• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The invention relates to a method for detecting the normal operating state of a working process, in which a machine (M ') is operated in a working environment with external specifications, wherein the machine (M') sensors (S'1, ..., S'm ) for measuring physical parameters at predetermined positions in the area of the machine M ', wherein - a standard process is started with a standard machine (M1) in the standard environment with external standard settings, - the standard machine (M1) being equipped with a number of sensors (S1, ..., Sn), - with the sensors (S1, ..., Sn) of the standard machine (M1) measured values are repeatedly determined at predetermined times, and - according to rules (R1, ..., Rp) is searched which describe the relationship of the measured values determined during the normal operating state of the standard machine (M1) in the relevant working environment with the external specifications, and which are fulfilled during the normal operating state. The invention provides that subsequently - the rules (R1, ..., Rp) determined for the standard process are transferred to the working process of a machine (M ') similar or identical to the standard machine (M1) in the work environment, - the machine (M ') has a number of sensors (S'1, ..., S'm) of the machine (M'), whose position and sensitivity with the on the standard machine (M1) located sensors (S1, ..., Sn ) at least partially coincides and - with the sensors (S'1, ..., S'm) of the machine (M ') repeatedly measured values are determined at predetermined times, - it is examined whether the transmitted rules (R1, ... , Rp) by the measured values of those sensors (S'1, ..., S'm) of the machine (M '), which with the sensors (S1, ..., Sn) of the standard machine (M1) with regard to arrangement and sensitivity are met, and, if so, the rules for characterizing the normal operating state of the work process are adopted, and, if not, adapt the rules to the new measurements or discard them.
    公开号:AT519777A1
    申请号:T50233/2017
    申请日:2017-03-22
    公开日:2018-10-15
    发明作者:Wurzenberger Markus；Skopik Florian
    申请人:Ait Austrian Institute Tech Gmbh；
    IPC主号:
    专利说明:

    (19) Austrian Patent Office do) AT519777A1 2018-10-15 (12) Austrian patent application (21) Application number:
    (22) Filing date:
    (43) Published on:
    A 50233/2017
    03/22/2017
    15.10.2018 (51) Int. CI .:
    G05B 13/04
    G05B 17/02
    G05B 23/02
    G06Q 10/06 (2006.01) (2006.01) (2006.01) (2012.01) (56) Citations:
    EP 2523150 A1
    US 9092802 B1
    DE 4436658 A1
    US 2009093892 A1 (71) patent applicant:
    AIT AUSTRIAN INSTITUTE OF TECHNOLOGY GMBH
    1220 WIEN (AT) (72) Inventor:
    Wurzenberger Markus
    1100 Vienna (AT)
    Skopik Florian
    2000 Stockerau (AT) (74) representative:
    Wildhack & Jellinek Patentanwälte OG
    1030 Vienna (AT)
    AT 519777 A1 2018-10-15 (54) Method for recognizing the normal operating state of a work process (57) The invention relates to a method for identifying the normal operating state of a work process, in which a machine (M ') is operated in a work environment with external specifications The machine (M ') has sensors (57, ..., S' m ) for measuring physical parameters at predetermined positions in the area of the machine M ', wherein
    - a standard process is started with a standard machine (M ^ in the standard environment with external standard specifications,
    the standard machine (M ^ has a number of sensors (St, ..., S n ),
    - With the sensors (St, ..., S n ) of the standard machine (Mt), measured values are repeatedly determined at predetermined times, and
    - rules (Ri, ..., R p ) are searched which describe the relationship of the measured values determined during the normal operating state of the standard machine (M ^ in the relevant working environment with the external specifications, and which are fulfilled during the normal operating state,
    The invention provides that subsequently
    - the rules (Rt, ..., R p ) determined for the standard process are transferred to the work process of a machine (Mß similar or identical (M ') in the work environment),
    - wherein the machine (M ') has a number of sensors (S't,. S' m ) of the machine (M '), the position and sensitivity of which are on the standard machine (M ^) sensors (St, .. ., S n ) at least partially matches and
    - With the sensors (S't, ..., S ' m ) of the machine (M'), measured values are repeatedly determined at predetermined times,
    - It is examined whether the transmitted rules (Ri, ..., R p ) are determined by the measured values of those sensors (S't, ..., S ' m ) of the machine (M') that are connected to the sensors (St, ..., S n ) of the standard machine (M1) match in terms of arrangement and sensitivity, are met, and
    - if this is the case, the rules for characterizing the normal operating state of the work process are adopted, and
    - if this is not the case, the rules are adapted to the new measured values or rejected.
    Fig. 1
    DVR 0078018
    Summary:
    The invention relates to a method for detecting the normal operating state of a work process, in which a machine (M ') is operated in a work environment with external specifications, the machine (M') sensors (S'1, ..., S ' m ) for measuring physical parameters at predetermined positions in the area of the machine M ', wherein
    a standard process is started with a standard machine (M1) in the standard environment with external standard specifications,
    the standard machine (M1) has a number of sensors (S1, ..., S n ),
    - With the sensors (S1, ..., S n ) of the standard machine (M1), measured values are repeatedly determined at predetermined times, and
    - rules (R1, ..., Rp) are searched which describe the relationship of the measured values determined during the normal operating state of the standard machine (M1) in the relevant working environment with the external requirements, and which are fulfilled during the normal operating state,
    The invention provides that subsequently
    - the rules (R1, ..., R p ) determined for the standard process are transferred to the work process of a machine (M ') similar or identical to the standard machine (M1) in the work environment,
    - The machine (M ') has a number of sensors (S'1, ..., S' m ) of the machine (M '), whose position and sensitivity with the sensors (S1.) located on the standard machine (M1) , ..., Sn) at least partially matches and
    - With the sensors (S'1, ..., S'm) of the machine (M '), measured values are repeatedly determined at predetermined times,
    - It is examined whether the transmitted rules (R1, ..., Rp) are determined by the measured values of those sensors (S'1, ..., S'm) of the machine (M ') that are connected to the sensors (S1 ,. .., Sn) of the standard machine (M1) match in terms of arrangement and sensitivity, are met, and
    - if this is the case, the rules for characterizing the normal operating state of the work process are adopted, and
    - if this is not the case, the rules are adapted to the new measured values or rejected.
    / 26
    The invention relates to a method for detecting the normal operating state of a work process according to the preamble of claim 1.
