Production management method, system and storage medium for high-elastic polyester fabric

Abstract

The present invention relates to the technical field of production management of high-elastic polyester fabric, specifically a production management method, system, and storage medium for high-elastic polyester fabric. A heating device is provided at the discharge port of a fiber combing device, a fiber mixing image is obtained by a camera device, and a fiber infrared image is obtained by an infrared imaging device. Temperature data is obtained based on a standard heating temperature and the fiber temperature. A fiber region division model is constructed to identify the fiber mixing image, the fiber infrared image, and the temperature data. First fiber distribution data, second fiber distribution data, and impurity distribution data are obtained. An impurity warning coefficient is obtained by identifying the impurity distribution data. A fiber uniformity recognition model is constructed to identify the first fiber distribution data and the second fiber distribution data to obtain a mixed fiber warning coefficient. By using the impurity warning coefficient and the mixed fiber warning coefficient, the present invention achieves intelligent management of the fiber mixed combing production process.

Description

Technical Field

[0001] The present invention relates to the technical field of production management of high-elastic polyester fabric, and in particular to a production management method, system and storage medium for high-elastic polyester fabric. Background Art

[0002] High-elastic polyester fabric is a fabric with high elasticity that is made of polyester fiber as the main material. Compared with ordinary fabrics, high-elastic polyester fabric is more outstanding in softness, elasticity and comfort, and is suitable for a variety of application scenarios that require high elasticity.

[0003] The production process of high-elastic polyester fabric usually includes multiple steps such as spinning, weaving, dyeing and shaping; among them, in the spinning stage, polyester fiber and elastic fiber (such as spandex, lycra, etc.) are mixed in a certain proportion through a fiber mixer to obtain a mixed fiber; then a fiber carding machine is used to comb the mixed fiber, and then the yarn is made.

[0004] During the fiber mixing and combing stages, the uniformity and quality of the blended fibers significantly impact the final high-elastic polyester fabric. Uneven fiber mixing, irregular combing, and impurities can easily lead to partially elastic yarns or uneven thickness, affecting the yarn's appearance and strength, and even causing equipment failure.

[0005] Therefore, it is necessary to carry out production management of the fiber mixing and combing process of the fiber blending machine and the fiber combing machine, and take corresponding treatment according to the test results to ensure the quality of the yarn and the safety of the equipment.

[0006] To this end, a production management method, system and storage medium for high-elastic polyester fabric are proposed. Summary of the Invention

[0007] The present invention aims to provide a production management method, system, and storage medium for high-elastic polyester fabric. The present invention installs a heating device at the discharge port of a fiber combing device, obtains a fiber mixing image using a camera device, and obtains a fiber infrared image using an infrared imaging device. Temperature data is obtained based on a standard heating temperature and the fiber temperature. A fiber region division model is constructed to identify the fiber mixing image, the fiber infrared image, and the temperature data. First fiber distribution data, second fiber distribution data, and impurity distribution data are obtained. An impurity warning coefficient is obtained by identifying the impurity distribution data. A fiber uniformity recognition model is constructed to identify the first and second fiber distribution data to obtain a mixed fiber warning coefficient. Through the impurity warning coefficient and the mixed fiber warning coefficient, the present invention achieves intelligent management of the fiber mixed combing production process.

[0008] To achieve the above objectives, the present invention provides the following technical solutions:

[0009] A production management method for high-elastic polyester fabric, comprising:

[0010] A heating device is provided at the discharge port of the fiber combing equipment to heat the output mixed fiber according to the standard temperature; a camera device is provided to photograph the mixed fiber at the discharge port to obtain a mixed fiber image; an infrared imaging device is provided to identify the mixed fiber at the discharge port to obtain a fiber infrared image and fiber temperature;

[0011] Constructing a fiber region division model to identify fiber mixed images, fiber infrared images, and temperature data; dividing the mixed fibers according to fiber characteristic differences and temperature distribution differences to obtain first fiber distribution data, second fiber distribution data, and impurity distribution data; the temperature data includes standard temperature and fiber temperature;

[0012] The distribution position, distribution area and distribution shape of impurities are identified based on the impurity distribution data, and an impurity warning coefficient is calculated; when the impurity warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixing image is transported to the fiber combing equipment for reprocessing;

[0013] A fiber uniformity recognition model is constructed to identify the first fiber distribution data and the second fiber distribution data to obtain a fiber ratio abnormality coefficient, a fiber neatness abnormality coefficient, and a fiber uniformity abnormality coefficient; and a mixed fiber warning coefficient is obtained based on the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient, and the fiber uniformity abnormality coefficient.

[0014] The mixed fiber warning coefficient is identified, and when the mixed fiber warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixing image is transported to the fiber mixing equipment for further processing.

[0015] The fiber area division model includes a first image recognition layer, a second image recognition layer and a fiber type recognition layer;

[0016] The first image recognition layer recognizes the fiber mixed image and temperature data to obtain fiber feature data; the fiber feature data includes color features and texture features of the fiber;

[0017] The second image recognition layer recognizes the fiber infrared image and temperature data by a clustering algorithm to obtain the temperature distribution data of the mixed fiber;

[0018] The fiber type recognition layer recognizes the fiber mixed image and the fiber infrared image according to the fiber characteristic data, the temperature distribution data and the temperature data to obtain the first fiber distribution data, the second fiber distribution data and the impurity distribution data.

