False twist machine abnormality early warning method
By measuring input parameters and training a machine learning model in a false twisting machine, the predicted yield of the false twisting machine is predicted, which solves the problem of quality decline and downtime caused by the instability of raw materials and environment, and realizes timely maintenance and improved production stability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- TAIWAN TEXTILE RESEARCH INSTITUTE
- Filing Date
- 2025-07-07
- Publication Date
- 2026-06-05
AI Technical Summary
The quality decline and abnormal shutdown problems of false twisting machines during production are difficult to effectively predict and prevent with existing technologies due to factors such as inconsistent raw material quality, unstable ambient temperature and humidity of the heater, and wear of the friction disc.
By measuring the input parameters and actual yield of the reference yarn using a false twisting machine, a machine learning model is trained to predict the predicted yield of the false twisting machine. Based on the prediction results, a decision is made on whether to carry out maintenance, including maintenance measures such as replacing the friction disc.
This enabled timely maintenance of the false twisting machine, reduced abnormal downtime, and improved production stability and efficiency.
Smart Images
Figure CN122147582A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to a false twisting machine abnormality early warning technology, and more particularly to a false twisting machine abnormality early warning method. Background Technology
[0002] False twisting machines convert yarn spindles into yarn cakes through a false twisting production process. However, the false twisting production process often faces challenges such as inconsistent raw material quality, instability of heaters due to ambient temperature and humidity, and wear on friction discs and belts, all of which can lead to quality degradation, or even abnormal shutdowns such as yarn breakage or complete machine stoppage. Consequently, inspecting abnormal false twisting machines is time-consuming and costly. Therefore, developing technologies to overcome these problems is an important task in this field. Summary of the Invention
[0003] This disclosure includes a method for early warning of anomalies in a false twisting machine. The method includes: converting multiple reference spindles into multiple reference yarn cakes using a false twisting machine; measuring multiple sets of reference input parameters for the multiple reference yarns of the reference spindles within the false twisting machine; measuring multiple actual yields of the reference yarn cakes; training a machine learning model based on the reference input parameter sets and the actual yields; converting multiple spindles into multiple yarn cakes using the false twisting machine; measuring multiple sets of input parameters for the multiple yarns of the spindles within the false twisting machine; generating multiple predicted yields based on the input parameter sets using the machine learning model; and determining whether to repair the false twisting machine based on the predicted yields.
[0004] In some implementations, the reference input parameter set includes multiple reference yarn breakage numbers, multiple reference maximum variation rates of the reference yarn, multiple reference tension alarm counts of the reference yarn, and multiple reference untwisting coefficients of the reference yarn, and the input parameter set includes multiple yarn breakage numbers, multiple maximum variation rates of the yarn, multiple tension alarm counts of the yarn, and multiple untwisting coefficients of the yarn.
[0005] In some embodiments, the false twisting machine abnormality early warning method further includes: untwisting a reference yarn and a yarn through a first friction disc in the false twisting machine; measuring multiple reference rotation speeds of the first friction disc and multiple reference movement speeds of the reference yarn when the first friction disc is untwisting the reference yarn; generating a reference untwisting coefficient based on the reference rotation speeds and the multiple reference movement speeds of the reference yarn; measuring multiple rotation speeds of the first friction disc and multiple movement speeds of the yarn when the first friction disc is untwisting the yarn; and generating the untwisting coefficient based on the rotation speeds and the multiple movement speeds of the yarn.
[0006] In some implementations, the reference untwisting coefficient decreases when the reference moving speed increases, increases when the reference rotating speed increases, decreases when the moving speed increases, and increases when the rotating speed increases.
[0007] In some implementations, one of the reference untwisting coefficients is equal to the corresponding reference rotational speed divided by the corresponding reference moving speed, and one of the untwisting coefficients is equal to the corresponding rotational speed divided by the corresponding moving speed.
[0008] In some implementations, the direction of rotational speed is perpendicular to the direction of movement speed.
[0009] In some embodiments, the false twisting machine abnormality early warning method further includes: processing a first spindle, a second spindle, and a third spindle in a filament spindle sequentially and continuously using the false twisting machine; generating a first predicted yield, a second predicted yield, and a third predicted yield, respectively, among the predicted yields of the first, second, and third spindles; comparing the first, second, and third predicted yields with a preset yield; and repairing the false twisting machine when the first predicted yield is less than the preset yield, the second predicted yield is less than the preset yield, and the third predicted yield is less than the preset yield.