    The steadily increasing digitization or automation of industrial processes accelerates and simplifies production processes and manufacturing, but makes maintenance of machines and production lines, for example, increasingly difficult or complex due to the increasing complexity. The increasing number of sensors per meter of a production line makes it impossible to manually record and assess all available data. For this reason, the focus is increasingly on the development of methods for predictive maintenance, for example, which help to optimize production processes and maintenance and to identify impending errors before failures occur. On the other hand, such methods can also be used to optimize production processes, since anomalies can also be caused by suboptimal process configurations that offer scope for optimizing machine operation.
    From the prior art, for example from Hartung, Joachim, Bärbel Elpelt, and Karl-Heinz Klösener. Statistics: teaching and handbook of applied statistics. Walter de Gruyter, 2009 and Murphy, Kevin P. Machine learning: a probabilistic perspective. MIT press, 2012, a variety of methods from the fields of statistics and machine learning are known, which can be used for example for process and maintenance optimization. In the case of such methods from the prior art, characteristics of a normal operating state, also referred to as ground truth, are first determined. These describe the normal operating state of a machine during a work process on the basis of relationships between measured values determined during the normal operating state and are valid during the normal operating state. Deviations from the ground truth are recognized as abnormal conditions of the machine by such a method and interpreted as indications of defective machine components or incorrect configurations of the working process of the machine.
    In the case of methods known from the prior art, however, it is assumed that at the latest when a machine is started up, it is known, for example, in which component of the machine which errors are to be identified and by measuring which physical quantities these can be identified, usually only individual machine components or component groups and not relationships that affect the entire machine. The transfer of a method successfully used / 26 to a machine of the same type, which has minor structural changes or is used in another area of application, has proven to be almost impossible and individual adjustments are necessary in order to laboriously readjust the methods. Another problem is the fact that in most cases the same machine is used to produce different products and orders and that the raw material used does not always correspond to the same quality and can therefore lead to variability in the production process. The same machine can also be used in countries in different geographical latitudes, which means that machines of the same type are inevitably exposed to different environmental and climatic influences, making the normal operating state of a machine difficult to grasp.
    It is therefore the object of the present invention to solve these problems and to provide a method in which rules for the normal operating state of machines are set up during the respective work processes and the rules automatically for different orders, for example, which are carried out on the same machine, different environmental conditions to which the machines are exposed or can also be transferred and adapted to other, largely identical machines.
    These objects are achieved by the characterizing features of claim 1.
    The invention provides that the rules determined for the standard process are then transferred to the work process of a machine similar or identical to the standard machine in the work environment, the machine having a number of sensors of the machine, the position and sensitivity of which are the same as those on the machine Sensors at least partially coincide with the standard machine and measured values are repeatedly determined with the sensors of the machine at predetermined points in time, it is examined whether the transmitted rules are met by the measured values of those sensors of the machine which match the sensors of the standard machine with regard to arrangement and sensitivity, and if this is the case, the rules for characterizing the normal operating state of the work process are adopted, and if this is not the case, the rules are adapted to the new measured values or are rejected.
    In order not to make the number of rules transmitted unnecessarily high, it is advantageously provided that it is examined whether sensors are present in the standard machine that are not matched by the sensor in terms of arrangement and sensitivity
    Machine can be assigned, and if there are no sensors that can be assigned to the machine, the rules that apply to these sensors in the standard machine are deleted.
    In the event that the same physical quantity is determined on the standard machine and the machine with sensors with different functions, it is advantageously provided that it is examined whether sensors of the machine which are assigned to one another have the same characteristic curve with regard to the physical quantity to be determined, with each other in the case assigned sensors of the machine with a different characteristic curve, the rules that are valid for these sensors in the standard machine are adapted to the new characteristic curve of the machine.
    A particularly flexible application of the invention is ensured by examining whether there are sensors in the machine to which no sensor of the standard machine can be assigned in terms of arrangement and sensitivity, and that new rules are established for the measured values of the sensors of the machine determined in this way characterize the normal operating state, in particular these rules being created in the same way that are also used for the creation of the rules for the measured values of the standard process, these rules in particular also being able to take into account measured values from sensors both on the machine and are present on the standard machine.
    A method according to the invention can advantageously also be used if the machine is created by modifying the standard machine or a further standard machine that is identical to the standard machine, by adding components and / or sensors to the standard machine, and / or removing components and / or sensors from the standard machine are, and / or components of the standard machine are modified, and / or sensors of the standard machine are repositioned or changed in terms of their sensitivity or calibration.
    For an application of the invention to machines in, for example, different geographical latitudes and climates, it is advantageously provided that the work process is defined in which the machine or the standard machine is operated in a work environment that differs from the standard environment, the work environment being different from the standard environment in particular the following environmental parameters:
    a) ambient temperature, and / or / 26
    b) ambient air humidity, and / or
    c) ambient air pressure, ...
    A method according to the invention can advantageously also be transferred to machines after a change in an order or work process, by checking whether the machine has external specifications that at least partially match the external standard specifications of the standard machine and that it is examined whether the transmitted rules by the external specifications of the machine, which correspond to the external standard specifications of the standard machine, and, if this is the case, the rules for characterizing the normal operating state of the work process are adopted, and if this is not the case, the rules to the adjusted or rejected new external specifications or new rules are set for these new external specifications.
    To detect suboptimal process configurations or impending errors before failures occur, the invention provides that the machine is operated in the work environment, that measured values are determined during the work process at predetermined, recurring times, that the rules determined for the work process apply to the Measured values are applied, and if there are measured values that do not meet the rules for the normal operating state, an abnormal state of the machine and / or the work process is identified.
    A further advantageous embodiment of the invention provides that rules are determined on at least two standard machines, which describe the relationship of the measured values determined during the normal operating state of the standard machines in the relevant working environments with the external specifications, and which are fulfilled during the normal operating states, and wherein the rules determined for at least two standard machines are transferred to the work process of a machine similar or identical to the standard machines in the work environment.
    In order to be able to use rule sets of various standard machines and to ensure a simple transfer of rules, it is provided that rule sets, in particular comprising a number of determined rules, are stored in a central switching point, the rule sets showing the relationships between the measured values determined during the normal operating states of a large number from / 26
    Describe standard machines in the relevant work environment with the external specifications and which are fulfilled during normal operating conditions, whereby suitable control sets are selected for a work process with external specifications in a work environment of a machine similar or identical to the standard machines, the control sets according to the sensors and / or the environmental parameters of the working environment and / or the external specifications of the machine are selected, and the selected rule sets are transferred to the working process of the machine in the working environment.