[0019] The first fiber is polyester fiber; the second fiber is elastic fiber; the first fiber distribution data includes distribution data of the first fiber in the fiber mixed image and distribution data of the first fiber in the fiber infrared image; the second fiber distribution data includes distribution data of the second fiber in the fiber mixed image and distribution data of the second fiber in the fiber infrared image; the impurity distribution data includes distribution data of impurities in the fiber mixed image and the fiber infrared image.

[0020] Identify the impurity distribution data to obtain the distribution position, distribution area and distribution shape of the impurities in the fiber mixing image;

[0021] According to the distribution position of impurities, the minimum circumscribed circle that can cover all impurities is obtained; according to the area ratio of the minimum circumscribed circle to the fiber mixed image, the distribution anomaly coefficient of impurities is obtained;

[0022] According to the distribution area of ​​the impurities, the total area of ​​the impurities in the fiber mixed image is obtained; according to the ratio of the total area of ​​the impurities to the area of ​​the fiber mixed image, the area anomaly coefficient of the impurities is obtained;

[0023] According to the distribution shape of the impurities, the circularity of the impurities is obtained; the circularity average value of all impurities is obtained, and the shape anomaly coefficient of the impurities is obtained by subtracting the circularity average value from 1;

[0024] The impurity warning coefficient is obtained based on the impurity distribution abnormality coefficient, area abnormality coefficient and shape abnormality coefficient.

[0025] The calculation formula of the impurity warning coefficient is:

[0026] WARF=β1*Dist+β2*Area+β3*Shape;

[0027] Among them, WARF represents the impurity warning coefficient; β1 represents the first impurity weight; Dist represents the impurity distribution anomaly coefficient; β2 represents the second impurity weight; Area represents the impurity area anomaly coefficient; β3 represents the third impurity weight; Shape represents the impurity shape anomaly coefficient.

[0028] The fiber uniformity recognition model includes a mixing ratio recognition layer, a fiber neatness recognition layer, a fiber mixing recognition layer and a fiber abnormality recognition layer;

[0029] The mixing ratio identification layer identifies the first fiber distribution area and the second fiber distribution area based on the first fiber distribution data and the second fiber distribution data; and calculates the fiber ratio anomaly coefficient based on the first fiber distribution area, the second fiber distribution area, and the standard mixing ratio; the standard mixing ratio is the fiber ratio in the fiber mixing stage;

[0030] The fiber regularity recognition layer recognizes the fiber texture features in the first fiber distribution data and the second fiber distribution data to obtain the fiber regularity abnormality area; and obtains the fiber regularity abnormality coefficient according to the area ratio of the regularity abnormality area;

[0031] The fiber mixing identification layer identifies the fiber characteristics and temperature distribution in the first fiber distribution data and the second fiber distribution data to obtain first fiber thickness distribution data and second fiber position distribution data; obtains a first thickness anomaly coefficient based on the data difference between the first fiber thickness distribution data and the standard thickness; obtains distance data between second fiber distribution positions based on the second fiber position distribution data, and obtains a second distribution anomaly coefficient based on the variance of the distance data; and calculates a fiber mixing anomaly coefficient based on the first thickness anomaly coefficient and the second distribution anomaly coefficient;

[0032] The fiber abnormality identification layer obtains a mixed fiber warning coefficient according to the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient and the fiber mixing abnormality coefficient.

[0033] The calculation formula of the mixed fiber warning coefficient is:

[0034] UNIF=λ1*Hunh+λ2*Neat+λ3*Juny;

[0035] Juny=exp(μ1*Thic+μ2*Tanc);

[0036] Among them, UNIF represents the mixed fiber warning coefficient; λ1 represents the first fiber weight; Hunh represents the fiber ratio abnormality coefficient; λ2 represents the second fiber weight; Neat represents the fiber neatness abnormality coefficient; λ3 represents the third fiber weight; Juny represents the fiber mixing abnormality coefficient; Thic represents the first thickness abnormality coefficient; μ1 represents the thickness weight; Tanc represents the second distribution abnormality coefficient; μ2 represents the distribution weight; exp represents the exponential function with the natural constant e as the base.