[0010] In some implementations, repairing a false twisting machine includes replacing the first friction disc with a second friction disc, which is different from the first friction disc.
[0011] In some embodiments, the false twisting machine abnormality warning method further includes: when the first predicted yield is greater than the preset yield, the second predicted yield is greater than the preset yield, or the third predicted yield is greater than the preset yield, after the false twisting machine processes the third spindle, processing a fourth spindle among the spindles using the false twisting machine.
[0012] In some embodiments, the false twisting machine abnormality warning method further includes: generating a fourth predicted yield corresponding to the fourth spindle in the predicted yield; comparing the fourth predicted yield with a preset yield; and replacing the first friction disc via the second friction disc when the second predicted yield is less than the preset yield, the third predicted yield is less than the preset yield, and the fourth predicted yield is less than the preset yield. Attached Figure Description
[0013] Figure 1 This is a schematic diagram illustrating a false twisting machine abnormality early warning system according to some embodiments disclosed herein;
[0014] Figure 2 A flowchart illustrating a method for predicting fabric stiffness according to some embodiments of this disclosure;
[0015] Figure 3 A flowchart illustrating other operations of the fabric stiffness prediction method according to some embodiments of this disclosure;
[0016] Figure 4A flowchart illustrating other operations of the fabric stiffness prediction method according to some embodiments of this disclosure;
[0017] Figure 5 A flowchart illustrating other operations of the fabric stiffness prediction method according to some embodiments of this disclosure.
[0018] [Symbol Explanation]
[0019] 100: False Twist Machine Abnormal Early Warning System
[0020] 110: Processor
[0021] 120: False twisting machine
[0022] RYS1: Reference spindle
[0023] RS1: Reference wire
[0024] RCS1: Reference Silk Pancake
[0025] MD1: Machine Learning Model
[0026] AYS1: Silk Ingot
[0027] ACS1: Silk Cake
[0028] AS1: Silk thread
[0029] FD1: First friction disc
[0030] X, Y, Z: Direction
[0031] FD2: Second friction disc
[0032] 200: Early Warning Method for Abnormalities in False Twist Machines
[0033] OP21~OP28, OP31~OP35, OP41~OP44, OP51~OP54: Operation Detailed Implementation
[0034] Throughout this document, although terms such as "first," "second," etc., are used to describe different elements, these terms are merely used to distinguish elements or operations described using the same technical terms. Unless the context clearly indicates otherwise, these terms do not specifically refer to or imply any order or sequence, nor are they intended to limit this disclosure.
[0035] Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains. It will be further understood that terms such as those defined in commonly used dictionaries should be interpreted as having meanings consistent with their meanings in the relevant technical context and this disclosure, and will not be interpreted as having idealized or overly formal meanings unless expressly defined as such herein.
[0036] The terminology used herein is for the purpose of describing particular embodiments only and is not restrictive. As used herein, unless the content clearly indicates otherwise, the singular forms “a,” “an,” and “the” are intended to include the plural forms, including “at least one.” “Or” means “and / or.” As used herein, the term “and / or” includes any and all combinations of one or more of the associated listed items. It should also be understood that, when used in this specification, the terms “comprising” and / or “including” specify the presence of the stated features, areas, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, areas, integrals, steps, operations, elements, components, and / or combinations thereof.
[0037] The following describes several embodiments of this disclosure with reference to the accompanying drawings. For clarity, many practical details will be described in the following description. However, it should be understood that these practical details should not be used to limit this disclosure. That is, in some embodiments of this disclosure, these practical details are not essential. Furthermore, for the sake of simplicity in the drawings, some conventional structures and elements will be shown in a simple schematic manner.
[0038] Figure 1 This is a schematic diagram illustrating a false twisting machine abnormality early warning system 100 according to some embodiments of this disclosure. Figure 1 As shown, the false twisting machine abnormality early warning system 100 includes a processor 110 and a false twisting machine 120.
[0039] In some embodiments, the false twister 120 is used to process a plurality of reference spindles RYS1 to convert the reference spindles RYS1 into a plurality of reference cakes RCS1.
[0040] In some implementations, processor 110 is coupled to false twister 120 and used to monitor the operation of false twister 120 to determine whether to repair false twister 120.
