Computer Program Product, Industrial Installation, Method and Apparatus for Determining or Predicting a Position of a Web Break

Abstract

A computer program product, an industrial installation in particular a paper-making machine, an apparatus and method for predicting a position of a web break of a fibrous material web that has occurred or is imminent, wherein the method includes capturing parameters, in particular speeds of rollers for transporting the fibrous material web or a web tension thereof, where the parameters are advantageously stored in the form of time series, a self-learning algorithm is used to detect the imminent web break and to determine the position of the web break which is imminent and/or has occurred, where the basis for the detection or the determination is a deviation of the respective parameter, such as from a temporal mean of the respective parameter.

Claims

1.-17. (canceled)

18. A method for determining and predicting a position of a web break of a fibrous material web which has occurred or is imminent in an industrial installation having a plurality of rollers, the rollers which are rotatably coupled to a motor transport the fibrous material web in the industrial installation, the method comprising first steps of: providing and optionally storing parameters, each parameter being associated with at least one roller; performing a check to determine whether a parameter has a temporal progression which corresponds to a specifiable distribution comprising a normal distribution; performing a check of a further parameter to determine whether the further parameter has a temporal progression which corresponds to the specifiable distribution, if the temporal progression of a respective parameter corresponds to the specifiable distribution; performing a check of the temporal progression of the respective parameter for a first deviation if the temporal progression of the further parameter does not correspond to the specifiable distribution.

19. The method as claimed in claim 18, the method further comprising second steps of: determining a second deviation of at least one parameter or a temporal change of the respective parameter; comparing at least one of (i) the respective parameter and (ii) a respective temporal change of the respective parameter with a specifiable limit value in each case; wherein the first deviation is at least one of (i) recognized and (ii) associated with at least one roller for the web break of the fibrous material web that has occurred or is imminent aided by a learning-capable algorithm.

20. The method as clamed in claim 19, wherein the first deviation of the respective parameter is determined with the first steps and the second deviation is determined with the aid of the second steps; wherein a first position of the web break which is imminent or has occurred is determined aided by the first deviation; and wherein a second position of the web break which has occurred or is imminent is determined aided by the second deviation, a relevant position being displayed to a user if the first position and second positions match.

21. The method as claimed in claim 18, further comprising: determining the position of the web break aided by visual monitoring, if the first position and the second position of a web break which has occurred are different; and providing the determined position of the web break which has occurred to the learning-capable algorithm, such that the learning-capable algorithm is improved based on the position of the web break which has occurred.

22. The method as claimed in claim 19, further comprising: determining the position of the web break aided by visual monitoring, if the first position and the second position of a web break which has occurred are different; and providing the determined position of the web break which has occurred to the learning-capable algorithm, such that the learning-capable algorithm is improved based on the position of the web break which has occurred.

23. The method as claimed in claim 20, further comprising: determining the position of the web break aided by visual monitoring, if the first position and the second position of a web break which has occurred are different; and providing the determined position of the web break which has occurred to the learning-capable algorithm, such that the learning-capable algorithm is improved based on the position of the web break which has occurred.

24. The method as claimed in claim 18, wherein at least one of (i) the parameters, (ii) the determined position of the web break which has occurred, (iii) the respectively determined first deviation and (iv) the respectively determined second deviation are provided to the learning-capable algorithm.

25. The method as claimed in claim 19, wherein at least one of (i) the first steps and (ii) the second steps are executed repeatedly in a periodic manner to ascertain an imminent web break.

26. The method as claimed in claim 19, wherein at least one of (i) the first steps and (ii) the second steps are executed after the web break has occurred to ascertain the position of the web break which has occurred.

27. The method as claimed in claim 26, wherein the respectively stored parameter is investigated in a time region for a first deviation; and wherein the time region is arranged immediately before the web break which has occurred.

28. The method as claimed in claim 18, wherein the deviation of the relevant parameter is determined based on a difference of the relevant parameter respectively from a temporal mean of the respective parameter.

29. The method as claimed in claim 8, wherein the respective parameter is associated with a plurality of rollers; and wherein the respective parameter maps a movement of respective adjacent rollers.

