Method for monitoring and modifying an intelligent algorithm for setting operating parameters of a container treatment machine and device for executing the method

Abstract

The invention relates to a method for monitoring and modifying an intelligent algorithm (1) for setting operating parameters (4) of a container treatment machine (5), wherein the method comprises determining a current reward (13) from a comparison of a setpoint (9) of a container property to a measured actual value (8) of the container property, determining an expected reward (16) from a comparison of the setpoint to an expected, simulated actual value (10) of the container property, wherein the expected, simulated actual value is based on the intelligent algorithm (1), automatically analyzing (19) the current reward (13) and the expected reward (14), and modifying (21) at least one manipulated variable (2) of the intelligent algorithm based on a result of the automatic analysis for setting operating parameters of a container treatment machine.The invention further relates to a device comprising a computer-readable storage device with instructions stored thereon, which, when executed by a processor of the device, cause it to execute the method.

Description

The invention relates to a method for monitoring and modifying an intelligent algorithm for setting operating parameters of a container treatment machine and to an apparatus for carrying out the method according to the independent claims.Prior ArtUS 2022 / 0147876 A1 discloses a method for providing an reconnaissable agent for estimating a reconnaissable reward function. The method includes detecting an observed space that includes one or more states and one or more actions, and modeling the observed space as a plurality of state-action pairs that can be elucidated. Further, the method includes forming one or more clearable models that include a simulated environment and a reward function, and returning at least one explanation of at least one model of the one or more clearable models that corresponds to at least one state and at least one reward.For setting parameters of a container treatment machine, monitoring can be provided on the machine, for example by means of a programmable logic controller (PLC) (PLC), or on an edge device. Monitoring of the work and performance of an artificial intelligence (AI) can take place. The monitoring can pass on messages or information about problems and weaknesses of the container treatment machine. A human operator can carry out an analysis in the event of anomalies occurring, make decisions in the event of incorrect behavior and initiate corresponding measures.The object is to provide a method for performing a processThe object of the invention is to provide a method for monitoring and modifying an intelligent algorithm for setting operating parameters of a container treatment machine and an apparatus for carrying out the method, which allow self-healing of the intelligent algorithm with changed boundary conditions.SolutionThe object is achieved by the method for monitoring and modifying an intelligent algorithm for setting operating parameters of a container treatment machine and the apparatus for carrying out the method according to the independent claims. Further embodiments are disclosed in the dependent claims.The method according to the invention for monitoring and modifying an intelligent algorithm for setting operating parameters of a container treatment machine comprises determining a current reward from a comparison of a setpoint value of a container property with a measured actual value of the container property, determining an expected reward from a comparison of the setpoint value with an expected simulated actual value of the container property, wherein the expected simulated actual value is based on the intelligent algorithm, automatically analyzing the current reward and the expected reward and modifying at least one manipulated variable of the intelligent algorithm based on a result of the automatic analysis for setting operating parameters of a container treatment machine.A container treatment machine can comprise a blow molding machine for producing PET containers from preforms, a filling system for filling product into containers, a closing device for closing containers filled with product, a pasteurization device for pasteurizing containers filled with product, a labeling machine for applying labels to containers and / or a cleaning device for cleaning containers or the like.The operating parameters can comprise a heating setting for heating the preforms and / or a blowing pressure for forming a container from a preform in the blow mold (for example in a blow molding machine), a filling speed and / or filling temperature of the product (for example in a filling system), a mounting pressure of a closure and / or mounting speed of a closure (for example in a closing device), a pasteurization temperature and / or a dwell time below the pasteurization temperature (for example in a pasteurization device), an operating speed (for example in a labeling machine), a heating temperature of a solution bath and / or a dwell time of a container in a solution bath (for example in a cleaning device).Adjusting operating parameters may include adjusting a single operating parameter or adjusting multiple operating parameters. The