[0037] A production management system for high-elastic polyester fabric, comprising:

[0038] The fiber data acquisition module sets a heating device at the discharge port of the fiber combing equipment to heat the output mixed fiber according to the standard temperature; sets a camera device to shoot the mixed fiber at the discharge port to obtain a fiber mixed image; sets an infrared imaging device to identify the mixed fiber at the discharge port to obtain a fiber infrared image and fiber temperature;

[0039] The fiber data recognition module constructs a fiber area division model to identify the fiber mixed image, fiber infrared image and temperature data; the mixed fibers are divided according to the fiber characteristic differences and temperature distribution differences to obtain the first fiber distribution data, the second fiber distribution data and the impurity distribution data; the temperature data includes the standard temperature and the fiber temperature;

[0040] The impurity abnormality processing module identifies the distribution position, distribution area and distribution shape of impurities based on the impurity distribution data, and calculates the impurity warning coefficient; when the impurity warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixing image is transported to the fiber combing equipment for reprocessing;

[0041] A fiber abnormality recognition module constructs a fiber uniformity recognition model, identifies the first fiber distribution data and the second fiber distribution data, and obtains a fiber ratio abnormality coefficient, a fiber neatness abnormality coefficient, and a fiber uniformity abnormality coefficient; and then obtains a mixed fiber warning coefficient based on the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient, and the fiber uniformity abnormality coefficient;

[0042] The fiber abnormality processing module identifies the mixed fiber warning coefficient. When the mixed fiber warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixing image is transported to the fiber mixing equipment for further processing.

[0043] In the impurity abnormality processing module, the calculation process of the impurity warning coefficient is as follows:

[0044] Identify the impurity distribution data to obtain the distribution position, distribution area and distribution shape of the impurities in the fiber mixing image;

[0045] According to the distribution position of impurities, the minimum circumscribed circle that can cover all impurities is obtained; according to the area ratio of the minimum circumscribed circle to the fiber mixed image, the distribution anomaly coefficient of impurities is obtained;

[0046] According to the distribution area of ​​the impurities, the total area of ​​the impurities in the fiber mixed image is obtained; according to the ratio of the total area of ​​the impurities to the area of ​​the fiber mixed image, the area anomaly coefficient of the impurities is obtained;

[0047] According to the distribution shape of the impurities, the circularity of the impurities is obtained; the circularity average value of all impurities is obtained, and the shape anomaly coefficient of the impurities is obtained by subtracting the circularity average value from 1;

[0048] The impurity warning coefficient is obtained based on the impurity distribution abnormality coefficient, area abnormality coefficient and shape abnormality coefficient.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] 1. The present invention photographs mixed fibers with a camera device to obtain a mixed fiber image; identifies the mixed fibers with an infrared imaging device to obtain a fiber infrared image; simultaneously obtains the fiber temperature and the standard heating temperature as temperature data; identifies the mixed fiber image and the temperature data to obtain fiber characteristic data; identifies the fiber infrared image and the temperature data with a clustering algorithm to obtain temperature distribution data; and then accurately identifies the distribution data of different types of fibers in the mixed fiber using the fiber characteristic data, the temperature distribution data, and the temperature data.

[0051] 2. The present invention obtains the distribution position, distribution area and distribution shape of impurities in the fiber mixture image based on the impurity distribution data; obtains the impurity distribution anomaly coefficient based on the impurity distribution position; obtains the impurity area anomaly coefficient based on the impurity distribution area; obtains the impurity shape anomaly coefficient based on the impurity distribution shape; and obtains the impurity warning coefficient based on the impurity distribution anomaly coefficient, area anomaly coefficient and shape anomaly coefficient; the impurity warning coefficient can be used to accurately identify and warn of impurity distribution anomalies in mixed fibers.

[0052] 3. The present invention calculates the fiber proportion anomaly coefficient based on the first fiber distribution area, the second fiber distribution area and the standard mixing ratio; identifies the fiber neatness anomaly coefficient based on the fiber texture characteristics; identifies the first fiber thickness distribution data and the second fiber position distribution data based on the fiber characteristics and temperature distribution; obtains the first thickness anomaly coefficient based on the data difference between the first fiber thickness distribution data and the standard thickness; obtains the distance data between the second fibers based on the second fiber position distribution data, and obtains the second distribution anomaly coefficient based on the distance data; and obtains the mixed fiber warning coefficient based on the fiber proportion anomaly coefficient, the fiber neatness anomaly coefficient and the fiber mixing anomaly coefficient; the mixed fiber warning coefficient can be used to accurately identify the mixing anomalies of different types of fibers in the mixed fibers. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Schematic diagram of a process for the production and management of a high-elastic polyester fabric according to the present invention;

[0054] Figure 2 It is a structural schematic diagram of the fiber area division model of the present invention;

[0055] Figure 3 Schematic diagram of the structure of the fiber uniformity recognition model of the present invention;

[0056] Figure 4 This is a schematic structural diagram of a production management system for high-elastic polyester fabric according to the present invention;

[0057] Figure 5 This is a logic judgment diagram of a production management method for high-elastic polyester fabric of the present invention. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] High-elastic polyester fabric is a fabric with high elasticity, comfort and durability, and is widely used in products such as clothing, sportswear, underwear, yoga wear, etc. The elasticity of high-elastic polyester fabric comes from the type and proportion of elastic yarn used; elastic yarn is usually woven from a mixture of elastic fiber and polyester fiber.

[0060] The production process for high-stretch polyester fabric involves a series of steps, including spinning, weaving, dyeing, and finishing. During the spinning stage, elastic yarn is produced through fiber blending, combing, stretching, and merging. During the blending and combing stages, uneven fiber blending, irregular combing, and fiber impurities can easily lead to localized poor elasticity or uneven thickness in the yarn, affecting its appearance and strength, and even causing equipment failure.