30. The method as claimed in claim 18, wherein a teaching of the learning-capable algorithm occurs based a comparison of the respective parameter in a time region before a web break that has taken place with the respective parameters in a time legion during a normal operation of the industrial installation.

31. The method as claimed in claim 18, wherein the respective parameter comprises at least one of: a rotary speed or a torque of respective roller or a motor coupled to the respective roller; a web tension of the fibrous material web between a first roller and a second roller which are respectively adjacent; a difference in the rotary speeds of two rollers; a supply current or a frequency of the supply current of the motor which is coupled to the respective roller; and a regulating parameter for the respective motor.

32. The method as claimed in claim 18, wherein the first deviation is determined solely after the web break has occurred; and wherein at least one of the first deviation and (ii) the second deviation is determined repeatedly in a periodic manner.

33. The method as claimed in claim 18, wherein at least one first roller is associated with a first group and wherein at least one second roller is associated with a second group; wherein at least one of: at least one of (i) a first deviation and a second deviation of the parameter of the at least one first roller is determined; at least one of (i) a first deviation and (ii) a second deviation of the parameter of the at least one second roller is determined; and at least one of (i) a first deviation and (ii) a second deviation of the parameter of the at least one first roller from the at least one second roller is determined.

34. An apparatus for determining a web break of a fibrous material web which has occurred or is imminent in an industrial installation, the apparatus comprising: at least one parameter acquiror, a respective parameter being established to define a movement of at least one roller; and a display for displaying the web break which has occurs or is imminent; wherein the apparatus is associated with a computer unit; and wherein the computer unit determines the position of the web break which has occurred or is imminent by: providing and optionally storing parameters, each parameter being associated with at least one roller; performing a check to determine whether a parameter has a temporal progression which corresponds to a specifiable distribution comprising a normal distribution; performing a check of a further parameter to determine whether the further parameter has a temporal progression which corresponds to the specifiable distribution, if the temporal progression of a respective parameter corresponds to the specifiable distribution; performing a check of the temporal progression of the respective parameter for a first deviation if the temporal progression of the further parameter does not correspond to the specifiable distribution.

35. A non-transitory computer program product encoded with program code which, when executed on a computer unit, causes determination and prediction of a position of a web break of a fibrous material web which has occurred or is imminent in an industrial installation having a plurality of rollers, the rollers which are rotatably coupled to a motor transport the fibrous material web in the industrial installation wherein the computer program product is configured, when it runs on a computer unit, the program code comprising: program code for providing and optionally storing parameters, each parameter being associated with at least one roller; program code for performing a check to determine whether a parameter has a temporal progression which corresponds to a specifiable distribution comprising a normal distribution; program code for performing a check of a further parameter to determine whether the further parameter has a temporal progression which corresponds to the specifiable distribution, if the temporal progression of a respective parameter corresponds to the specifiable distribution; program code for performing a check of the temporal progression of the respective parameter for a first deviation if the temporal progression of the further parameter does not correspond to the specifiable distribution.

36. An industrial installation comprising a drying section of a paper-making machine or a paper-making machine having the apparatus as claimed in claim 34.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0119] The invention will now be described and explained in greater detail making reference to the drawings. The features shown in the drawings can be combined into new embodiments of the invention. The embodiments illustrated in the drawings are merely exemplary and in no way restrict the invention in which:

[0120] FIG. 1 shows an exemplary industrial installation in accordance with the invention;

[0121] FIG. 2 shows a flowchart of the method in accordance with the invention;

[0122] FIG. 3 shows a further flowchart of the method in accordance with an exemplary embodiment of the invention;

[0123] FIG. 4 shows a further flowchart of the method in accordance with an exemplary embodiment of the invention;

[0124] FIG. 5 shows graphical plots of time series of parameters in accordance with the invention.

DETAILED DESCRIPTION OF THE EXEMPLARY EMBODIMENTS

[0125] FIG. 1 shows an exemplary industrial installation 100. The installation comprises rollers 3, 3′, where the rollers 3, 3′ are configured for transporting a fibrous material web in particular a paper web. The respective roller 3, 3′ is associated, in each case, with a motor 5. The motor 5 serves for rotary movement of the respectively associated roller 3, 3′. Sensors 9 serve to determine the parameters x1, x2, x3, for example, the rotary speed w, the torque D of a roller 3, 3′ or the web tension BS of the fibrous material web 1. The respective acquirer 9 can be configured as a detector or a sensor. The acquirer can be associated with a roller 3, 3′ or a motor 5.

