot be definitely attributed to one class or cause.

In the first case, ie when the attribution to the class or the cause is clear, the appropriate measures can be taken directly to eliminate this defect. If the characteristics of the printed matter, for example, clearly show the attribution to the cause “wrong wetting agent supply is wrong”, for example a mechanical detection of the wetting agent supply is carried out, on the basis of which the wetting agent supply can be defective in quality (for example (Increased power) is adjusted to a new value.
In the second case, where attribution to a class or cause is unclear, multiple measures are used. Once, it would be possible to evenly indicate the means and let the printer decide. Similarly, the choice regarding the cause could be determined by chance and the defect removed via this cause accordingly. Therefore, it is possible to divide the cause U1 "wrong ink supply" into several sub-causes. At that time, each sub-cause is assigned a change necessary for removing the wrong ink supply (for example, a necessary ink density change). Therefore, this measure should be performed in the learning phase.

In summary, the method of the invention in its application is as follows.

Learning Phase-Attribution vector indicating the attribution to a predetermined class of printed matter or the cause of printing trouble is assigned to a point in the feature space. At that time, the printing trial result evaluated by the process technology is evaluated.

Execution phase-the measurement technical result of the measurement field, for example by a densitometer, represents the point in the feature space for the print. As a result, the assignment of the printed matter (actual quality) from the corresponding attribution vector (maximum component) to the set grade or the cause of the printing fault (printing defect) is performed. Based on this, the print fault is removed through the cause thus detected.

Up to this point, the method of the present invention has been described based on the feature space. The axes of this feature space are formed exclusively by measuring technical (densitometer) quantities. However, as is known, the type of printing material also has a great influence on the print quality achieved. Therefore, in the learning phase, it is also possible to assign corresponding attribution vectors in the feature space (formed by the measurement technical quantity) for different printing materials that produce the same printing-technical properties. Therefore, the feature space for the strongly absorbent printing material has, besides the feature space for glossy or very smooth paper types, the occupancy of the attribution vector for a given class and cause. The same can be envisaged, for example for different rubber bracket types. In general, the choice of material to be printed and the type of rubber bracket are made once before the start of inking, so during the execution phase of the method according to the invention a corresponding feature space is determined, including the occupancy of membership by the attribution vector. , Applied.

For example, specially constructed quality diagnostic fields (shift fields, blur fields) can also be evaluated video-wise when applying the method of the invention. The measured values thus formed and processed by the image processing method (point increase, geometrical position of halftone dots, linear processing) can form one or more axes of the feature space. At that time, the attribution vector is assigned to the point in the feature space according to the print trial result in accordance with the set class or cause. Full detection of printed matter is likewise conceivable. The class to be set by the expert can be defined in the RGB color space of the camera by a predetermined contrast histogram. Predetermined target values (eg color density values) can be set for the ink supply, depending on which grade the printed matter belongs to most strongly.

The method of the invention can advantageously be implemented in a computing device. The calculation device is connected to a color density measuring device, a measuring device operating in colorimetric or video technology. After the measurements have been made on the prints via the assigned image screen device, the execution phase of the invention indicates the cause of the problem in relation to the control to be recommended for eliminating the print disturbance. .. The computer for carrying out the method of the invention can also be directly connected to the corresponding adjusting means of the printing machine via a special interface and bus system. Depending on the cause required for printing difficulty,
For example, the print adjustment can be changed automatically, the wetting agent supply can be corrected or the adjusting element for the ink supply can be activated.

[0025]

According to the monitoring and controlling method of the present invention,
The quality deviation of the printed matter can be controlled to reach the target quality by an optimum control variable, for example, the actual cause.

[Brief description of drawings]

FIG. 1 is a diagram showing a part of a feature space for explaining the present invention.

FIG. 2 is a diagram showing a part of a feature space for explaining the present invention.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Arno Vuehl Mühlheim Main Waldheimer Strasse 26

Claims (3)

[Claims] 1. A method for monitoring and controlling a printing process, for example in an offset printing press, characterized in that a characteristic representative of the quality of the printed matter is detected by measuring technology and the printed matter is associated with a variable controlling the printing process. Adjust and operate the printing process to approach the target quality,
In a method for monitoring and controlling a printing process, according to a result of a printing trial evaluated by a process technology, a point in a feature space formed by a measurement technical feature is assigned a predetermined preset grade and / Or, a attribution vector to the cause of printing difficulty (quality defect) is assigned (learning phase), and when monitoring the printing process, from the characteristics obtained by measurement technology in the printed matter, and at the relevant point in the characteristic space. The attribution of the printed matter to a given grade or cause via the corresponding attribution vector of, and based on which the attribution to the grade is indicated and / or the printing process is detected in this way A method for monitoring and controlling a printing process, which is characterized by performing an adjusting operation (execution phase) through a cause. 2. A method according to claim 1, wherein the coordinates of the feature space are formed by a full tone ink density value (DV) and a screen tone value defining at least one screen tone ink density (DR). 3. A method as claimed in claim 1, in which differently printed substrates each having a similar printing characteristic are assigned a correspondingly assigned feature space of the attribution vector.