METHOD AND CONTROLLING SYSTEM FOR CONTROLLING POLYMER VISCOSITY QUALITY

Abstract

A method for controlling polymer viscosity quality in a compounding process of polymers (110) using at least one extruder (111) is disclosed. The method comprises: a) at least one measurement step (112), wherein at least one influence variable affecting viscosity of the compound is determined by using at least one sensor (114); b) at least one prediction step (116), wherein an expected viscosity (117) of the compound is determined considering the influence variable by using at least one prediction unit (118), wherein the prediction unit (118) comprises at least one analysis tool comprising at least one trained model; c) at least one evaluation step (120), wherein the expected viscosity (117) of the compound is compared to at least one pre-defined and/or pre-determined threshold value, wherein at least one item of output information is generated depending on said comparison; and d) at least one control step (122), wherein the item of output information is displayed using at least one display device (124), wherein the output information comprises at least one handling recommendation (126) for at least one setting of the extruder (111). Further disclosed are a computer program, specifically an application, and a controlling system (138) for controlling polymer viscosity quality in a compounding process of polymers (110).

Claims

1. A method for controlling polymer viscosity quality in a compounding process of polymers (110) using at least one extruder (111), wherein the compound at least partially comprises at least one thermoplastic material selected from the group consisting of: polybutylene terephthalate (PBT); polyethylene terephthalate (PET); polyamide (PA); polystyrene (PS), the method comprising: a) at least one measurement step (112), wherein at least one influence variable affecting viscosity of the compound is determined by using at least one sensor (114); b) at least one prediction step (116), wherein an expected viscosity (117) of the compound is determined considering the influence variable by using at least one prediction unit (118), wherein the prediction unit (118) comprises at least one analysis tool comprising at least one trained model; c) at least one evaluation step (120), wherein the expected viscosity (117) of the compound is compared to at least one pre-defined and/or pre-determined threshold value, wherein at least one item of output information is generated depending on said comparison; and d) at least one control step (122), wherein the item of output information is displayed using at least one display device (124), wherein the output information comprises at least one handling recommendation (126) for at least one setting of the extruder (111).

2. The method according to the preceding claim, wherein the influence variable is selected from the group consisting of: a head pressure; a torque; a temperature; a transfer rate; a rotational speed; a shear rate; a mass flow; an electric power consumption.

3. The method according to any one of the preceding claims, wherein the method comprises generating at least one database, wherein the database comprises one or both of quantitative and qualitative information on the compounding process, wherein the information is selected from the group consisting of: sensor data; a material; a process parameter of the compounding process; a pre-determined target viscosity value.

4. The method according to any one of the preceding claims, wherein the trained model comprises at least one model selected from the group consisting of: at least one random forest algorithm; at least one neural network; at least one linear model; at least one non-linear model; at least one grey box model; at least one elastic net algorithm.

5. The method according to any one of the preceding claims, wherein step a) further comprises at least one pre-processing step, wherein data determined by the sensor (114) is pre-processed, wherein the pre-processing comprises eliminating of outliers, wherein outliers are eliminated from the data determined by the sensor (114) by using at least one clustering algorithm.

6. The method according to any one of the preceding claims, wherein: i) if the at least one expected viscosity (117) of the compound falls within a predefined tolerance range of the threshold value, the handling recommendation (126) comprises information on maintaining the setting of the extruder (111); and ii) if the at least one expected viscosity (117) of the compound differs from the threshold value by more than the predefined tolerance range, the handling recommendation (126) comprises information on changing the setting of the extruder (111).

7. The method according to the preceding claim, wherein the predefined tolerance range is one or both of a relative and an absolute tolerance, wherein the predefined tolerance range is ±5%, preferably ±2%, and/or ±3 ml/g, preferably ±2 ml/g.

8. The method according to any one of the preceding claims, wherein the output information further comprises at least one information selected from the group consisting of: an information on the expected viscosity of the compound; an information on historical viscosity of the compound such as measured viscosity and/or expected viscosity.

9. The method according to any one of the preceding claims, wherein the method further comprises regulating at least one process parameter of the compounding process depending on the output information, wherein the process parameter comprises at least one rotational speed of the extruder (111).