    A program for carrying out a method according to the invention on a machine can advantageously be stored on a data carrier.
    Further advantages and refinements of the invention result from the description and the accompanying drawings. The invention is illustrated below on the basis of particularly advantageous, but not restrictive, exemplary embodiments and is described by way of example with reference to the drawings.
    A particularly advantageous embodiment of a method according to the invention is shown in more detail in connection with FIGS. 1 and 2.
    1 schematically shows a machine M1 which has a feed unit 11 on which there are three sensors S1, S 2 , S3, the feed unit 11 having, for example, a first conveyor belt section 12 with a printing unit 13 is connected, with a sensor S4 on the first conveyor belt section 12 and a further sensor S5 on the printing unit 13. This machine M1 will later be used as a standard machine in the exemplary embodiment shown. The printing unit 13 is connected to a punching unit 15 via a second conveyor belt section 14, a sensor S6 being located on the second conveyor belt section 14 and a sensor S7 being located on the punching unit 15. The punching unit 15 is finally connected via a third conveyor belt section 16 to an ejection unit 17, a sensor S8 being located on the third conveyor belt section 16. A machine M1 constructed in this way can be used, for example, for printing and punching corrugated cardboard. In the example, the machine M1 is driven by a V-belt and the printing unit 13 comprises a fixing roller.
    / 26
    The following procedure is usually used by such methods for determining the normal operating state of a machine M1 as ground truth:
    First, a parameterization is carried out, with the machine M1 in question being assigned a number of n attributes. In the present exemplary embodiment of the invention, the individual attributes are provided with an attribute identification i in the form of a predetermined natural numbers, for example in the number range between 1 and n.
    In the following, any point in time chosen is designated by t, such a point in time t being selected from a set {T o , T1, T 2 , ...}.
    At a predetermined number of discrete times t, an attribute value x 'is set for each individual attribute with the attribute identification i during the operation of the machine M1. The points in time are designated starting from an initial point in time T o up to the point in time T, the last acquisition of attribute values. According to the attribute identification i used in each case, the individual concrete values available for the time t in question are entered in a line vector m '= [χ1 *, ... x n '] assigned to the time t.
    The individual line vectors m 'are grouped together in the order in which they were recorded to form a machine attribute matrix A M , where Τ' denotes the time of the last recording.
    On =
    An attribute can be a measured value that was created by a sensor assigned to machine M1. In the present exemplary embodiment of the invention, sensors 51,... 58 repeatedly determine physical measured values at predetermined times t. The attribute identifications i = 1..8 are assigned the measured values determined by the sensors S1,... S 8 , which are explained in more detail in Table 1, ie the measured values determined by the sensors are transferred to the columns 1 at the relevant times t = 1 ..8 entered in the relevant line vector m '.
    Table 1: Sensors, physical measured values and measured value units of machine M1.
    7.26
    Si Speed (min -1 ) Motor of the machine drives (roller r = 0.1 m) S 2 Limit switch (binary) Indicates whether cardboard is being fed or not S3 Speed (m / s) Specifies the speed at which cardboard is fed s 4 Speed (m / s) Specifies the speed at which cardboard is moved S5 Temperature (° C) Surface temperature of the fuser roller s 6 Speed (m / s) Specifies the speed at which cardboard is moved 67 Pressure (bar) Indicates the pressure used for punching S 8 Speed (m / s) Specifies the speed at which cardboard is moved
    Furthermore, the attributes can also be defined by external specifications and boundary conditions, which can change over time, such as an order number, a default setting such as the punching force to be used when executing the order, or an external one, not by means of an on the machine Mt. located sensor to be determined, such as the ambient temperature or ambient humidity.
    The external specifications for executing an order on the machine Mt and the environmental parameters such as ambient temperature and / or ambient air humidity and / or ambient air pressure to which the machine Mt is exposed during a production process are characterized by four attributes A ^ A 2 , A 3 , A 4 characterized. The attributes of the machine M ^ which describe external specifications and environmental parameters are explained in more detail in Table 2.
    Table 2: Attributes of the machine M ^ that relate to external specifications and environmental parameters.
    Αχ Carton thickness (mm) AE2 Size (1 in cm x b in cm) A3 Color (binary) a 4 Relative humidity (%)
    Values determined for these attributes are entered at the relevant times t in the columns i = 9 ... 12 in the relevant row vector m '.
    Finally, the attributes can be time-independent attributes, such as an identification ID and / or a machine type. These attributes are fixed and unchangeable for the machine Mt in question, ie if the attribute with the attribute identification i is time-independent, then x, 1 = x, 2 = ... = χ, τ . Values determined for these attributes are entered at the relevant times t in the columns i = 13... 14 in the relevant row vector m '.
    8/26
    In such a method, the normal operating state of a machine Mt is usually described with a rule set R, the following types of rules from the prior art being used, for example:
    - statistical rules see Georgii, Hans-Otto. Stochastics: Introduction to Probability Theory and Statistics. Walter de Gruyter GmbH & Co KG, 2015. Hartung, Joachim, Bärbel Elpelt, and Karl-Heinz Klösener. Statistics: teaching and handbook of applied statistics. Walter de Gruyter, 2009.,
    - Correlation rules see Friedberg, Ivo, et al. Combating advanced persistent threats: From network event correlation to incident detection. Computers & Security 48 (2015): 35-57., Or Skopik F., Fiedler R. (2016): EP 1416597.2-1853 - Method for detecting deviations from a given standard state, June 2016.,
    - Neural network models see Rojas, Raul. Theory of neural networks: a systematic introduction. Springer-Verlag, 2013.,
    - Regression models see Fahrmeir, Ludwig, Thomas Kneib, and Stefan Lang. Regression: models, methods and applications. Springer publishing house, 2007.
    Statistical rules characterize a single attribute of a machine M ^ wherein a sequence (x i) of an attribute with tez / € {1, ..., n} j via e ne finite period Z will be considered. A rule R is based on the values that the attribute Xi. assumes in the so-called training phase Z o . Such a rule R defines, for example, an interval within which the mean value of the values of the attribute in a time period Z is considered, or the probability distribution with which the distribution of the values of the attribute in a time period Z is described. Statistical tests such as B. a Kolmogorov-Smirnov test, see Hartung, Joachim, Bärbel Elpelt, and Karl-Heinz Klösener. Statistics: textbook and manual of applied statistics. Walter de Gruyter, 2009, can be used for this.