[0126] The industrial installation shown here comprises first rollers 3, where the first rollers are collected together in a first group 4. The industrial installation 100 shown here further comprises second rollers 3′, where the second rollers 3′ are associated with a second group 4.

[0127] Advantageously, the first rollers have, as the parameters x1, x2, x3, substantially the same values. Accordingly, a web break BA of the fibrous material web 1 occurs between respectively adjacent first rollers 3 and second rollers 3′.

[0128] The motors 5 are each connected to a power supply 8. The power supply 8 is advantageously configured as a frequency converter. The power supply 8 is coupled to a control facility 7. The control facility serves to control and/or regulate the torques D and the rotary speed w of the respective roller 3, 3′.

[0129] The control facility 7 is associated with a computer unit RE. The computer unit comprises an interface for receiving the parameters x1, x2, x3. Advantageously, on the basis of the parameters x1, x2, x3, the rotary speeds w of the respective rollers 3, 3′ and/or their torques D are stored in the computer unit RE. Alternatively or additionally, the computer unit RE is associated with a further computer unit. The further computer unit is preferably configured as a cloud and serves for collecting the parameters x1, x2, x3 stored as the time series A, B, C.

[0130] Preferably, the learning-capable algorithm A is installed on the computer unit RE. Preferably, a further computer unit serves for teaching the learning-capable algorithm Alg.

[0131] The parameters x1, x2, x3 can be torques D, rotary speeds w and/or the web tension BS acquired via sensors 9.

[0132] The torque D and/or the rotary speed w is advantageously provided to the power supply for the respective motors.

[0133] Parameters x1, x2, x3 can also be control variables and/or regulating variables which, for example, are provided by the control facility 7.

[0134] FIG. 2 shows an exemplary method. The exemplary method serves for determining the position of a web break BA of a fibrous material web 1.

[0135] In a first step V101, the parameters x1, x2, x3, in particular, the rotary speeds w and the torques D of the respective rollers 3, 3′ are acquired. Optionally, the web tension BS is determined based on rotary speeds w of two adjacent rollers 3, 3′ in each case. Optionally, the web tension BS or another parameter x1, x2, x3 is stored as a time series A, B, C.

[0136] In a second method step V201, the determination of a first deviation Δ1 and/or a second deviation Δ2 is performed. The first deviation Δ1 is determined with the aid of first steps (V102, . . . , V102). The second deviation Δ2 is determined with the aid of second steps (V103, . . . , V603).

[0137] Advantageously, the first steps (V102, . . . , V702) and second steps (V103, . . . , V603) occur together in order to determine the position of a web break that has occurred or is imminent.

[0138] The first and second deviation Δ1, Δ2 is, for example, a difference of a value of a web tension BS from a mean web tension (BS). The respective deviation Δ1, Δ2 can also be a difference between two rotary speeds w, in particular, a rotary speed w of a first roller 3 from the rotary speed w of a second roller 3′. Alternatively, the deviation Δ can be a difference of torques D of a first roller 3 and a second roller 3′.

[0139] In a third method step V301, the determination of the position of the web break BA based on the at least one deviation Δ1, Δ2 advantageously occurs.

[0140] The position of the web break BA is preferably displayed to a user with the aid of a display. Advantageously, the display is configured mobile, for example, as a tablet or a handheld device.

[0141] FIG. 3 shows a further exemplary method. The further exemplary method preferably serves for predicting the position of a web break BA that has taken place or is imminent.

[0142] In the method shown here, first steps (V102, . . . , V102) are performed.

[0143] In a first step V102, the parameters x1, x2, x3 are acquired from the industrial installation 100. The acquisition of the parameters x1, x2, x3 advantageously occurs via sensors 9. With the aid of the parameters x1, x2, x3, the movement of each of the rollers 3, 3′ in the industrial installation 100 is described. Advantageously, the parameters x1, x2, x3 are stored in the form of time series A, B, C. The storing advantageously occurs in a database that is stored on a further computer unit, in particular a cloud.