[0061] Therefore, it is necessary to carry out production management of the fiber mixing and carding process of the fiber mixing machine and the fiber carding machine.

[0062] Example 1

[0063] The present invention proposes a production management method for high-elastic polyester fabric, the process of which is as follows: Figure 1 As shown, including:

[0064] S10. A heating device is provided at the discharge port of the fiber combing device to heat the output mixed fiber according to the standard temperature; a camera is provided to capture the mixed fiber at the discharge port to obtain a mixed fiber image; an infrared imaging device is provided to identify the mixed fiber at the discharge port to obtain an infrared image of the fiber and the fiber temperature;

[0065] S20. Constructing a fiber region division model to identify the fiber mixed image, fiber infrared image, and temperature data; dividing the mixed fibers according to the fiber characteristic differences and temperature distribution differences to obtain first fiber distribution data, second fiber distribution data, and impurity distribution data; the temperature data includes a standard temperature and a fiber temperature;

[0066] The fiber area division model is constructed based on a deep neural network, and its structure is as follows: Figure 2 As shown, it includes a first image recognition layer, a second image recognition layer and a fiber type recognition layer;

[0067] The first image recognition layer recognizes the fiber mixed image and temperature data to obtain fiber feature data; the fiber feature data includes color features and texture features of the fiber;

[0068] The second image recognition layer recognizes the fiber infrared image and temperature data by a clustering algorithm to obtain the temperature distribution data of the mixed fiber;

[0069] The fiber type recognition layer recognizes the fiber mixed image and the fiber infrared image according to the fiber characteristic data, the temperature distribution data and the temperature data to obtain the first fiber distribution data, the second fiber distribution data and the impurity distribution data.

[0070] The first fiber is polyester fiber; the second fiber is elastic fiber; the first fiber distribution data includes distribution data of the first fiber in the fiber mixed image and distribution data of the first fiber in the fiber infrared image; the second fiber distribution data includes distribution data of the second fiber in the fiber mixed image and distribution data of the second fiber in the fiber infrared image; the impurity distribution data includes distribution data of impurities in the fiber mixed image and the fiber infrared image.

[0071] In order to verify the recognition effect of the fiber area division model described in the present invention, it is tested.

[0072] The test includes a fiber area division model, a first comparison model and a second comparison model; the first comparison model is used to identify the fiber mixing image and temperature data to obtain fiber feature data; the fiber mixing image and the fiber infrared image are identified based on the fiber feature data to obtain corresponding distribution data.

[0073] The second comparison model identifies the fiber infrared image and temperature data according to the clustering algorithm to obtain temperature distribution data of the mixed fiber; and identifies the fiber mixed image and the fiber infrared image according to the temperature distribution data to obtain corresponding distribution data.

[0074] The fiber region partition model, the first comparison model, and the second comparison model were used to identify 100 test samples, obtaining the corresponding test distribution data. The F1 index of the fiber region partition model, the first comparison model, and the second comparison model was calculated based on the test distribution data and the actual distribution data. The average F1 index was used as a measure of the model's recognition effectiveness. The resulting test data is shown in Table 1.

[0075] Table 1 Fiber area division model test data table

[0076] Model Testing rounds The average value of F1 index Fiber region partition model 100 0.9425 First comparison model 100 0.9031 The second comparison model 100 0.8863

[0077] From the data in Table 1, it can be seen that the fiber area division model has the best recognition effect.

[0078] The present invention photographs mixed fibers with a camera device to obtain a fiber mixing image; identifies the mixed fibers with an infrared imaging device to obtain a fiber infrared image; simultaneously obtains the fiber temperature and the standard heating temperature as temperature data; identifies the fiber mixing image and the temperature data to obtain fiber characteristic data; identifies the fiber infrared image and the temperature data with a clustering algorithm to obtain temperature distribution data; and then accurately identifies the distribution data of different types of fibers in the mixed fibers using the fiber characteristic data, the temperature distribution data, and the temperature data.

[0079] S30. The distribution position, distribution area and distribution shape of impurities are identified according to the impurity distribution data, and the impurity warning coefficient is calculated; when the impurity warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixed image is transported to the fiber combing device for reprocessing;

[0080] Identify the impurity distribution data to obtain the distribution position, distribution area and distribution shape of the impurities in the fiber mixing image;

[0081] According to the distribution position of impurities, the minimum circumscribed circle that can cover all impurities is obtained; according to the area ratio of the minimum circumscribed circle to the fiber mixed image, the distribution anomaly coefficient of impurities is obtained;

[0082] According to the distribution area of ​​the impurities, the total area of ​​the impurities in the fiber mixed image is obtained; according to the ratio of the total area of ​​the impurities to the area of ​​the fiber mixed image, the area anomaly coefficient of the impurities is obtained;

[0083] According to the distribution shape of the impurities, the circularity of the impurities is obtained; the circularity average value of all impurities is obtained, and the shape anomaly coefficient of the impurities is obtained by subtracting the circularity average value from 1;

[0084] The impurity warning coefficient is obtained based on the impurity distribution abnormality coefficient, area abnormality coefficient and shape abnormality coefficient.

