Method and device for texturing a synthetic thread

Patent metadata

Field	Value
Publication number	US-2021188590-A1
Application number	US-201816638702-A
Country	US
Kind code	A1
Filing date	Jun 13, 2018
Priority date	Aug 23, 2017
Publication date	Jun 24, 2021
Grant date	—

Abstract

Techniques involve texturing a synthetic thread. The thread is pulled off a feed bobbin, which is connected via the thread end thereof to a beginning of a thread of a reserve bobbin by way of a thread knot. In order to monitor the texturing, a thread tension of the thread is measured and analyzed in a measuring point. Additionally, measuring signals of the thread tension are analyzed at the measuring point using a machine learning program, in order to identify the thread knot. To this end, a device has a diagnostic unit, which interacts with the thread tension measuring device in such a way that the measuring signals of the thread tension can be analyzed by means of a machine learning program for identifying a thread knot.

First claim

1 . A method for texturing a synthetic thread, in which said thread is drawn off a supply bobbin and stretched,
wherein a thread end of the supply bobbin is connected to a thread start of a reserve bobbin by a thread knot, wherein a thread tension of the thread is measured and analyzed continuously in a measuring station in order to monitor the texturing, and wherein, in order to identify the thread knot at the measuring station, the measured signals of the thread tension are analyzed by a machine learning program.

2 . The method as claimed in claim 1 , wherein a chronological sequence of measured signals of the thread tension is detected as an analysis graph and analyzed.

3 . The method as claimed in claim 1 , wherein a chronological sequence of measured signals of the thread tension when a threshold value of the thread tension is overshot is detected as a fault graph and analyzed.

4 . The method as claimed in claim 3 , wherein analysis of the measured signals of the thread tension is carried out by at least one machine learning algorithm of the machine learning program.

5 . The method as claimed in claim 4 , wherein the thread knot is identified by the machine learning algorithm from analyzed analysis graphs or analyzed fault graphs.

6 . The device as claimed in claim 5 , wherein the fault graphs are assigned to multiple fault categories, and wherein the thread knot is one of the fault graph categories.

7 . The method as claimed in claim 6 , wherein other fault graph categories are assigned to at least one of a specific process fault, a specific operating error, a specific interfering parameter, and a specific product fault.

8 . The method as claimed in claim 7 , wherein following identification of the thread knot or following the assignment of the fault graph to one of the fault graph categories, a control command relating to a process change is triggered.

9 . A device for texturing a synthetic thread, comprising:
a creel for holding a supply bobbin and a reserve bobbin, multiple delivery devices, a texturing unit, a winding station, a thread tension measuring device for measuring a thread tension at a measuring station, and a diagnostic unit, wherein the thread tension measuring device cooperates with a diagnostic unit in such a way that measured signals of the thread tension (T) can be analyzed by a machine learning program to identify a thread knot.

10 . The device as claimed in claim 9 , wherein the diagnostic unit has a programmable learning processor for executing the machine learning program.

11 . The device as claimed in claim 10 , wherein the learning processor is coupled to an input unit, by means of which one or more determined analysis graphs of the thread tension can be read.

12 . The device as claimed in claim 11 , wherein the learning processor is coupled to an output unit, by means of which at least one of an identification of the thread knot and an assignment of the analyzed fault graphs to one of the fault graph categories can be visualized.

13 . The device as claimed in claim 12 , wherein the learning processor has a neural network for executing the machine learning program.

14 . The device as claimed in claim 13 , wherein the learning processor is separated physically from the input unit and the output unit.

15 . The device as claimed in claim 14 , wherein the diagnostic unit is connected to a machine control unit, by means of which a control command relating to the process change can be carried out.

Assignees

• Oerlikon Textile Gmbh & Co Kg

Inventors

• Huthmacher Jörg

Classifications

• G01L5/10

  using electrical means · CPC title

• B65H2511/52

  Defective operating conditions · CPC title

• B65H2701/31

  Textiles threads or artificial strands of filaments · CPC title

• B65H59/40 Primary

  Applications of tension indicators · CPC title

• G01L5/102

  using sensors located at a non-interrupted part of the flexible member · CPC title