    Using a statistical rule established according to the prior art, the mean value Λ ί of a sample in the training phase Z o can be calculated as xF = Σ, εΖο 4
    Assuming that the sample is representative, an interval is defined that allows a tolerance range of twice q100%. The corresponding rule is defined as R ^ tez): if Äf £ [Ä ° - oxf °, + aX ^ j => alarm!
    9.26
    To check the representativity of the estimated mean, a confidence interval can be calculated instead of a point estimate.
    In the case of correlation rules known from the prior art, for example, an attribute Xi is considered at one point in time or at several points in time, ie two or more values of a column in the machine attribute matrix, correlation rules of the form : if not < a 'V> alarm! be set up. Two or more different attributes can also be considered at a time Tu, ie two or more values of a line of the machine attribute matrix, with correlation rules of the form: if not <a => x > b) => alarm! result in another variant de | Creation of correlation rules are considered two or more different attributes at two or more times Tt, , ···, ie values from different columns and different rows of the machine attribute matrix, with correlation rules of the form : not 1 < 0 χ : 1 ~> b) alarm! au fg it t e m d s who.
    Neural network models are used in the prior art to assess the current state of a machine Mu of a machine component or of a work process carried out on the machine Mt. Neural networks are trained and the values of selected attributes of the are used as input parameters
    Machine Mt uses, where I is a set of attributes. With neural network models, rules of the form if> a no alarm! V if < a alarm! created
    Regression models are created by means of regression analysis and are used in the prior art for predicting the state of a machine Mu of a machine component or a work process carried out on the machine Mt. Similar to neural network models, values of a defined time interval% = l T o, T '] (x ) I e i of selected attributes of the machine Mt are used as input parameters for regression models, where 7 is a set of attributes um predict the state of the machine Mt after the elapsed time up to Τ '. Regression models are used, for example, to create rules of the form = state at T '+ At.
    10/26
    In a first exemplary embodiment of a method according to the invention, the machine Mt is used as the standard machine, the process carried out on the machine Mt as the standard process and the working environment in which the machine Mt is located as the standard environment and the external specifications for the machine Mt as external standard specifications to be viewed as.
    As soon as the machine Mt is in a stable state, that is to say, for example, after a start-up phase, such a method provides that the first step is to search for rules for the machine M ^, that is to say the standard machine, in a learning phase using, for example, the methods described above. The rules describe the relationship between the measured values determined by the sensors Si, ... S 8 and the relationship between the determined measured values and the environmental parameters during the normal operating state of the machine Mt in the relevant working environment with the external specifications and are during the normal operating state of the Machine Mt met. For machine Mt, for example, all attributes Xm can be used as input parameters for a rule and rules for machine Mt have the general form = R (Xm
    The control set R M i learned for the machine Mt in the present case contains eight
    Regulate. The rules R ^ .-. Rs have the following form, where T, the times with
    Ai = T i + i - Ti = 0.5s marked, at which measured values from sensors Si, ... S 8 are collected:
    • 7 I (5i, 5 3 ): Si G [24.9,25.1] min 1 =>S3> 0.25 m / s; (v = 2r r x -) • 7 2 (A 2 , 5 3 ): l = 50cm S3 <0.262m / s • B3 (i, S 2 ): 5 <E ' = i _ 9 5 2 (T,) <8 • 7 4 (A 3 , 5 5 ): A3 = 1 => S 5 G [178.182] ° C • 7 5 (Ai, 5γ): Ai = 3 mm => S7 G [9.8,10.2] bar • 72 6 (5 3 , 5 4 , 5 6 , 5 8 ): 0.25 m / s <5 3 <5 4 <5 6 <5 8 • T 7 (i, 5i): i> 1000 => Ej = i _ 999 G [24.95,25.05] min “ 1 • 7 8 (ä 4 , 5 7 ): A 4 > 40% => 5 7 > 10 bar
    Rule R 3 checks, for example, whether sufficient boxes are being drawn in within 10 magazines or whether the feeder 11 is open for a sufficiently long time, with a box duration of approximately 3 seconds for the feed unit 11 being opened
    11/26 is assumed. If the feed unit 11 is opened too long, for example, too many cartons would be drawn in.
    In order to transmit the rules R ^ .-. Rs set up for the machine M ^, ie the standard machine, a method according to the invention provides as a next step that the rules R ^ .-. Rs are initially presented in an abstract form. Based on the set of attributes of the machine Mt = {Si, S 2 ,. , .ββ, Α ^ Αζ, Αζ, Α ^} unc | In the learned rule set R M for the machine Mt R-Mi = {Ri, ·, Rs}, a function X {R): R—> X is defined, which delivers all attributes that a rule uses as an input parameter. In addition, a second function RpoE) is: set up, all
    Returns rules that use attributes of the set Xi as input parameters, where I represents an index set. As the next step of a method according to the invention, all variable control attributes are learned in a learning phase. In the case of a correlation rule, for example, parameters a, b, ... can be introduced as variable attributes (see abstract rules in the next paragraph).
    In the following, the rules Ri, ... R 8 valid for the machine Mt are given in abstract form and an explanation of each rule Ri, ... R 8 , with Latin letters during the learning phase being values of constants and Greek letters Represent values of tolerances:
    • Ri; a, a,: Si = [a - cv, a + a] min -1 S3> v - ß m / s; (y - 2πτ α 1 )
    Since the machine Mt is driven by a V-belt, a negative evaluation of the rule Rt would indicate a problem with the power transmission. The expected speed at which the V-belt moves can be calculated from the speed of the drive of the machine Mt.
    • ^ 2-, a (A 2 , S3): l = A 2 (l) cm => S3 <α m / s
    This rule describes the relationship between the box length and the speed of the conveyor belt. The function A 2 (l) provides the value for the box length, which is specified in the external specifications for the order carried out on the machine Mt.
    • R3; a, b, cÜ, 52) = «<<&
    Depending on the size and machine speed of the machine M ^, the feeder must be open for a certain time so that a box can be drawn in. Depending on the processing time required per box, the feed 11 must remain closed for a certain time in order to prevent a cardboard jam. Out
    These factors result in the limit values a, b, how often the sensor S 2 delivers the value 1, which corresponds to the open state, in the time window c.