[0085] The calculation formula of the impurity warning coefficient is:

[0086] WARF=β1*Dist+β2*Area+β3*Shape;

[0087] Where WARF represents the impurity warning coefficient; β1 represents the first impurity weight; Dist represents the impurity distribution anomaly coefficient; β2 represents the second impurity weight; Area represents the impurity area anomaly coefficient; β3 represents the third impurity weight; and Shape represents the impurity shape anomaly coefficient. The weights in the impurity warning coefficient calculation formula are obtained by collecting corresponding verification data.

[0088] The present invention obtains the distribution position, distribution area and distribution shape of impurities in a fiber mixing image based on impurity distribution data identification; obtains the impurity distribution anomaly coefficient based on the impurity distribution position; obtains the impurity area anomaly coefficient based on the impurity distribution area; obtains the impurity shape anomaly coefficient based on the impurity distribution shape; and obtains the impurity early warning coefficient based on the impurity distribution anomaly coefficient, area anomaly coefficient and shape anomaly coefficient; and the impurity early warning coefficient can accurately identify and warn of impurity distribution anomalies in mixed fibers.

[0089] S40. Constructing a fiber uniformity recognition model to identify the first fiber distribution data and the second fiber distribution data to obtain a fiber ratio abnormality coefficient, a fiber neatness abnormality coefficient, and a fiber uniformity abnormality coefficient; and then obtaining a mixed fiber warning coefficient based on the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient, and the fiber uniformity abnormality coefficient;

[0090] The fiber uniformity recognition model is constructed based on a deep neural network, and its structure is as follows: Figure 3 As shown, it includes a mixing ratio identification layer, a fiber neatness identification layer, a fiber mixing identification layer and a fiber abnormality identification layer;

[0091] The mixing ratio identification layer identifies the first fiber distribution area and the second fiber distribution area based on the first fiber distribution data and the second fiber distribution data; and calculates the fiber ratio anomaly coefficient based on the first fiber distribution area, the second fiber distribution area, and the standard mixing ratio; the standard mixing ratio is the ratio of the first fiber to the second fiber in the fiber mixing stage;

[0092] The calculation formula of the fiber ratio abnormality coefficient is:

[0093]

[0094] Wherein, Hunh represents the fiber ratio anomaly coefficient; A1 represents the first fiber distribution area; A2 represents the second fiber distribution area; Indicates standard mixing ratio.

[0095] The fiber regularity recognition layer recognizes the fiber texture features in the first fiber distribution data and the second fiber distribution data to obtain the fiber regularity abnormality area; and obtains the fiber regularity abnormality coefficient according to the area ratio of the regularity abnormality area;

[0096] The fiber mixing identification layer identifies the fiber characteristics and temperature distribution in the first fiber distribution data and the second fiber distribution data to obtain first fiber thickness distribution data and second fiber position distribution data; obtains a first thickness anomaly coefficient based on the data difference between the first fiber thickness distribution data and the standard thickness; obtains distance data between second fiber distribution positions based on the second fiber position distribution data, and obtains a second distribution anomaly coefficient based on the variance of the distance data; and calculates a fiber mixing anomaly coefficient based on the first thickness anomaly coefficient and the second distribution anomaly coefficient;

[0097] The fiber abnormality identification layer obtains a mixed fiber warning coefficient based on the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient and the fiber mixing abnormality coefficient. The calculation formula of the mixed fiber warning coefficient is:

[0098] UNIF=λ1*Hunh+λ2*Neat+λ3*Juny;

[0099] Juny=exp(μ1*Thic+μ2*Tanc);

[0100] Among them, UNIF represents the mixed fiber warning coefficient; λ1 represents the first fiber weight; Hunh represents the fiber ratio abnormality coefficient; λ2 represents the second fiber weight; Neat represents the fiber neatness abnormality coefficient; λ3 represents the third fiber weight; Juny represents the fiber mixing abnormality coefficient; Thic represents the first thickness abnormality coefficient; μ1 represents the thickness weight; Tanc represents the second distribution abnormality coefficient; μ2 represents the distribution weight; exp represents the exponential function with the natural constant e as the base.

[0101] Among them, the weight of the calculation formula of the mixed fiber warning coefficient is obtained by collecting corresponding verification data.

[0102] The present invention calculates and obtains a fiber proportion anomaly coefficient based on a first fiber distribution area, a second fiber distribution area and a standard mixing ratio; identifies and obtains a fiber neatness anomaly coefficient based on fiber texture characteristics; identifies and obtains first fiber thickness distribution data and second fiber position distribution data based on fiber characteristics and temperature distribution; obtains a first thickness anomaly coefficient based on a data difference between the first fiber thickness distribution data and the standard thickness; obtains distance data between second fibers based on the second fiber position distribution data, and obtains a second distribution anomaly coefficient based on the distance data; and obtains a mixed fiber warning coefficient based on the fiber proportion anomaly coefficient, the fiber neatness anomaly coefficient and the fiber mixing anomaly coefficient; and the mixed fiber warning coefficient can accurately identify different types of fiber mixing anomalies in the mixed fiber.