[0041] In some embodiments, the false twister 120 is used to measure multiple reference yarns RS1 of the reference yarn RYS1 in the false twister 120 when processing the reference yarn RYS1, and to transmit the reference input parameter sets to the processor 110.
[0042] In some embodiments, the false twister 120 is further configured to measure multiple reference actual yields of the reference yarn cake RCS1 when generating the reference yarn cake RCS1, and transmit the reference actual yields to the processor 110.
[0043] In some implementations, processor 110 trains a machine learning model MD1 based on a set of reference input parameters and a reference actual yield. For example, machine learning model MD1 generates multiple reference predicted yields based on the set of reference input parameters. Then, processor 110 generates a loss function based on the difference between the reference predicted yield and the reference actual yield. The larger the difference between the reference predicted yield and the reference actual yield, the larger the loss function. Processor 110 adjusts the weight parameters of machine learning model MD1 based on the changes in the loss function, and generates new reference predicted yields again after adjusting the weight parameters to generate a new loss function. Processor 110 can repeatedly adjust the weight parameters and generate corresponding reference predicted yields to reduce the loss function. In some implementations, machine learning model MD1 is trained when the loss function reaches its minimum value.
[0044] In some implementations, the machine learning model MD1 can be implemented using the XGBoost algorithm.
[0045] In some implementations, after the machine learning model MD1 has been trained, the false twisting machine 120 is used to process multiple spools AYS1 to convert the spools AYS1 into multiple yarn cakes ACS1. The false twisting machine 120 is further used to measure multiple sets of input parameters of multiple yarns AS1 of the spools AYS1 in the false twisting machine 120.
[0046] In some implementations, the processor 110 is used to generate multiple predicted yields based on a set of input parameters using a machine learning model MD1, and to determine whether to repair the false twisting machine 120 based on the predicted yields.
[0047] like Figure 1 As shown, the false twisting machine 120 includes a first friction disc FD1. In some embodiments, the first friction disc FD1 is used to rotate in the XY plane to untwist the reference yarn RS1 of the reference spindle RYS1. Specifically, the reference yarn RS1 of the reference spindle RYS1 moves along the Z direction at multiple reference moving speeds. Simultaneously, the first friction disc FD1 contacts the reference yarn RS1 of the reference spindle RYS1 and is used to rotate in the XY plane at corresponding multiple reference rotational speeds to untwist the corresponding reference yarn RS1. The X, Y, and Z directions are perpendicular to each other. In other words, the direction of the reference rotational speed of the first friction disc FD1 is perpendicular to the direction of the reference moving speed of the reference yarn RS1.
[0048] Similarly, the first friction disc FD1 rotates in the XY plane to untwist the yarn AS1 of the spindle AYS1. Specifically, the yarn AS1 of the spindle AYS1 moves along the Z direction at multiple speeds. Simultaneously, the first friction disc FD1 contacts the yarn AS1 of the spindle AYS1 and rotates in the XY plane at corresponding speeds to untwist the corresponding yarn AS1. The X, Y, and Z directions are perpendicular to each other. In other words, the direction of the rotational speed of the first friction disc FD1 is perpendicular to the direction of the moving speed of the yarn AS1.
[0049] In some implementations, the reference input parameter set includes multiple reference yarn breakage numbers of reference yarn RS1, multiple reference maximum variation rates of reference yarn RS1, multiple reference tension alarm counts of reference yarn RS1, and multiple reference untwisting coefficients of reference yarn RS1.
[0050] In some embodiments, during the process of converting the reference spindle RYS1 into the reference yarn cake RCS1 using the false twisting machine 120, the false twisting machine 120 is used to measure the number of breaks of the reference yarn RS1 of the reference spindle RYS1 to calculate the reference yarn breakage count. The false twisting machine 120 is further used to measure the change in tension of the reference yarn RS1 to generate a reference maximum variability rate.
[0051] In some embodiments, during the process of converting the reference spindle RYS1 into the reference yarn cake RCS1 using the false twister 120, the false twister 120 is further used to measure the tension of the reference yarn RS1 to generate a reference tension alarm count. For example, when the tension of the reference yarn RS1 exceeds a preset tension range and persists for more than a preset time length, the reference tension alarm count increases by 1.