plurality of operating parameters may be independent of one another, at least partially dependent on one another, or dependent on one another.In addition to the operating parameters of the container treatment machine, further boundary conditions may have an influence on a container which is treated in the container treatment machine. For example, an ambient temperature (i.e. a boundary condition) in which the container treatment machine is operated can have a first temperature in a first operating cycle and a second temperature in a second operating cycle; for example, a difference between the first and second temperatures can be 30% or more or even less. The humidity of the air can also have an influence on the results of the container treatment machine. For example, a wall thickness (i.e. a boundary condition) of preforms per batch can vary, so that an influence on a wall thickness of a container to be achieved results therefrom and, for example, a desired wall thickness (setpoint value) of the container cannot be achieved by means of the container treatment machine under the prevailing operating parameters for one of the batches.The containers may comprise bottles, for example of PET, glass, fibre material (such as wood fibres or other biodegradable fibres) or the like, or cans.A set point of a container property may be a wall thickness or a wall thickness distribution of the container, a number of pasteurization units of the product received in a container, the position of a label on the container and / or the position of a closure on the container and / or the like.The measured actual value of the container property can be a wall thickness determined by measurement or a wall thickness distribution of the container, a number of recorded pasteurization units of the product in a container determined by measurement, the position of a label on the container determined by measurement and / or the position of a closure on the container determined by measurement and / or the like.The expected, simulated actual value can result from a simulation of a process of the container treatment machine, wherein the process can comprise the corresponding container treatments (for example blow molding a container from a preform in a blow mold which is heated to a temperature and shaped with a blow pressure) and wherein the control variables can be used which are available to the intelligent algorithm.From the comparison of the current forward and the expected forward, conclusions can be drawn as to whether the intelligent algorithm with the control variables made available to it can regulate the container treatment machine by setting operating parameters in such a way that the measured actual value corresponds to the setpoint value to a predefined extent.Such conclusions can be obtained, for example, by automatically analyzing the current forward and the expected forward. Based on the result of the automatic analysis for setting operating parameters of a container treatment machine, at least one manipulated variable of the intelligent algorithm can then be modified, if appropriate.The method may further comprise measuring an actual value of the container property to obtain the measured actual value.The method may further comprise simulating an actual value for obtaining the expected simulated actual value. The simulating may comprise a simulation of a process of the container treatment machine, wherein the process may comprise the corresponding container treatments and wherein the manipulated variables available to the intelligent algorithm may be used.The automatically analyzing may include determining that the amount of difference between current forward and expected forward is greater than a threshold, and the modifying may include automatically expanding a design of experiment (DOE) run dataset by automatically initiating a training run with other set points and automatically de-mapping the smart algorithm. The limit value may be, for example, 50.For example, during operation of the container treatment machine, it can be determined that the current reward falls during operation or sets at a low level. For example, the current forward during operation may decrease to a value below 100 from a maximum of 250 or may be set at a low level below 100 from a maximum of 250. It can mean that the intelligent algorithm cannot provide a better result on the basis of the at least one manipulated variable available to it. The expected, simulated actual value can be target-compliant. The amount of difference between current and expected rewards may be large. It may be useful to expand the data set of the DOE with which the intelligent algorithm operates. The training data set of the intelligent algorithm can be augmented by further set points. The new training run can be initiated and the intelligent algorithm can be depleted thereby.The automatically analyzing may include determining that the amount of difference between current and expected rewards is greater than a threshold and / or that the amount of difference from the target value to the expected simulated actual value is greater than another threshold, and the modifying may include a complete new design-of-experiment run by initiating a new training run with set points of the new design-of-experiment run and automatically de-multiplexing the smart algorithm. The