[0103] S50. Identify the mixed fiber warning coefficient. When the mixed fiber warning coefficient does not meet the requirement, transport the mixed fiber corresponding to the fiber mixing image to a fiber mixing device for further processing.

[0104] The present invention also proposes a production management system for high-elastic polyester fabric, the structure of which is as follows: Figure 4 As shown, including:

[0105] The fiber data acquisition module sets a heating device at the discharge port of the fiber combing equipment to heat the output mixed fiber according to the standard temperature; sets a camera device to shoot the mixed fiber at the discharge port to obtain a fiber mixed image; sets an infrared imaging device to identify the mixed fiber at the discharge port to obtain a fiber infrared image and fiber temperature;

[0106] The fiber data recognition module constructs a fiber area division model to identify the fiber mixed image, fiber infrared image and temperature data; the mixed fibers are divided according to the fiber characteristic differences and temperature distribution differences to obtain the first fiber distribution data, the second fiber distribution data and the impurity distribution data; the temperature data includes the standard temperature and the fiber temperature;

[0107] The impurity abnormality processing module identifies the distribution position, distribution area and distribution shape of impurities based on the impurity distribution data, and calculates the impurity warning coefficient; when the impurity warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixing image is transported to the fiber combing equipment for reprocessing;

[0108] A fiber abnormality recognition module constructs a fiber uniformity recognition model, identifies the first fiber distribution data and the second fiber distribution data, and obtains a fiber ratio abnormality coefficient, a fiber neatness abnormality coefficient, and a fiber uniformity abnormality coefficient; and then obtains a mixed fiber warning coefficient based on the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient, and the fiber uniformity abnormality coefficient;

[0109] The fiber abnormality processing module identifies the mixed fiber warning coefficient. When the mixed fiber warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixing image is transported to the fiber mixing equipment for further processing.

[0110] The present invention also provides a storage medium having a computer program stored thereon. When the computer program is executed by a processor, the above-mentioned production management method for high-elastic polyester fabric is implemented.

[0111] The present invention installs a heating device at the discharge port of the fiber combing equipment, obtains a fiber mixing image using a camera device, and obtains a fiber infrared image using an infrared imaging device; obtains temperature data based on the standard heating temperature and the fiber temperature; constructs a fiber region division model to identify the fiber mixing image, fiber infrared image, and temperature data; obtains first fiber distribution data, second fiber distribution data, and impurity distribution data; obtains an impurity warning coefficient based on the impurity distribution data; and constructs a fiber uniformity recognition model to identify the first and second fiber distribution data to obtain a mixed fiber warning coefficient. Through the impurity warning coefficient and the mixed fiber warning coefficient, the present invention achieves intelligent management of the fiber mixing and combing production process.

[0112] Example 2

[0113] In order to verify the actual effect of the production management method of high-elastic polyester fabric proposed by the present invention, it is verified according to actual scenarios; the logical judgment of the production management method of high-elastic polyester fabric is as follows: Figure 5 As shown, including:

[0114] S10. A heating device is provided at the discharge port of the fiber combing device to heat the output mixed fiber according to the standard temperature; a camera is provided to capture the mixed fiber at the discharge port to obtain a mixed fiber image; an infrared imaging device is provided to identify the mixed fiber at the discharge port to obtain an infrared image of the fiber and the fiber temperature;

[0115] Polyester fiber itself is generally transparent or slightly milky white, with a smooth surface, flat and uniform appearance; it has good heat resistance and can remain stable at higher temperatures, but it is still affected by thermoplasticity to a certain extent.

[0116] Commonly used elastic fibers include spandex fibers and lycra fibers. The present invention uses spandex fibers as the elastic fibers.

[0117] Spandex fibers are typically white or transparent, lighter in color, and typically have a round cross-section. However, they are thinner and longer than polyester fibers, and their surface is generally smoother. Spandex fibers are sensitive to high temperatures and tend to lose their elasticity and even deform after prolonged exposure to high temperatures. Typically, the elastic yarn in high-elastic polyester fabrics contains 5% to 20% elastic fiber; in this invention, the spandex fiber content is set at 10%.

[0118] S20. Constructing a fiber region division model to identify the fiber mixed image, fiber infrared image, and temperature data; dividing the mixed fibers according to the fiber characteristic differences and temperature distribution differences to obtain first fiber distribution data, second fiber distribution data, and impurity distribution data; the temperature data includes a standard temperature and a fiber temperature;

[0119] The fiber area division model includes a first image recognition layer, a second image recognition layer and a fiber type recognition layer;

[0120] The first image recognition layer recognizes the fiber mixed image and temperature data to obtain fiber feature data; the fiber feature data includes color features and texture features of the fiber;

[0121] The second image recognition layer recognizes the fiber infrared image and temperature data by a clustering algorithm to obtain the temperature distribution data of the mixed fiber;

[0122] The fiber type recognition layer recognizes the fiber mixed image and the fiber infrared image according to the fiber characteristic data, the temperature distribution data and the temperature data to obtain the first fiber distribution data, the second fiber distribution data and the impurity distribution data.