[0052] In some embodiments, the false twisting machine 120 is further configured to generate a reference untwisting coefficient based on the reference rotational speed of the first friction disc FD1 and the reference moving speed of the reference yarn RS1 of the reference spindle RYS1. In some embodiments, the reference untwisting coefficient decreases when the reference moving speed increases, and increases when the reference rotational speed increases. For example, one of the reference untwisting coefficients is equal to the corresponding value of the reference rotational speed divided by the corresponding value of the reference moving speed.
[0053] In some implementations, the input parameter set includes multiple yarn breakage numbers of yarn AS1, multiple maximum variation rates of yarn AS1, multiple tension alarm counts of yarn AS1, and multiple untwisting coefficients of yarn AS1.
[0054] In some embodiments, during the process of converting the spindle AYS1 into the yarn cake ACS1 using the false twisting machine 120, the false twisting machine 120 is used to measure the number of breaks in the yarn AS1 of the spindle AYS1 to calculate the number of broken yarns. The false twisting machine 120 is also used to measure the change in tension of the yarn AS1 to produce the maximum rate of variation.
[0055] In some embodiments, during the process of converting the spindle AYS1 into the yarn cake ACS1 using the false twister 120, the false twister 120 is further used to measure the tension of the yarn AS1 to generate a tension alarm count. For example, when the tension of the yarn AS1 exceeds a preset tension range and continues for more than a preset time length, the tension alarm count increases by 1.
[0056] In some embodiments, the false twisting machine 120 is further configured to generate a detwisting coefficient based on the rotational speed of the first friction disc FD1 and the moving speed of the yarn AS1 on the spindle AYS1. In some embodiments, the detwisting coefficient decreases when the moving speed increases, and increases when the rotational speed increases. For example, one of the detwisting coefficients is equal to the corresponding rotational speed divided by the corresponding moving speed.
[0057] In some embodiments, the false twisting machine 120 is used to measure the rate of variation of the tension of a reference yarn package RCS1 to generate a reference predicted yield. In some embodiments, the reference predicted yield decreases when the rate of variation of the tension of the reference yarn package RCS1 increases, and increases when the rate of variation of the tension of the reference yarn package RCS1 decreases. In some embodiments, the reference predicted yield may be equal to the rate of variation of the tension of the reference yarn package RCS1 divided by a preset rate of variation.
[0058] In some embodiments, the false twisting machine malfunction warning system 100 further includes a second friction disc FD2. When servicing the false twisting machine 120, the first friction disc FD1 can be removed from the false twisting machine 120, and the second friction disc FD2 can be installed in the original position of the first friction disc FD1 to replace the first friction disc FD1.
[0059] Figure 2 This is a flowchart illustrating a false twisting machine abnormality early warning method 200 according to some embodiments disclosed herein. Figure 2 As shown, the false twisting machine abnormality early warning method 200 includes operations OP21 to OP28. In some embodiments, operations OP21 to OP28 can be performed by, for example... Figure 1 The false twisting machine abnormality warning system 100 shown is executed. However, the embodiments disclosed herein are not limited to this, and operations OP21 to OP28 can also be executed by other false twisting machine abnormality warning systems.
[0060] During operation OP21, multiple reference spindles RYS1 are converted into multiple reference yarn cakes RCS1 via false twisting machine 120.
[0061] During operation OP22, the false twisting machine 120 measures multiple reference yarns RS1 of the reference spindle RYS1 in multiple reference input parameter groups in the false twisting machine 120.
[0062] During operation of OP23, the false twisting machine 120 measures multiple reference actual yields of the reference yarn cake RCS1.
[0063] During operation OP24, processor 110 trains machine learning model MD1 based on reference input parameter set and reference actual yield.
[0064] During operation OP25, multiple AYS1 spindles are converted into multiple ACS1 yarn cakes via false twisting machine 120.
[0065] During operation OP26, the false twisting machine 120 measures multiple sets of input parameters for multiple filaments AS1 of the spindle AYS1 in the false twisting machine 120.
[0066] During operation OP27, processor 110 uses machine learning model MD1 to generate multiple predicted yields based on the input parameter set.
[0067] During operation OP28, processor 110 determines whether to repair false twisting machine 120 based on the predicted yield.