    Ri: a, a (Ä3, S5): A3 => S5 G [ft - O; ft + O '] ° C
    This rule controls whether the fuser roller is at the right temperature for printing.
    • R5-, a, a (Ai> S7) Aj => S7 G [η - a; a 4- a] bar
    This rule checks whether the appropriate force is used for punching, so that creases are not punched too deeply, or whether other parts are actually punched out completely.
    • A-j.A-S'i, S4, 5ß, 5g): a m / s <S3 <S4 <Re <, ¾
    This rule ensures that there is no cardboard jam due to the decreasing speeds of the conveyor belts connecting the individual machine sections.
    • R ; A, b, c, a (^ Rl) i O, => jljü Rl θ [c CÜ, C + Cl]
    This rule checks whether the mean value of the revolutions per minute of the conveyor belt drive, measured with sensor St over the last b + 1 measurements, lies in the correct interval. This evaluation is based on a statistical rule that uses an estimated mean.
    • R8; a.b (A4, R7) 'A4> a% - S7> b bar
    This rule establishes a connection between the environmental parameters of the working environment in which the machine Mt is operated and the sensor S 7 of the machine Mt. If e.g. B. the humidity in the production hall in which the machine Mt is operated exceeds a certain value, more pressure must be applied to the punching.
    2 schematically shows a machine M ', the rules of the machine M ^, which is referred to below as the standard machine, to be transferred to the machine M'. The machine M 'has a feed unit 21, which is essentially identical to the feed unit 11 of the standard machine Mt and on which there are three sensors S' ^ S ' 2 , S' 3 , the feed unit 21 via a first conveyor belt Section 22 is connected to a punching unit 25 which is essentially identical to the punching unit 15 of the standard machine Mt and is located on the
    13/26 a first conveyor belt section 21 has a sensor S ' 4 and a sensor S' 7 on the punching unit 25. The punching unit 25 is connected via a second conveyor belt section 26 to an ejection unit 27 which is essentially identical in construction to the ejection unit 17 of the standard machine Mt, a sensor S ′ 8 being located on the second conveyor belt section 26. The sensors S '^ S' 2 , S ' 3 , S' 4 , S ' 7 , S' 8 repeatedly determine physical measured values at predetermined times.
    The machine M 'thus has a number of sensors S' ^ S ' 2 , S' 3 , S ' 4 , S' 7 , S ' 8 , the position and sensitivity of which at least partially match the sensors on the standard machine Mt. Compared to the machine M ^, ie the standard machine, the machine M 'has structural differences, since the conveyor belt section 12 running between the feed unit 11 and the pressure unit 13 and the sensor S 5 thereon and the pressure unit 13 and the sensor S 6 located on the printing unit 13, which is present on the machine Mt, is missing. The external specifications for executing an order on the machine M 'and the environmental parameters to which the machine M' is exposed during a production process are identified by three attributes A '^ A' 2i A ' 4 . Thus the machine M 'has the following attributes = {S ^, S 2 , S' 3 , S ' 4 , S' 7 , S ' s , A {, A ^, A' 4 }
    In order to transfer rules set up for the standard machine Mt to the machine M ', a method according to the invention provides as a next step that those sensors of the machine M' are determined which can be assigned to a sensor of the standard machine Mt in terms of arrangement and sensitivity, whereby in the specific example, the sensors Si, S 2 , S 3 , S 4 , S 7 , S 8 or S'i, S ' 2 , S' 3 , S ' 4 , S' 7 , S ' 8 are concerned. Furthermore, it is provided that those external specifications and environmental parameters of the machine M 'are determined which correspond to the external standard specifications and standard environmental parameters of the standard machine Mt. In the specific example, these are the attributes A't, A ' 2 , A' 4 . It is further provided that those sensors of the standard machine M are also determined which cannot be assigned a sensor of the machine M 'in terms of arrangement and sensitivity, the sensors S 5 and S 6 being in the example shown. In addition, those external standard specifications and standard environmental parameters of the standard machine Mt are determined which do not match the external specifications and environmental parameters of the machine M ', which in the example is the attribute A 3 .
    14/26
    According to the invention, the function Rm: X → R, which maps from the space of all attributes X to the space of all rules R, is used to identify those rules which are used on the standard machine Mt and can also be used on the machine M ' , because the rules apply to the same attributes. The following rules result in the example, which are transferred from the standard machine Mt to the machine M ': Rm 1 {^ m > ) = Rm' = {R'iiR ^ iR ^^ iR'hiR ^ iiR's}.
    If a machine M 'lacks sensors that are present in the standard machine Mt, a method according to the invention advantageously provides that the rules that apply to these sensors in the standard machine Mt are deleted. Likewise, rules are also deleted that apply to external standard specifications or standard environmental parameters in the standard machine Mt, which do not occur as external specifications or environmental parameters in the machine M '.
    As a further step in the transmission of rules according to the invention, it is examined whether the rules transmitted to the machine M 'are fulfilled by the measured values of those sensors of the machine M' which correspond to the sensors and sensors of the standard machine Mt with regard to arrangement and sensitivity. If this is the case, the rules for characterizing the normal operating state of the work process are adopted on the machine M '. If this is not the case, the rules are adapted to the new measured values of the machine M 'in that the constants and tolerances for the machine M' which are present in the abstract rules are relearned in a learning phase, or the rules are rejected. Furthermore, it is examined whether the rules transmitted to the machine M 'are fulfilled by the external specifications and environmental parameters of the machine M', which correspond to the external standard specifications and standard environment parameters of the standard machine Mt. If this is the case, the rules are applied to the machine M '. If this is not the case, the rules are adapted to the new external specifications and environmental parameters of the machine M ', as described above, i.e. by re-learning the constants and tolerances for the machine M 'in the abstract rules in a learning phase, or by ultimately rejecting the rules.
    A further variant of a method according to the invention provides that when rules are transferred, it is checked whether sensors assigned to one another on machine M 'and standard machine Mt have the same characteristic with regard to the physical quantity to be determined. In the event that a machine M 'on which
    15/26
    Rules of a standard machine M1 are to be transmitted, encompassing sensors which determine the same physical parameters as sensors of the standard machine M1, but have a different characteristic curve, those rules which apply to these sensors on the standard machine M1 are passed to the new characteristic curve of the machine M ' customized.