[0123] The first fiber is polyester fiber; the second fiber is elastic fiber; the first fiber distribution data includes distribution data of the first fiber in the fiber mixed image and distribution data of the first fiber in the fiber infrared image; the second fiber distribution data includes distribution data of the second fiber in the fiber mixed image and distribution data of the second fiber in the fiber infrared image; the impurity distribution data includes distribution data of impurities in the fiber mixed image and the fiber infrared image.

[0124] The present invention photographs mixed fibers with a camera device to obtain a fiber mixing image; identifies the mixed fibers with an infrared imaging device to obtain a fiber infrared image; simultaneously obtains the fiber temperature and the standard heating temperature as temperature data; identifies the fiber mixing image and the temperature data to obtain fiber characteristic data; identifies the fiber infrared image and the temperature data with a clustering algorithm to obtain temperature distribution data; and then accurately identifies the distribution data of different types of fibers in the mixed fibers using the fiber characteristic data, the temperature distribution data, and the temperature data.

[0125] S30. The distribution position, distribution area and distribution shape of impurities are identified according to the impurity distribution data, and the impurity warning coefficient is calculated; when the impurity warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixed image is transported to the fiber combing device for reprocessing;

[0126] Identify the impurity distribution data to obtain the distribution position, distribution area and distribution shape of the impurities in the fiber mixing image;

[0127] According to the distribution position of impurities, the minimum circumscribed circle that can cover all impurities is obtained; according to the area ratio of the minimum circumscribed circle to the fiber mixed image, the distribution anomaly coefficient of impurities is obtained;

[0128] According to the distribution area of ​​the impurities, the total area of ​​the impurities in the fiber mixed image is obtained; according to the ratio of the total area of ​​the impurities to the area of ​​the fiber mixed image, the area anomaly coefficient of the impurities is obtained;

[0129] According to the distribution shape of the impurities, the circularity of the impurities is obtained; the circularity average value of all impurities is obtained, and the shape anomaly coefficient of the impurities is obtained by subtracting the circularity average value from 1;

[0130] The impurity warning coefficient is obtained based on the impurity distribution abnormality coefficient, area abnormality coefficient and shape abnormality coefficient.

[0131] By identifying the mixed fiber samples, the impurity distribution data is obtained; then, the corresponding parameters are collected and calculated based on the impurity distribution data, as shown in Table 2.

[0132] Table 2 Impurity distribution data parameter table

[0133] Sample serial number area Distribution anomaly coefficient Area anomaly coefficient Shape anomaly coefficient 1 1㎡ 0.457 0.147 0.231 2 1㎡ 0.512 0.205 0.689 3 1㎡ 0.345 0.138 0.452 4 1㎡ 0.628 0.115 0.573 5 1㎡ 0.379 0.195 0.312 6 1㎡ 0.598 0.217 0.596 7 1㎡ 0.289 0.101 0.764 8 1㎡ 0.416 0.191 0.754 9 1㎡ 0.523 0.211 0.635 10 1㎡ 0.561 0.126 0.482

[0134] The formula for calculating the impurity warning coefficient using the distribution anomaly coefficient, area anomaly coefficient, and shape anomaly coefficient is:

[0135] WARF=β1*Dist+β2*Area+β3*Shape;

[0136] Among them, WARF represents the impurity warning coefficient; β1 represents the first impurity weight; Dist represents the impurity distribution anomaly coefficient; β2 represents the second impurity weight; Area represents the impurity area anomaly coefficient; β3 represents the third impurity weight; Shape represents the impurity shape anomaly coefficient.

[0137] S40. Constructing a fiber uniformity recognition model to identify the first fiber distribution data and the second fiber distribution data to obtain a fiber ratio abnormality coefficient, a fiber neatness abnormality coefficient, and a fiber uniformity abnormality coefficient; and then obtaining a mixed fiber warning coefficient based on the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient, and the fiber uniformity abnormality coefficient;

[0138] The fiber uniformity recognition model includes a mixing ratio recognition layer, a fiber neatness recognition layer, a fiber mixing recognition layer and a fiber abnormality recognition layer;

[0139] The mixing ratio identification layer identifies the first fiber distribution area and the second fiber distribution area based on the first fiber distribution data and the second fiber distribution data; and calculates the fiber ratio anomaly coefficient based on the first fiber distribution area, the second fiber distribution area, and the standard mixing ratio; the standard mixing ratio is the fiber ratio in the fiber mixing stage;

[0140] The fiber regularity recognition layer recognizes the fiber texture features in the first fiber distribution data and the second fiber distribution data to obtain the fiber regularity abnormality area; and obtains the fiber regularity abnormality coefficient according to the area ratio of the regularity abnormality area;

[0141] The fiber mixing identification layer identifies the fiber characteristics and temperature distribution in the first fiber distribution data and the second fiber distribution data to obtain first fiber thickness distribution data and second fiber position distribution data; obtains a first thickness anomaly coefficient based on the data difference between the first fiber thickness distribution data and the standard thickness; obtains distance data between second fiber distribution positions based on the second fiber position distribution data, and obtains a second distribution anomaly coefficient based on the variance of the distance data; and calculates a fiber mixing anomaly coefficient based on the first thickness anomaly coefficient and the second distribution anomaly coefficient;

[0142] The fiber abnormality identification layer obtains a mixed fiber warning coefficient according to the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient and the fiber mixing abnormality coefficient.