limit value may be, for example, 50. The further limit value may be a difference in light transmittance. A magnitude of the further limit value can be given, for example, by maximum deviations from the setpoint value during a DOE run.For example, during operation of the container treatment machine, it can be determined that the current reward falls during operation or sets at a low level. For example, the current forward during operation may decrease to a value below 100 from a maximum of 250 or may be set at a low level below 100 from a maximum of 250. The difference between current and expected rewards may be large. The expected reward may be good, which may mean that the difference from the target value to the expected simulated actual value may be small. The expected forward may be good at values greater than 180 of 250. The difference from the setpoint value to the expected simulated actual value can be a difference in light transmittance. The order of magnitude can be given, for example, by small deviations from the setpoint value during a DOE run, for example within 10% of the maximum deviation. It may be useful to execute a complete, new DOE run with all set points, for example 126 pieces. Retraining of the smart algorithm may be required based on the currently prevailing situation at the container handling machine. Thus, a new training data set over all set points can result by initiating a new training run and de-multiplexing the intelligent algorithm.The automatic analysis may comprise determining that the amount of the difference between the current forward and the expected forward is less than a limit value and that at least one of the manipulated variables lies at a limit of a value range, and the modifying may comprise automatically enlarging the value range for the at least one manipulated variable and automatic setting via an extrapolation of the at least one manipulated variable. The limit value may be, for example, 50.For example, during operation of the container treatment machine, it can be determined that the current reward falls during operation or sets at a low level. For example, the current forward during operation may decrease to a value below 100 from a maximum of 250 or may be set at a low level below 100 from a maximum of 250. It can mean that the intelligent algorithm cannot provide a better result on the basis of the at least one manipulated variable available to it. The expected, simulated actual value can be real-world. It may be useful to allow a larger value range for the at least one manipulated variable. The settings can be obtained by extrapolation of the manipulated variables.The determination of the current reward and the determination of the expected reward may be performed during operation of the container treatment machine. A temporal behavior of the current and expected reward can thus be analyzed. Likewise, the time profile of the measured actual value and of the expected, simulated actual value can be analyzed.The automatically analyzing may include determining that both the current forward and the expected forward are greater or less than a threshold and that a difference from the target value to the expected simulated actual value is less or greater than another threshold, and the modifying may include automatically shifting base set points defined via base manipulated variables and automatically assigning at least one new manipulated variable offset to the smart algorithm. The further limit value may be a difference in light transmittance. The order of magnitude can be given, for example, by small deviations from the setpoint value during a DOE run, for example within 10% of the maximum deviation or, for example, 60% of the maximum deviation.This can occur when the intelligent algorithm encounters limits of a range of values of the manipulated variables. The current forward may be poor and the measured actual value may deviate greatly from the desired value. The expected forward may also be poor. The current forward may be less than 100 out of a maximum of 250. The measured actual value may be a difference in light transmittance. The order of magnitude may be given, for example, by small deviations from the setpoint value during a DOE run, for example 60% of the maximum deviation. The expected forward may be less than 100 out of a maximum of 250.The base set points may be or include a portion of the set points of a DOE run.The method may further comprise automatically calculating the offset. Alternatively, a manual definition of the offset can be provided, for example by a human operator of the container treatment machine.Automatically calculating the offset may include using boundaries of a range of values of the base set points as new base set points.Furthermore, fault messages of the container treatment machine can be taken into account in the automatic analysis.The error message can comprise malfunctions such as container burst on the blow molding machine or container burst on the filling system or breakage of containers in the container treatment machine.The automatically analyzing may include determining that a number of error messages occurring increases and that the current reward is greater than a threshold, and the modifying may include using new operating parameters and obtaining therefrom new manipulated variables and automatically