    Alternatively, a machine M 'to which the rules of a standard machine M1 are to be transferred can also be built in such a way that, in comparison to the standard machine M1, no components and / or sensors are omitted, but rather components and / or sensors are added and / or components are modified are and / or sensors of the standard machine M1 repositioned or changed in terms of their sensitivity or calibration. In a method according to the invention, it can therefore also be provided that it is examined whether the machine M 'to which the rules of the standard machine M1 are transferred has sensors to which no sensor of the standard machine M1 can be assigned in terms of arrangement and sensitivity. In this case, a method according to the invention provides that new rules are set up for the measured values of the sensors of the machine M 'thus determined which characterize the normal operating state of the machine M'. These new rules for the machine M 'are created in the manner described above, which are also used for the creation of rules for the measured values of the standard process of the standard machine M1, these rules in particular also taking into account measured values from sensors which are both on the machine M ', as well as on the standard machine M1. It is also examined whether the external specifications and environmental parameters of machine M 'match the external standard specifications and standard environmental parameters of standard machine M1. In this case it can be provided that the rules of the standard machine M1, which relate to these external standard specifications and standard environment parameters, are transmitted to the machine M 'and that it is checked whether the transmitted rules are met by the external specifications and environmental parameters of the machine M' , If this is the case, the rules are applied to the machine M '. If this is not the case, the rules are adapted to the new external specifications and environmental parameters of the machine M ', as described above.
    A further variant of a method according to the invention provides that rules from at least two standard machines M1, M2, which include sensors and are operated with the respectively valid external standard specifications and with the prevailing standard environment parameters, are transferred to a further machine M '/ 26. In this case, for both standard machines M1, M 2, as described above, rules for the respectively valid set of attributes are first searched for, which relate to the relationship between the measured values determined by the sensors of the standard machines M1, M 2 during the normal operating state of the standard machines M1, M 2 Describe in the relevant standard working environment with the external standard specifications, and which are fulfilled during the normal operating state of the standard machines M1, M 2 . In order to transfer the rules drawn up for the standard machines M1, M 2 to the machine M ', the next step is to present the rules valid for the standard machines M1, M 2 in an abstract form. It is then checked whether sensors of machine M 'match sensors of standard machines M1, M 2 in terms of sensitivity and arrangement, and whether external specifications or environmental parameters of machine M' match the respective external standard specifications or standard environmental parameters of standard machines M1, M 2 .
    If sensors and / or external specifications and / or environmental parameters of the machine M 'match with at least one of the standard machines M1 and M2, those relating to the same attributes and on at least one of the standard machines M1, M 2 or on both standard machines M1 , M 2 valid rules transferred to the machine M '.
    As a further step, it is examined whether the rules transmitted from one of the standard machines M1, M 2 or both standard machines M1, M 2 to the machine M 'by the measured values of those sensors of the machine M' which are connected to the sensors of at least one of the standard machines M1, M 2 match in terms of arrangement and sensitivity, are met. Furthermore, it is examined whether the rules transmitted to the machine M 'are fulfilled by the external specifications and environmental parameters of the machine M', which correspond to the external standard specifications and standard environment parameters of at least one of the standard machines M1, M 2 . If this is the case, the rules are applied to the machine M '. Furthermore, it can advantageously be provided in a method according to the invention that the abstract rule forms of the rules transmitted to the machine M 'are compared. In the event that two or more rules have the same abstract form and these rules are valid for the machine M ', the respectively more suitable values for variables and tolerances are selected. When adopting rules, the lower tolerances and narrower intervals are advantageously selected.
    / 26
    If the rules transmitted to the machine M 'are not met by the external specifications and environmental parameters of the machine M', which correspond to the external standard specifications and standard environmental parameters of at least one of the standard machines M1, M2, the constants and tolerances present in the abstract rules become learned again for the machine M 'in a learning phase and the rules are adapted to the new external specifications and environmental parameters of the machine M', or the rules are rejected.
    If a machine M 'lacks sensors which are present in at least one of the standard machines M1, M2, it is provided that the rules which apply to these sensors in at least one of the standard machines M1, M2 are deleted. Rules are also deleted that apply to at least one of the standard machines M1, M2 for external standard specifications or
    Standard environmental parameters apply, which do not occur with the machine M 'as external specifications or environmental parameters.
    In the event that the machine M 'has sensors to which no sensor of at least one of the standard machines M1, M2 can be assigned in terms of arrangement and sensitivity, new rules are set up as described above, which characterize the normal operating state of the machine M'.
    Alternatively, sets of rules can also be stored in a central switching point, the rules describing the interrelationships of the measured values determined during the normal operating states of a large number of standard machines M1,..., M n in the relevant working environments with the external specifications that occurred during the normal operating conditions are met. According to the invention, suitable sets of rules are selected for a work process with external specifications in a work environment of a machine M 'similar or identical to the standard machines M1,..., M n . To select suitable rules, it is checked whether environmental parameters of the working environment and / or external specifications and / or sensors of the machine M 'with standard environmental parameters of the working environment and / or external standard specifications and / or with sensors with regard to arrangement and sensitivity of at least one of the standard machines M1, .. ., M n match. The selected rules are transferred to the work process of the machine M 'in the work environment.
    / 26
    In a further variant of a method according to the invention it is provided that, for example, a machine M ^ as shown in FIG. 1, which is operated in a working environment and comprises sensors S ^ .-. Ss, the sensors during the on the machine Mt. carried out work process at predetermined recurring times determine measured values, as described above rules R ^ .-. Rs are determined for the normal operating state of the machine Mt and these rules R ^ .-. Rs are applied to the measured values of the sensors of the machine Mt. Since the rules R ^ -.- Rs are met with the external specifications during the normal operating state of the machine Mt in the relevant working environment, an abnormal state of the machine Mt and / or the work process can be identified if measured values are available that comply with the rules Ri, ... R 8 are not sufficient for normal operating conditions.
    Table 3 shows an example of the evaluation of the rules R ^ .-. Rs described above for the machine M · ,, shown in FIG. 1, the rules R ^ .-. Rs for the measured values of the sensors S ^ .-. Ss and the external specifications and environmental parameters Ai, A 2 , A 3 , A 4 apply and are viewed over a period R to T 140 .