[0068] In some practices, false twisting machines can convert yarn spindles into yarn cakes through a false twisting production process. However, the false twisting production process often faces challenges such as inconsistent raw material quality, instability of heaters due to ambient temperature and humidity, or wear and tear on friction discs and belts, all of which can lead to a decline in quality and even abnormal shutdowns such as yarn breakage or complete machine stoppage.
[0069] Compared to the above approach, in the embodiment disclosed herein, the processor 110 monitors the input parameter set of the spindle AYS1 to generate a predicted yield, and decides whether to repair the false twisting machine 120 based on the predicted yield. In this way, the false twisting machine abnormality early warning system 100 can repair the false twisting machine 120 in a timely manner.
[0070] Figure 3 The flowchart illustrates other operations of the false twisting machine abnormality early warning method 200 according to some embodiments of this disclosure. For example... Figure 3 As shown, the false twisting machine abnormality early warning method 200 may further include operating OP31 to OP35. In some embodiments, operating OP31 to OP35 can be achieved through methods such as... Figure 1 The false twisting machine abnormality warning system 100 shown is executed. However, the embodiments disclosed herein are not limited to this, and operations OP31 to OP35 can also be executed by other false twisting machine abnormality warning systems.
[0071] During operation OP31, the reference yarn RS1 and yarn AS1 are untwisted by the first friction disc FD1 in the false twisting machine 120. In some embodiments, the first friction disc FD1 first untwistries the reference yarn RS1, and then untwistries the yarn AS1 after training the machine learning model MD1 based on the reference yarn RS1.
[0072] When operating OP32, the false twisting machine 120 measures multiple reference rotation speeds of the first friction disc FD1 and multiple reference movement speeds of the reference yarn RS1 while the first friction disc FD1 untwises the reference yarn RS1.
[0073] For example, when the false twisting machine 120 untwists the first reference yarn in the reference yarn RS1 on the first friction disc FD1, it measures a first reference rotation speed and a first reference movement speed of the first reference yarn. Then, when the false twisting machine 120 untwists the second reference yarn in the reference yarn RS1 on the first friction disc FD1, it measures a second reference rotation speed and a second reference movement speed of the first reference yarn, and so on.
[0074] During operation OP33, processor 110 generates reference untwisting coefficients based on a reference rotation speed and multiple reference movement speeds of the reference yarn. For example, processor 110 divides a first reference rotation speed by a first reference movement speed to generate a first reference untwisting coefficient corresponding to a first reference yarn. Processor 110 divides a second reference rotation speed by a second reference movement speed to generate a second reference untwisting coefficient corresponding to a second reference yarn, and so on.
[0075] During operation OP34, when the false twisting machine 120 untwists the yarn AS1 on the first friction disc FD1, it measures multiple rotational speeds of the first friction disc FD1 and multiple movement speeds of the yarn AS1.
[0076] For example, when the false twisting machine 120 untwists the first yarn in the yarn AS1 using the first friction disc FD1, it measures a first rotational speed of the first friction disc FD1 and a first moving speed of the first yarn. Then, when the false twisting machine 120 untwists the second yarn in the yarn AS1 using the first friction disc FD1, it measures a second rotational speed of the first friction disc FD1 and a second moving speed of the first yarn, and so on.
[0077] During operation OP35, processor 110 generates untwisting coefficients based on the rotation speed and multiple movement speeds of the yarn AS1. For example, processor 110 divides a first rotation speed by a first movement speed to generate a first untwisting coefficient corresponding to the first yarn. Processor 110 divides a second rotation speed by a second movement speed to generate a second untwisting coefficient corresponding to the second yarn, and so on.
[0078] Figure 4 The flowchart illustrates other operations of the false twisting machine abnormality early warning method 200 according to some embodiments of this disclosure. For example... Figure 4 As shown, the false twisting machine abnormality early warning method 200 may further include operating OP41 to OP44. In some embodiments, operating OP41 to OP44 can be achieved through methods such as... Figure 1 The false twisting machine abnormality warning system 100 shown is executed. However, the embodiments disclosed herein are not limited to this, and operations OP41 to OP44 can also be executed by other false twisting machine abnormality warning systems.
[0079] During operation OP41, the false twisting machine 120 sequentially and continuously processes a first spindle, a second spindle, and a third spindle in AYS1. In other words, after untwisting the first spindle, the false twisting machine 120 continuously untwistries the second spindle, and after untwisting the second spindle, it continuously untwistes the third spindle.