10. A computer program for controlling polymer viscosity quality in a compounding process of polymers (110), wherein in the compounding process at least one extruder (111) is used, wherein the compound at least partially comprises at least one thermoplastic material selected from the group consisting of: polybutylene terephthalate (PBT); polyethylene terephthalate (PET); polyamide (PA); polystyrene (PS), the computer program comprises instructions which, when the program is executed by a computer or computer network, cause the computer or computer network to carry out the following steps retrieving at least one influence variable affecting viscosity of the compound; predicting an expected viscosity (117) of the compound considering the influence variable by using at least one analysis tool comprising at least one trained model; comparing the expected viscosity (117) of the compound to at least one pre-defined and/or pre-determined threshold value, wherein at least one item of output information is generated depending on said comparison; and displaying the item of output information, wherein the output information comprises at least one handling recommendation (126) for at least one setting of the extruder (111).

11. A controlling system (138) for controlling polymer viscosity quality in a compounding process of polymers (110), wherein the compound at least partially comprises at least one thermoplastic material selected from the group consisting of: polybutylene terephthalate (PBT); polyethylene terephthalate (PET); polyamide (PA); polystyrene (PS), wherein in the compounding process at least one extruder (111) is used, wherein the controlling system (138) comprises at least one sensor (114) configured for determining at least one influence variable affecting viscosity of the compound, wherein the controlling system (138) comprises at least one prediction unit (118) configured for determining an expected viscosity (117) of the compound considering the influence variable, wherein the prediction unit (118) comprises at least one analysis tool comprising at least one trained model, wherein the controlling system (138) comprises at least one evaluation unit (140) configured for comparing the expected viscosity (117) of the compound to at least one pre-defined and/or pre-determined threshold value, wherein the evaluation unit (140) is configured for generating at least one item of output information depending on said comparison, wherein the controlling system (138) comprises at least one display device (124) configured for displaying the item of output information, wherein the output information comprises at least one handling recommendation (126) for at least one setting of the extruder (111).

12. The controlling system (138) according to the preceding claim, wherein the sensor (114) comprises at least one sensor (114) selected from the group consisting of: a pressure sensor, specifically a melt pressure sensor; an inductive sensor, specifically an inductive slot sensor; a thermocouple.

13. The controlling system (138) according to any one of the two preceding claims, wherein the controlling system (138) further comprises a regulator unit configured for regulating at least one process parameter of the compounding process depending on the output information, wherein the process parameter comprises at least one rotational speed of the extruder (111).

14. The controlling system (138) according to any one of the three preceding claims, wherein the controlling system (138) is configured for performing the method for controlling at least one quality of a compound in a compounding process according to any one of the preceding claims referring to a method.

Description

SHORT DESCRIPTION OF THE FIGURES

[0107] Further optional features and embodiments will be disclosed in more detail in the subsequent description of embodiments, preferably in conjunction with the dependent claims. Therein, the respective optional features may be realized in an isolated fashion as well as in any arbitrary feasible combination, as the skilled person will realize. The scope of the invention is not restricted by the preferred embodiments. The embodiments are schematically depicted in the Figures. Therein, identical reference numbers in these Figures refer to identical or functionally comparable elements.

[0108] In the Figures:

[0109] FIG. 1: shows a flow chart of an embodiment of a method for controlling polymer viscosity quality in a compounding process of polymers using at least one extruder;

[0110] FIG. 2: shows an embodiment of an interface of a computer program for controlling polymer viscosity quality in a compounding process of polymers; and

[0111] FIG. 3: shows an embodiment of a controlling system for controlling polymer viscosity quality in a compounding process of polymers.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0112] In a first aspect of the invention, a method for controlling polymer viscosity quality in a compounding process of polymers 110 using at least one extruder 111 is disclosed. FIG. 1 shows a flow chart of an embodiment of the method. The method comprises the following steps, which may specifically be performed in the given order. Still, a different order may also be possible. It may further be possible to perform two or more of the method steps fully or partially simultaneously. It may further be possible to perform one or more method steps or even all of the method steps once or repeatedly. The method may comprise additional method steps, which are not listed herein. The method steps are the following:

[0113] a) at least one measurement step 112, wherein at least one influence variable affecting viscosity of the compound is determined by using at least one sensor 114;

[0114] b) at least one prediction step 116, wherein an expected viscosity 117 of the compound is determined considering the influence variable by using at least one prediction unit 118, wherein the prediction unit 118 comprises at least one analysis tool comprising at least one trained model;

[0115] c) at least one evaluation step 120, wherein the expected viscosity 117 of the compound is compared to at least one pre-defined and/or pre-determined threshold value, wherein at least one item of output information is generated depending on said comparison; and

[0116] d) at least one control step 122, wherein the item of output information is displayed using at least one display device 124, wherein the output information comprises at least one handling recommendation 126 for at least one setting of the extruder 111.

[0117] The method may further comprise automatically retrieving information from a database, such as from a sorted accumulation of information, and providing the information to the prediction unit 118. In particular, the retrieving of information from the database may be performed between the measurement step 112 and the prediction step 116.

[0118] The measurement step 112 may further comprise at least one pre-processing step, wherein in the pre-processing step, data determined by the sensor 114 may be pre-processed. In particular, the pre-processing may comprise an elimination of outliers. Thus, as an example, outliers may be eliminated from the data determined by the sensor 114 by using at least one clustering algorithm. In particular, the at least one clustering algorithm may be or may comprise one or more of an expectation-maximization algorithm, a k-means algorithm and a k-median algorithm.

[0119] The evaluation step 120 may further comprise comparing the expected viscosity 117 of the compound with at least one target specification 121. Therein the target specification may specifically be a customer specification, such as a specification given and/or set by a customer and/or buyer of the compound.

[0120] FIG. 2 shows an embodiment of an interface of a computer program for controlling polymer viscosity quality in a compounding process of polymers. Specifically, FIG. 2 illustrates an interface of a computer or computer network, when an embodiment of a computer program, e.g. an application, for controlling polymer viscosity quality in a compounding process of polymers 110 is executed on the computer or computer network. As an example, the interface may specifically be or may comprise a display device 124, wherein on the display device 124 output information comprising at least one handling recommendation 126 for at least one setting of the extruder 111 may be displayed. As an example, the handling recommendation 126 may be given in text, such as by using letters of an arbitrary language. However, other forms of displaying handling recommendation 126 may be possible, such as by using one or more of a pictogram, a color code, a video-sequence and the like. As an example, the displayed handling recommendation 126 may even be accompanied by at least one audio signal, such as by an acoustic signal, e.g. an acoustic warning signal.

[0121] Further, on the interface, e.g. on the display device 124, expected viscosities 117 of more than one compounding process of polymers 110 using at least one extruder 111 may be shown. Thus, as an example, in a first box 128 the expected viscosity 117 of a first compounding process of polymers 110 using at least one extruder 111, e.g. at least one first extruder, may be illustrated, and in a second box 130, the expected viscosity 117 of a second compounding process of polymers 110 using at least one extruder 111, e.g. at least one second extruder, may be illustrated. Further information may be illustrated in the first box 128 and in the second box 130. Such further information may for example be or may comprise one or more of a viscosity deviation value 132 and a specification interval 134. Therein, the specification interval 134 may, for example, be an admissibility range of the viscosity, such as a specification limit and/or boundary. The viscosity deviation value 132, as an example, may specifically be a discrepancy between the expected viscosity 117 and a target specification 121, such as a difference between the expected viscosity 117 and a mean value of the specification interval 134, e.g. a difference between the expected viscosity 117 and a target viscosity value. In a third box 136, at least one input parameter, such as a production version and a current process order, may be illustrated. As an example, based on the input parameters, the computer program may automatically select respective information from a database, such as information on one or more of the trained model, the polymer viscosity quality and a product, e.g. a compound to be produced using the respective compounding process of polymers. Such a selection may be automatically performed, such as in a pre-set interval P. As an example, the pre-set interval P for automatically performing the selection may be 0 min<P≤5 min, specifically 0.5 min≤P≤3 min, more specifically 1 min≤P≤2 min.