passing the new manipulated variables to the intelligent algorithm. The limit value may be, for example, 180 of 250.For example, although unfavourable material distributions can lead to a good measured actual value, errors can nevertheless occur, for example due to too much material in the bottom and too little material in the neck of the container or vice versa.The automatically analyzing may include determining that a number of error messages occurring increases and that the current forward is greater than a threshold, and the modifying may include using new operating parameters and a complete new design-of-experiment run by initiating a new training run with set points of the new design-of-experiment run and automatically de-facing the smart algorithm. The limit value may be, for example, 180 of 250.Further provided is an apparatus comprising a computer readable storage device having instructions stored thereon which, when executed by a processor of the apparatus, cause the apparatus to carry out the method as described above or below.Brief description of the FiguresThe accompanying figure presents, by way of example, aspects and / or exemplary embodiments of the invention for better understanding and for illustrative purposes. It shows: FIG. 1 is a block diagram for describing the method for monitoring and modifying an intelligent algorithm for setting operating parameters of a container treatment machine.Detailed Figure DescriptionFIG. 1 shows a block diagram for describing a method for monitoring and modifying an intelligent algorithm 1 for a setting 3 of operating parameters 4 of a container treatment machine 5. A container 6 can be treated in the container treatment machine 5. A measured actual value 8 of a container property can be determined from this container 6 by means of a measurement 7. For this container property, a set value 9 can be known.A current forward 13 can be determined by an input 11 of the measured actual value 8 and an input 12 of the setpoint value 9, from a comparison of the setpoint value 9 with the measured actual value 8.An expected, simulated actual value 10 can result from a simulation of a process of the container treatment machine 5, wherein the process can comprise the corresponding container treatments and wherein the manipulated variables 2 which are available to the intelligent algorithm 1 can be used.An expected reward 16 can be determined by an input 14 of the setpoint value 9 and by an input 15 of the expected simulated actual value 10, from a comparison of the setpoint value 9 to the expected simulated actual value 10.An input 17 of the current forward 13 and an input 18 of the expected forward 16 can be made into an automatic analysis 19. The current forward 13 and the expected forward 16 as well as optionally the one or more error messages 20 may be analyzed in the automatic analysis 19.Based on a result of the automatic analysis 19, at least one of the control variables 2 of the intelligent algorithm 1 can be modified 21 for setting 3 operating parameters 4 of the container treatment machine 5.For example, the container treatment machine 5 can be a blow molding machine in which preforms are formed into containers 6 in a heated blow mold by inflation with a blow pressure. The operating parameters 4 of the blow molding machine can be a heating setting for heating the preforms and the blow molding pressure. With these operating parameters 4, a container 5 can now be formed in the blow molding machine, wherein a container property to be measured can be a wall thickness. By measuring 7 the wall thickness of the container 5, a measured actual value 8 of the wall thickness is obtained. For containers produced with the blow molding machine, a nominal value 9 of the wall thickness can be specified. In addition, an expected, simulated actual value 10 of the wall thickness can be determined. This value can correspond to a theoretical value of the wall thickness that a container should theoretically have if all operating parameters 4 of the blow molding machine and an assumed thickness of the preform and, for example, assumed environmental parameters of the blow molding machine are as they should be theoretical.From a comparison of the setpoint value 9 of the wall thickness to the measured actual value 8 of the wall thickness, the current forward 13 can be determined. From a comparison of the target value 9 of the wall thickness to the expected, simulated actual value 10 of the wall thickness, the expected reward 16 can be determined.An automatic analysis 19 of the current reject 13 and the expected reject 14 can be carried out in order, on the basis of a result of the automatic analysis 19, to optionally carry out a modification 21 of at least one manipulated variable 2 of the intelligent algorithm 1 for a setting 3 of operating parameters 4, i.e. for a setting 3 of the heating temperature and / or of the blowing pressure of the blowing machine. A modification 21 of at least one control variable 2 may be required if the assumed wall thickness of the preforms changes, for example in the case of different batches, and / or if the environmental parameters change significantly during operation of the blow molding machine.The automatic analysis 19 can also be incorporated with one or more fault messages 20 of the blow molding machine, for example if the blow molding process is not carried out correctly.Furthermore, an automatic analysis 19 can also initiate a modification 22 of the intelligent algorithm 1, wherein a new training and deployment replaces the intelligent algorithm 1 with an optimized algorithm.References included in the specificationThis list of documents cited by the applicant has been produced in an automated manner and is only included for the better information of the reader. The list is not part of the German patent application or utility model application. The DPMA does not take any adhesion for any faults or omissions.Patent Literature citedUS 2022 / 0147876 A1