[0080] During operation OP42, processor 110 generates a first predicted yield, a second predicted yield, and a third predicted yield, respectively, among the predicted yields of the first, second, and third wire spindles.
[0081] Specifically, the processor 110 inputs the first set of input parameters corresponding to the first spindle into the machine learning model MD1 to generate a first predicted yield, inputs the second set of input parameters corresponding to the second spindle into the machine learning model MD1 to generate a second predicted yield, and inputs the third set of input parameters corresponding to the third spindle into the machine learning model MD1 to generate a third predicted yield.
[0082] In operation OP43, the first predicted yield, the second predicted yield, and the third predicted yield are compared with a preset yield. In some embodiments, the preset yield is equal to 90%. However, the embodiments disclosed herein are not limited to this. In various embodiments, the preset yield can be various percentages.
[0083] During operation OP44, when the first predicted yield rate is less than the preset yield rate, the second predicted yield rate is less than the preset yield rate, and the third predicted yield rate is less than the preset yield rate, the false twisting machine malfunction warning system 100 repairs the false twisting machine 120. In some embodiments, the false twisting machine malfunction warning system 100 can issue a warning signal to trigger the repair operation on the false twisting machine 120. In some embodiments, three consecutive predicted yield rates (i.e., the first predicted yield rate, the second predicted yield rate, and the third predicted yield rate) being less than the predicted yield rate indicates that the false twisting machine 120 may have a malfunction.
[0084] Figure 5 The flowchart illustrates other operations of the false twisting machine abnormality early warning method 200 according to some embodiments of this disclosure. For example... Figure 5 As shown, the false twisting machine abnormality early warning method 200 may further include operating OP51 to OP54. In some embodiments, operating OP51 to OP54 can be achieved through methods such as... Figure 1 The false twisting machine abnormality warning system 100 shown is executed. However, the embodiments disclosed herein are not limited to this, and operations OP51 to OP54 can also be executed by other false twisting machine abnormality warning systems.
[0085] In operation OP51, when the first predicted yield is greater than the preset yield, the second predicted yield is greater than the preset yield, or the third predicted yield is greater than the preset yield, after the third spindle is processed by the false twisting machine 120, a fourth spindle in spindle AYS1 is processed by the false twisting machine 120.
[0086] In other words, when at least one of the first predicted yield, the second predicted yield, and the third predicted yield is greater than the preset yield, the processor 110 determines that the false twisting machine 120 is normal and instructs the false twisting machine 120 to continue processing the fourth spindle.
[0087] During operation OP52, processor 110 generates a fourth predicted yield corresponding to the fourth spindle in the predicted yield. Specifically, processor 110 inputs the fourth set of input parameters corresponding to the fourth spindle into machine learning model MD1 to generate the fourth predicted yield.
[0088] During operation OP53, processor 110 compares the fourth predicted yield with the preset yield.
[0089] During operation OP54, when the second predicted yield, the third predicted yield, and the fourth predicted yield are all less than the preset yield, the first friction disc FD1 is replaced via the second friction disc FD2. In some embodiments, three consecutive predicted yields (i.e., the second, third, and fourth predicted yields) being less than the predicted yield indicates that the false twisting machine 120 may be malfunctioning.
[0090] In summary, through the embodiments disclosed herein, the processor 110 can monitor the input parameter set of the spindle AYS1 to generate a predicted yield, and decide whether to repair the false twisting machine 120 based on the predicted yield. In this way, the false twisting machine anomaly early warning system 100 can promptly repair the false twisting machine 120.
[0091] Although the present disclosure has been described above with reference to embodiments, it is not intended to limit the present disclosure. Anyone skilled in the art may make some modifications and refinements without departing from the spirit and scope of the present disclosure. Therefore, the scope of protection of the present disclosure shall be determined by the appended claims.
Claims
1. A method for early warning of abnormalities in a false twisting machine, characterized in that, Include: Multiple reference spindles are converted into multiple reference yarn cakes using a false twisting machine; The multiple reference yarns of the multiple reference spindles are measured in the false twisting machine using multiple sets of reference input parameters; Measure multiple reference actual yields of the multiple reference yarn cakes; A machine learning model is trained based on the multiple sets of reference input parameters and the multiple reference actual yields. This false twisting machine converts multiple spools of silk into multiple spools of silk. Measure multiple sets of input parameters for multiple filaments of the multiple spindles in the false twisting machine; This machine learning model generates multiple predicted yields based on the multiple sets of input parameters. as well as Based on the aforementioned multiple predicted yield rates, a decision is made as to whether to repair the false twisting machine.