[0122] As an example, a color of the first box 128 and the second box 130 may be adapted according to a confidence interval. Thus, as an example, in case a 95%-confidence interval may be within the specification interval 134, the color of the respective box, e.g. of the first box 128 and/or of the second box 130, may comprise a green color. In case the specification interval 134 is violated by at least one limit of the 95%-confidence interval, in particular in case the 95%-confidence interval reaches and/or cuts the specification interval 134, the color of the respective box may be or may comprise a yellow or orange color. Further, in case the 95%-confidence interval may be outside of the specification interval 134, the respective box may comprise a red color.

[0123] In FIG. 3, an embodiment of a controlling system 138 for controlling polymer viscosity quality in a compounding process of polymers 110 is illustrated, wherein in the compounding process at least one extruder 111 is used. The controlling system comprises at least one sensor 114 configured for determining at least one influence variable affecting viscosity of the compound. The controlling system 138 further comprises at least one prediction unit 118 configured for determining an expected viscosity 117 of the compound considering the influence variable, wherein the prediction unit 118 comprises at least one analysis tool comprising at least one trained model. The controlling system 138 further comprises at least one evaluation unit 140 configured for comparing the expected viscosity 117 of the compound to at least one pre-defined and/or pre-determined threshold value, wherein the evaluation unit 140 is configured for generating at least one item of output information depending on said comparison. The controlling system 138 further comprises at least one display device 124 configured for displaying the item of output information. The output information comprises at least one handling recommendation 126 for at least one setting of the extruder 111.

[0124] The controlling system 138 may further comprise a regulator unit (not illustrated in FIG. 3) configured for regulating at least one process parameter of the compounding process depending on the output information. In particular, depending on the item of output information the at least one process parameter of the compounding process, e.g. the extruder setting, may be regulated by using the regulator unit. Specifically, the process parameter may be or may comprise a rotational speed of the extruder 111. Thus, as an example, when increasing the rotational speed, the viscosity, such as the polymer viscosity, may be reduced and vice versa.

[0125] The controlling system 138 may be configured for performing the method for controlling at least one quality of a compound in a compounding process as illustrated in FIG. 1.

LIST OF REFERENCE NUMBERS

[0126] 110 polymer

[0127] 111 extruder

[0128] 112 measurement step

[0129] 114 sensor

[0130] 116 prediction step

[0131] 117 expected viscosity

[0132] 118 prediction unit

[0133] 120 evaluation step

[0134] 121 target specification

[0135] 122 control step

[0136] 124 display device

[0137] 126 handling recommendation

[0138] 128 first box

[0139] 130 second box

[0140] 132 viscosity deviation value

[0141] 134 specification interval

[0142] 136 third box

[0143] 138 controlling system

[0144] 140 evaluation unit

LIST OF CITED REFERENCES

[0145] “A novel approach to dynamic modelling of polymer extrusion for improved process control” on pages 617-628 in “proceedings of the Institution of Mechanical Engineers, Vol. 221 Part I: J. Systems and Control Engineering”, by McAfee and Thompson

[0146] “Soft-sensor for real-time monitoring of melt viscosity in polymer extrusion process” in “49th IEEE Conference on decision and Control”, by Liu et al.

[0147] “Beständigkeit von Kunststoffen”, Munich: Hanser 2007, by Gottfried W. Ehrenstein and Sonja Pongratz

[0148] Computers & Chemical Engineering, 60, Pages 86 to 101, 2014, by Von Stoch et al.

[0149] “Random Forests” in Machine Learning 45.1, October 2001, by Leo Breiman “The Elements of Statistical Learning”, Springer New York, 2009, by Trevor Hastie, Robert Tibshirani and Jerome Friedman

[0150] “Conditional variable importance for random forests” in BMC Bioinformatics 9.1, July 2008, page 307, by Carolin Strobl et al; [0151] WO 2008/040943 A2; [0152] ES 2 331 720 A1; [0153] EP 3 020 530 A1;

[0154] Kumar A. et al.: “A model based approach for estimation and control for polymer compounding”, Proceedings of the 2003 IEEE International Conference on Control Applications, CCA 2003, ISTANBUL, TURKEY, Jun. 23-25, 2003; NEW YORK, N.Y.: IEEE, US, vol. I, 23 Jun. 2003, pages 729-735, XP010652701, ISBN: 978-0-7803-7729-5.