[0144] In a second step V202, a check is performed to determine whether a parameter x1, x2, x3 has changed ac all. If the parameter x1, x2, x3 has not changed, then a new parameter x1, x2, x3 is selected and checked for a change. This is a test, in particular, of whether the respective parameter x1, x2, x3 corresponds to the specifiable distribution Vert.

[0145] The previously selected parameter x1, x2, x3 is discarded in a third step V302.

[0146] Alternatively or additionally, in the second step V202 the parameter x1, x2, x3 can be investigated as to whether the respective parameter x1, x2, x3 corresponds to a specifiable distribution, for example, a Gaussian normal distribution. Thus, natural variations of the respective parameter x1, x2, x3 can be taken into account.

[0147] On rejection of the respective parameter x1, x2, x3, the investigation of a further parameter x1, x2, x3 occurs with the aid of the second step V202.

[0148] In a fourth step V402, the parameters x1, x2, x3 that have changed over time t and/or that do not correspond to the specifiable distribution, are investigated for first deviations Δ1 and the first deviations Δ1 are determined and stored.

[0149] In a fifth step V502, a check is performed to determine whether the respective first deviations Δ1 are similar in a plurality of rollers 3, 3′ in a group 4, 4′. Herein, the parameters x1, x2, x3 of a first roller 3 are advantageously each compared with the parameters x1, x2, x3 of a second roller 3′.

[0150] Alternatively, a check can be performed to determine whether a deviation Δ1 is ascertainable in a plurality of first rollers 3. In the case of a deviation Δ1 with only a first roller 3 or a single second roller 3′, typically, a measuring error should be assumed.

[0151] In a sixth step V602, on the basis of the first deviations Δ1, it is ascertained with the aid of a learning-capable algorithm Alg, whether a web break BA of the fibrous material web 1 is imminent or has already occurred. Based on an association of the first deviation Δ1 with a roller 3, 3′ of the region between two adjacently arranged groups 4, 4′, it is possible to determine where a web break BA will occur or has occurred.

[0152] In the case of an imminent web break BA, a warning signal is advantageously issued. Advantageously, the expected position of the web break BA is also communicated to the receiver of the warning signal.

[0153] Advantageously, the user provides information regarding whether the displayed position of the web break BA was correct and/or whether a web break BA has actually occurred.

[0154] In an optional seventh step V702, a teaching of the learning-capable algorithm Alg advantageously occurs. The learning-capable algorithm Alg is caught based on a comparison of the parameters x1, x2, x3 with the web break BA that has occurred. With the teaching, the detection of the web break BA of the learning-capable algorithm Alg can be improved.

[0155] Aside from the first deviations Δ1 and the respective position of the web break BA that has occurred or is imminent, the time series A, B, C of the respective parameter x1, x2, x3 are provided. By providing the parameters x1, x2, x3 and, in particular, the position of the web break BA of the fibrous material web 1 that has occurred, the learning-capable algorithm can be taught better and patterns and/or peaks can be more easily detected with a high degree of reliability.

[0156] FIG. 4 shows a further exemplary method. In this figure, two steps (V103, . . . , V603) are shown. In a first step V103, a start signal occurs. The start signal can be the occurrence of a web break BA. Alternatively or additionally, the start signal can represent a regularly repeating time signal.

[0157] In a second step V203, a reading-in of the time series A, B, C and therewith of the parameters x1, x2, x3 occurs. The time series A, B, C are typically stored parameters x1, x2, x3 as a function of time t.

[0158] In a third step V303, based on parameters provided x1, x2, x3, a determination of the respective web tensions BS of the fibrous material web 1 occurs. The web tension BS is associated in each case with a region of the fibrous material web 1 between two adjacent rollers 3, 3′ in each case and/or with a region between two groups 4, 4′ of rollers 3, 3′.

[0159] Alternatively or additionally, the regions of the fibrous material web 1 in which the web tension BS is severely reduced are determined.

[0160] Optionally, time points are determined at which at least one parameter x1, x2, x3 is subject to a significant change.

[0161] In a fourth step V403, whether there are time regions dt in which a plurality of parameters x1, x2, x3 show a corresponding behavior pattern is determined.