2. The method for early warning of abnormalities in a false twisting machine as described in claim 1, characterized in that, The plurality of reference input parameter sets include a plurality of reference yarn breakage numbers, a plurality of reference maximum variation rates, a plurality of reference tension alarm counts, and a plurality of reference untwisting coefficients for the plurality of reference yarns. The multiple input parameter groups include multiple yarn breakage numbers of the multiple yarns, multiple maximum variation rates of the multiple yarns, multiple tension alarm counts of the multiple yarns, and multiple untwisting coefficients of the multiple yarns.
3. The method for early warning of abnormalities in a false twisting machine as described in claim 2, characterized in that, Also includes: The plurality of reference yarns and the plurality of yarns are untwisted by a first friction disc in the false twisting machine; When the first friction disc untwistresses the plurality of reference threads, the plurality of reference rotation speeds of the first friction disc and the plurality of reference movement speeds of the plurality of reference threads are measured. The plurality of reference untwisting coefficients are generated based on the plurality of reference rotation speeds and the plurality of reference moving speeds of the plurality of reference filaments; When the first friction disc untwistwists the plurality of threads, multiple rotational speeds of the first friction disc and multiple moving speeds of the plurality of threads are measured; as well as The plurality of untwisting coefficients are generated based on the plurality of rotation speeds and the plurality of moving speeds of the plurality of filaments.
4. The false twisting machine abnormality early warning method as described in claim 3, characterized in that, Wherein, as the plurality of reference moving speeds increase, the plurality of reference untwisting coefficients decrease. As the plurality of reference rotational speeds increase, the plurality of reference untwisting coefficients increase. As the plurality of moving speeds increase, the plurality of untwisting coefficients decrease, and As the plurality of rotational speeds increase, the plurality of untwisting coefficients increase.
5. The method for early warning of abnormalities in a false twisting machine as described in claim 3, characterized in that, One of the plurality of reference untwisting coefficients is equal to a corresponding one of the plurality of reference rotational speeds divided by a corresponding one of the plurality of reference moving speeds, and One of the plurality of untwisting coefficients is equal to the corresponding one of the plurality of rotational speeds divided by the corresponding one of the plurality of movement speeds.
6. The method for early warning of abnormalities in a false twisting machine as described in claim 3, characterized in that, The directions of the plurality of rotational speeds are perpendicular to the directions of the plurality of moving speeds.
7. The method for early warning of abnormalities in a false twisting machine as described in claim 3, characterized in that, Also includes: The false twisting machine processes a first filament spindle, a second filament spindle, and a third filament spindle from the plurality of filament spindles sequentially and continuously. Generate a first predicted yield, a second predicted yield, and a third predicted yield from among the plurality of predicted yields corresponding to the first spindle, the second spindle, and the third spindle, respectively. The first predicted yield, the second predicted yield, and the third predicted yield are compared with a preset yield; and When the first predicted yield is less than the preset yield, the second predicted yield is less than the preset yield, and the third predicted yield is less than the preset yield, the false twisting machine is repaired.
8. The method for early warning of false twisting machine abnormalities as described in claim 7, characterized in that, The repair of this false twisting machine includes: The first friction disc is replaced by a second friction disc, which is different from the first friction disc.
9. The method for early warning of abnormalities in a false twisting machine as described in claim 8, characterized in that, Also includes: When the first predicted yield is greater than the preset yield, the second predicted yield is greater than the preset yield, or the third predicted yield is greater than the preset yield, after the false twisting machine processes the third spindle, the false twisting machine processes a fourth spindle among the plurality of spindles.
10. The method for early warning of abnormalities in a false twisting machine as described in claim 9, characterized in that, Also includes: Generate a fourth predicted yield among the plurality of predicted yields corresponding to the fourth spindle; The fourth predicted yield is compared with the preset yield; and When the second predicted yield is less than the preset yield, the third predicted yield is less than the preset yield, and the fourth predicted yield is less than the preset yield, the first friction disc is replaced by the second friction disc.