[0002]

Claims

Method for monitoring and modifying an intelligent algorithm (1) for a setting (3) of operating parameters (4) of a container treatment machine (5), the method comprising: - determining a current reward (13) from a comparison of a setpoint value (9) of a container property with a measured actual value (8) of the container property, - determining an expected reward (16) from a comparison of the setpoint value (9) with an expected, simulated actual value (10) of the container property, wherein the expected, simulated actual value (10) is based on the intelligent algorithm (1), - automatically analyzing (19) the current reward (13) and the expected reward (14), modifying (21) at least one control variable (2) of the intelligent algorithm (1) based on a result of the automatic analysis for a setting (3) of operating parameters (4) of a container treatment machine (5).The method according to claim 1, further comprising measuring (7) an actual value of the container property to obtain the measured actual value (8).The method of claim 1 or 2, further comprising simulating an actual value to obtain the expected simulated actual value (10).The method of any of claims 1 to 3, wherein automatically analyzing (19) comprises determining that a difference between current forward (13) and expected forward (16) is less than a threshold, and wherein modifying (21) comprises automatically expanding a set of data of a design-of-experiment run by automatically initiating a training run with further set points and automatically de-mapping the smart algorithm (1).The method of any of claims 1 to 3, wherein automatically analyzing (19) comprises determining that a difference of current forward (13) and expected forward (16) is greater than a threshold and that a difference from the target value (9) to the expected simulated actual value (10) is less than another threshold, and wherein modifying (21) comprises a complete new design-of-experiment run by initiating a new training run with set points of the new design-of-experiment run and automatically de-multiplexing the smart algorithm (1).The method according to any one of claims 1 to 3, wherein the automatic analysis (19) comprises determining that a difference between the current forward (13) and the expected forward (16) is less than a limit value and that at least one of the manipulated variables (2) lies at a boundary of a value range, and wherein the modifying (21) comprises automatically enlarging the value range for the at least one manipulated variable (2) and automatic setting via an extrapolation of the at least one manipulated variable (2).The method according to any one of claims 1 to 3, wherein the automatic analysis (19) comprises determining that both the current forward (13) and the expected forward (16) are greater or less than a threshold and that a difference from the target value (9) to the expected simulated actual value (10) is less or greater than a further threshold, and wherein the modifying (21) comprises an automatic shift of base set points defined via base manipulated variables and an automatic allocation of at least one new manipulated variable (2) with offset to the intelligent algorithm (1).The method of claim 7, further comprising automatically calculating the offset.The method of claim 8, wherein automatically calculating the offset comprises using limits of a range of values of the base set points as new base set points.The method according to any one of claims 1 to 9, wherein fault messages (20) of the container treatment machine (5) are further taken into account in the automatic analysis (19).The method according to claim 10, wherein the automatic analysis (19) comprises determining that a number of error messages (20) occurring increases and that the current forward (13) is greater than a threshold value, and wherein the modifying (21) comprises using new operating parameters (4) and obtaining therefrom new manipulated variables (2) and automatically passing the new manipulated variables (2) to the intelligent algorithm (1).The method of claim 10, wherein automatically analyzing (19) comprises determining that a number of error messages (20) occurring increases and that the current reward (13) is greater than a threshold, and wherein modifying (21) comprises using new operating parameters (4) and a complete new design-of-experiment run by initiating a new training run with set points of the new design-of-experiment run and automatically de-multiplexing the smart algorithm (1).An apparatus comprising a computer readable storage device having instructions stored thereon which, when executed by a processor of the apparatus, cause the apparatus to perform the method of any one of claims 1 to 12.

Images