[0162] A behavior is, in particular, the occurrence of a second deviation Δ2, where the second deviation Δ2 is a deviation of the respective parameter x1, x2, x3 from the respective temporal mean <x1>, <x2>, <x3>of the parameter x1, x2, x3. The web tension BS between two rollers 3, 3′ each serve as very suitable parameters x1, x2, x3.

[0163] Based the behavior pattern, it is possible to ascertain whether and where a web break BA of the fibrous material web 1 has occurred and is imminent.

[0164] In a fifth step V503, the indication to a user of the position of the web break BA of the fibrous material web 1 that has occurred or is imminent is provided.

[0165] In an optional sixth step V603, the teaching of a learning-capable algorithm Alg based on the position of the web break BA that has been ascertained or is to be expected is performed. Advantageously, a user provides indications as to whether and/or at which position in the industrial installation 100 the web break BA has occurred.

[0166] FIG. 5 shows graphical plots of time series A, B, C. A first time series A shows a variation of a parameter x1, x2, x3 as a function of time t. At one time point, a brief rise in the parameter x1, x2, x3 is visible. At no other time region dt does the parameter x1, x2, x3 exceed the specifiable limit value CR. In addition, a specifiable distribution Vert of the parameters x1, x2, x3 is assigned in the respective time series A, B, C. Such a deviation can be a first deviation Δ1.

[0167] In particular, if only one single parameter x1, x2, x3, for example, the rotary speed w of a roller 3, 3′ or a torque D of a roller 3, 3′ goes beyond the limit value GR, in general, a measurement error can be assumed. Advantageously, a taught learning capable algorithm Alg can detect this.

[0168] Also shown is a specifiable distribution Vert, according to which the noise of the respective parameter x1, x2, x3 is describable.

[0169] The second time series B shows a variation which accords with a web break BA. A linear drift of one parameter x1, x2, x3 is characteristic of a web break BA. A drift can also be non-linear. The drift occurs in the time region dt beyond the limit value GB. The second deviation Δ2 increases during the drift. Such a parameter x1, x2, x3 can be the web tension BS of the fibrous material web 1 between two groups 4, 4′.

[0170] In the further course, a fall occurs in the parameter x1, x2, x3 shown here. This fall is a characteristic sign of a web break BA and this behavior is advantageously recognized by a learning-capable algorithm Alg and notified to the user.

[0171] The third time series C shows a variation of a parameter x1, x2, x3 over time t. Apart from slight deviations, the respective parameter x1, x2, x3 corresponds to the respective mean value <x1>, <x2>, <x3>.

[0172] Summarizing, the invention relates to a method and an apparatus, in particular, for predicting a position of a web break BA of a fibrous material web 1 that has taken place or is imminent. The invention further relates to a computer program product and to an industrial installation 100, in particular a paper-making machine. The method comprises the acquisition of parameters x1, x2, x3, in particular, rotary speeds w of rollers 3, 3′ for the transport of the fibrous material web 1 or a web tension ES thereof. Advantageously, the parameters x1, x2, x3 are stored as time series A, B, C. A self-learning algorithm Alg serves to detect the web break BA that is imminent and for determining the position of the web break BA that has occurred and/or is imminent. The basis for the detection or the determination is a deviation Δ of the respective parameter x1, x2, x3, for example, from a temporal mean <x1>, <x2>, <x3>of the respective parameter x1, x2, x3.

[0173] Thus, while there have been shown, described and pointed out fundamental novel features of the invention as applied to a preferred embodiment thereof, it will be understood that various omissions and substitutions and changes in the form and details of the methods described and the devices illustrated, and in their operation, may be made by those skilled in the art without departing from the spirit of the invention. For example, it is expressly intended that all combinations of those elements and/or method steps which perform substantially the same function in substantially the same way to achieve the same results are within the scope of the invention. Moreover, it should be recognized that structures and/or elements and/or method steps shown and/or described in connection with any disclosed form or embodiment of the invention may be incorporated in any other disclosed or described or suggested form or embodiment as a general matter of design choice. It is the intention, therefore, to be limited only as indicated by the scope of the claims appended hereto.