    Table 3: Example evaluation of the rules for machine M ^
    t Si s 2 S 3 s 4 s 5 s e sy s 8 Ai a 2 ^ 3 a 4 Ti 25 1 00:26 0.2601 180 0.2601 9.9 0.2602 3 50 1 36 T 2 25 1 0259 00:26 180.2 00:26 9.95 0.2601 3 50 1 377140 25 0 00:26 0.2603 171 0.2604 9.9 0.2604 3 50 1 41
    In the example given in Table 3, the temperature of the fixing roller of the printing unit 13 of machine Mt has dropped so far that the pressure is no longer fully visible on the carton. In the example shown, the information about a deteriorating print quality is obtained with a method according to the invention by a failure of the R 4 rule. In addition, the air humidity at time T 140 has risen to over 40%, which leads to a failure of rule R 8 and the cutting die only exerts a pressure of 9.9 bar, as a result of which rule R 5 is violated. Since the measured values of the sensors of the machine Mt do not meet the rules R 4 , R 5 and R 8 in the example shown, an abnormal state of the machine Mt or of the production process carried out on the machine Mt is indicated by a method according to the invention.
    19/26
    With such a method, the time-consuming troubleshooting in a production process carried out on the machine M1 can be shortened and the duration of downtimes can be reduced, as a result of which the production process can be optimized and its efficiency increased. The automatic determination of the rules using the methods described above advantageously also eliminates the time-consuming manual determination and configuration of the rules.
    / 26
    权利要求:
    Claims (11)
    [1]
    claims
    1. A method for recognizing the normal operating state of a work process in which a machine (M ') is operated in a work environment with external specifications, the machine (M') sensors (S'1, ..., S ' m ) for Measurement of physical parameters at predetermined positions in the area of the machine M ', wherein
    a) - a standard process is started with a standard machine (M1) in the standard environment with external standard specifications,
    the standard machine (M1) has a number of sensors (S1, ..., S n ),
    - With the sensors (S1, ..., S n ) of the standard machine (M1), measured values are repeatedly determined at predetermined times, and
    - Searching for rules (R1, ..., Rp) that describe the relationship between the measured values determined during the normal operating state of the standard machine (M1) in the relevant working environment and the external specifications, and which are fulfilled during the normal operating state characterized that subsequently
    b) - the rules (Ri, ..., R p ) determined for the standard process are transferred to the work process of a machine (M ') similar or identical to the standard machine (M1) in the work environment,
    - wherein the machine (M ') has a number of sensors (S'1, ..., S'm) of the machine (M'), the position and sensitivity of which are with the sensors (S1) located on the standard machine (M1) , ..., Sn) at least partially matches and
    - With the sensors (S'1, ..., S'm) of the machine (M '), measured values are repeatedly determined at predetermined times,
    - It is examined whether the transmitted rules (R1, ..., Rp) are determined by the measured values of those sensors (S'1, ..., S'm) of the machine (M ') that are connected to the sensors (S1 ,. .., Sn) of the standard machine (M1) match in terms of arrangement and sensitivity, are met, and
    - if this is the case, the rules for characterizing the normal operating state of the work process are adopted, and
    - if this is not the case, the rules are adapted to the new measured values or rejected.
    [2]
    2. The method according to claim 1, characterized in that
    21/26
    - That it is examined whether sensors (S 1 , ..., S n ) are present in the standard machine (M1), which are not assigned to the machine (S'1, ..., S ' m ) in terms of arrangement and sensitivity ( M ') can be assigned,
    - Where in the absence of assignable sensors (S'1, ..., S ' m ) on the machine (M') the rules that apply to these sensors in the standard machine (M 1 ) are deleted.
    [3]
    3. The method according to claim 1 or 2, characterized in that it is examined whether mutually assigned sensors (S'i, ..., S'm) of the machine (M ') have the same characteristic with regard to the physical quantity to be determined,
    - In the case of mutually assigned sensors (S ' 1 , ..., S' m ) of the machine (M ') with a different characteristic curve, those rules that apply to these sensors (S1, ..., Sn) in the standard machine ( M 1 ) are valid, to be adapted to the new characteristic of the machine (M ').
    [4]
    4. The method according to any one of the preceding claims, characterized in
    - That it is examined whether sensors (S ' 1 , ..., S' m ) are present on the machine (M '), which are not in the arrangement and sensitivity of any sensor (S1, ..., Sn) of the standard machine (M 1 ) can be assigned,
    - that for the measured values of the sensors (S ' 1 , ..., S' m ) of the machine (M ') determined in this way, new rules (R' 1 , ..., R ' q ) are drawn up which reflect the normal operating state characterize,
    - In particular, these rules (R ' 1 , ..., R' q ) are created in the same way that are used to create the rules (R 1 , ..., R p ) for the measured values of the standard process .
    - whereby these rules (R ' 1 , ..., R' q ) in particular can also take into account measured values from sensors (S ' 1 , ..., S' m ), both on the machine (M ') and are available on the standard machine (M 1 ).
    [5]
    5. The method according to any one of the preceding claims, characterized in that the machine (M ') by modification of the standard machine (M 1 ) or with the standard machine (M 1 ) identical to the other standard machine (M2) is created by
    a) components and / or sensors (S 1 , ..., S n ) are added to the standard machine (M 1 ), and / or
    b) components and / or sensors (S 1 , ..., S n ) are removed from the standard machine (M 1 ), and / or
    c) components of the standard machine (M 1 ) are modified, and / or
    22/26
    d) Sensors (Si, ..., S n ) of the standard machine (M1) are repositioned or changed with regard to their sensitivity or calibration.
    [6]
    6. The method according to any one of the preceding claims, characterized in that the work process is determined in which the machine (M ') or the standard machine (M1) is operated in a work environment which differs from the standard environment, the work environment being different from the standard environment in particular differs in terms of the following environmental parameters:
    a) ambient temperature, and / or
    b) ambient air humidity, and / or
    c) ambient air pressure, ...
    [7]
    7. The method according to any one of the preceding claims, characterized in that it is checked whether the machine (M ') has external specifications that at least partially match the external standard specifications of the standard machine (M1) and that it is examined whether the transmitted Rules (R1, ..., R p ) are met by the external specifications of the machine (M '), which correspond to the external standard specifications of the standard machine (M1), and
    - if this is the case, the rules for characterizing the normal operating state of the work process are adopted, and
    - if this is not the case, the rules are adapted or rejected to the new external requirements or new rules are set for these new external requirements.
    [8]
    8. The method according to any one of the preceding claims, characterized in
    - that the machine (M ') is operated in the working environment,
    - that measured values are determined during the work process at predetermined recurring times,
    - that the rules R ' q ) determined for the work process are applied to the measured values, and
    - If there are measured values that do not meet the rules (R'1, ..., R ' q ) for the normal operating state, an abnormal state of the machine (M') and / or the work process is identified.
    [9]
    9. The method according to any one of the preceding claims, characterized in that rules (R1, ..., R p ) are determined on at least two standard machines (M1, M2), which determine the relationship between the measured values determined during the normal operating state
    Describe 23/26 of the standard machines (M1, M 2 ) in the relevant working environment with the external specifications, which are fulfilled during normal operating conditions, and
    - wherein measured the (2 M1, M) for at least two standard machines rules (R1, R n ...) are transferred to the work process of a standard machines (M1, M 2) similar or identical machine (M ') in the working environment ,
    [10]
    10. The method according to any one of the preceding claims, characterized in that rule sets (R), in particular comprising a number of determined rules, are stored in a central switching point,
    - The rule sets (R) describe the relationships between the measured values determined during the normal operating states of a large number of standard machines (M1, ..., Mn) in the relevant working environments with the external specifications, and which are fulfilled during the normal operating states,
    - Suitable control sets (R) are selected for a work process with external specifications in a working environment of a machine similar to the standard machines (M1, ..., M n ) or identical (M '), the control sets (R) being matched by the sensors and / or the environmental parameters of the working environment and / or the external specifications of the machine (M ') are selected, and
    - The selected rule sets (R) are transferred to the work process of the machine (M ') in the work environment.
    [11]
    11. A data carrier on which a program for carrying out a method according to one of the preceding claims is stored.
    24/26
    1.1
    类似技术:
    公开号 | 公开日 | 专利标题
    DE102017000536A1|2017-08-03|Cell controller for detecting a cause of an abnormality in a manufacturing machine
    DE102018203280A1|2018-09-13|State diagnostic device
    DE102012215984B4|2019-05-09|Method and device for cost-optimized model-based extension of system lifetime
    DE112008003195T5|2011-07-07|Electrical circuit with a physical transmission layer diagnostic system
    DE102018133196A1|2020-06-25|IMAGE-BASED MAINTENANCE PROPERTY AND MISUSE DETECTION
    EP2778817A1|2014-09-17|Monitoring of the initial equipment of a first technical system by means of benchmarks
    DE102017210787A1|2018-12-27|Method and apparatus for detecting anomalies in a communication network
    AT519777B1|2019-12-15|Method for recognizing the normal operating state of a work process
    DE102017211737B4|2019-03-28|Monitoring device and method for monitoring a system
    EP3474091A1|2019-04-24|Method and device for controlling a process within a system, in particular of a milling process in a milling device
    DE102008032885A1|2010-01-21|Method and device for checking and determining states of a sensor
    EP3151072B1|2020-07-29|Method and system for error detection and monitoring in an electronically controlled or regulated machine part
    DE102009047724A1|2011-06-16|Program sequence monitoring method for operating field device utilized for e.g. measuring physical parameter at defined areas in process plant, involves comparing test value with another test value to monitor program sequence
    DE102016012451A1|2018-01-04|Method for monitoring, analyzing and operating at least one production facility
    DE102018128640A1|2019-05-23|Device for estimating a noise generation cause
    DE102018221684A1|2020-06-18|Device and method for anomaly detection and for training a model for anomaly detection
    DE102005038889A1|2007-02-22|Method for testing a production material
    EP3282399B1|2020-05-13|Method for the improved detection of process anomalies of a technical installation and corresponding diagnostic system
    DE102017213060A1|2019-01-31|A computer-implemented method and apparatus for automatically generating tagged image data and analyzer for inspecting a component
    DE112012006909T5|2015-06-03|Control system management device
    EP1157317A1|2001-11-28|Method and device for reducing a number of measured values of a technical system
    DE102006003611A1|2007-08-02|Locating faulty data record in number of data records involves forming cells in input data space, characterizing as faulty any data record of selected cell whose target data item value fails check criterion derived from target data values
    DE102019120864B3|2020-12-17|Method and device for planning maintenance work on at least one machine
    EP3951529A1|2022-02-09|Monitoring device and method for detection of abnormalities
    DE102018128315A1|2020-05-14|Method, device, computer program and computer program product for checking a first adaptive system model
    同族专利:
    公开号 | 公开日
    EP3396477A1|2018-10-31|
    EP3396477B1|2022-01-05|
    AT519777B1|2019-12-15|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    DE4436658A1|1993-10-13|1995-04-20|Hitachi Ltd|Device and process for investigating disturbances|
    US20090093892A1|2007-10-05|2009-04-09|Fisher-Rosemount Systems, Inc.|Automatic determination of the order of a polynomial regression model applied to abnormal situation prevention in a process plant|
    EP2523150A1|2011-05-11|2012-11-14|General Electric Company|System and Method for Optimizing Plant Operations|
    US9092802B1|2011-08-15|2015-07-28|Ramakrishna Akella|Statistical machine learning and business process models systems and methods|
    US6760716B1|2000-06-08|2004-07-06|Fisher-Rosemount Systems, Inc.|Adaptive predictive model in a process control system|
    TWI336823B|2004-07-10|2011-02-01|Onwafer Technologies Inc|Methods of and apparatuses for maintenance, diagnosis, and optimization of processes|
    US8489218B2|2010-10-15|2013-07-16|Taiwan Semiconductor Manufacturing Company, Ltd.|Chamber match using important variables filtered by dynamic multivariate analysis|
    US9703902B2|2013-05-09|2017-07-11|Rockwell Automation Technologies, Inc.|Using cloud-based data for industrial simulation|
    WO2016145089A1|2015-03-09|2016-09-15|Skytree, Inc.|System and method for using machine learning to generate a model from audited data|EP3767404A1|2019-07-18|2021-01-20|Siemens Aktiengesellschaft|Implementing event rules for maintenance relevant events in a multitude of machines|
    法律状态:
    优先权:
    申请号 | 申请日 | 专利标题
    ATA50233/2017A|AT519777B1|2017-03-22|2017-03-22|Method for recognizing the normal operating state of a work process|ATA50233/2017A| AT519777B1|2017-03-22|2017-03-22|Method for recognizing the normal operating state of a work process|
    EP18160444.8A| EP3396477B1|2017-03-22|2018-03-07|Method for determining rules for characterizing the normal operating state of a work process|
    [返回顶部]