[0143] By identifying the mixed fiber sample, the first fiber distribution data and the second fiber distribution data are obtained; then the corresponding fiber abnormality parameters are obtained by identification and measurement, as shown in Table 3.

[0144] Table 3 Data table of fiber abnormality parameters

[0145] Sample serial number area Fiber ratio anomaly coefficient Fiber regularity anomaly coefficient Fiber mixing anomaly coefficient 1 1㎡ 0.278 0.345 0.217 2 1㎡ 0.168 0.312 0.289 3 1㎡ 0.137 0.205 0.411 4 1㎡ 0.069 0.220 0.134 5 1㎡ 0.218 0.270 0.386 6 1㎡ 0.132 0.227 0.200 7 1㎡ 0.413 0.309 0.176 8 1㎡ 0.320 0.420 0.148 9 1㎡ 0.053 0.329 0.314 10 1㎡ 0.193 0.185 0.402

[0146] The calculation formula of the mixed fiber warning coefficient is:

[0147] UNIF=λ1*Hunh+λ2*Neat+λ3*Juny;

[0148] Juny=exp(μ1*Thic+μ2*Tanc);

[0149] Among them, UNIF represents the mixed fiber warning coefficient; λ1 represents the first fiber weight; Hunh represents the fiber ratio abnormality coefficient; λ2 represents the second fiber weight; Neat represents the fiber neatness abnormality coefficient; λ3 represents the third fiber weight; Juny represents the fiber mixing abnormality coefficient; Thic represents the first thickness abnormality coefficient; μ1 represents the thickness weight; Tanc represents the second distribution abnormality coefficient; μ2 represents the distribution weight; exp represents the exponential function with the natural constant e as the base.

[0150] The present invention installs a heating device at the discharge port of the fiber combing equipment, obtains a fiber mixing image using a camera device, and obtains a fiber infrared image using an infrared imaging device; obtains temperature data based on the standard heating temperature and the fiber temperature; constructs a fiber region division model to identify the fiber mixing image, fiber infrared image, and temperature data; obtains first fiber distribution data, second fiber distribution data, and impurity distribution data; obtains an impurity warning coefficient based on the impurity distribution data; and constructs a fiber uniformity recognition model to identify the first and second fiber distribution data to obtain a mixed fiber warning coefficient. Through the impurity warning coefficient and the mixed fiber warning coefficient, the present invention achieves intelligent management of the fiber mixing and combing production process.

[0151] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A production management method for high-elastic polyester fabric, characterized in that: include: S10. A heating device is provided at the discharge port of the fiber carding device to heat the output mixed fiber according to the standard temperature; A camera device is set to shoot the mixed fibers at the discharge port to obtain a fiber mixed image; an infrared imaging device is set to identify the mixed fibers at the discharge port to obtain a fiber infrared image and fiber temperature; S20. Constructing a fiber region division model to identify the fiber mixed image, fiber infrared image, and temperature data; dividing the mixed fibers according to the fiber characteristic differences and temperature distribution differences to obtain first fiber distribution data, second fiber distribution data, and impurity distribution data; the temperature data includes a standard temperature and a fiber temperature; The first fiber is a polyester fiber; the second fiber is an elastic fiber; the first fiber distribution data includes distribution data of the first fiber in a fiber mixture image and distribution data of the first fiber in a fiber infrared image; the second fiber distribution data includes distribution data of the second fiber in a fiber mixture image and distribution data of the second fiber in a fiber infrared image; the impurity distribution data includes distribution data of impurities in a fiber mixture image and a fiber infrared image; S30. Identify the distribution position, distribution area, and distribution shape of the impurities based on the impurity distribution data, and obtain the circularity of the impurities based on the distribution shape of the impurities; obtain the average circularity of all impurities, and subtract the average circularity from 1 to obtain the impurity shape anomaly coefficient; obtain the impurity warning coefficient based on the impurity distribution anomaly coefficient, the impurity area anomaly coefficient, and the impurity shape anomaly coefficient; when the impurity warning coefficient does not meet the requirements, transport the mixed fiber corresponding to the fiber mixing image to the fiber combing equipment for reprocessing; S40. Constructing a fiber uniformity recognition model to identify the first fiber distribution data and the second fiber distribution data to obtain a fiber ratio abnormality coefficient, a fiber neatness abnormality coefficient, and a fiber uniformity abnormality coefficient; and then obtaining a mixed fiber warning coefficient based on the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient, and the fiber uniformity abnormality coefficient; The calculation formula of the mixed fiber warning coefficient is: UNIF=λ1*Hunh+λ2*Neat+λ3*Juny; Juny=exp(μ1*Thic+μ2*Tanc); Among them, UNIF represents the mixed fiber warning coefficient; λ1 represents the first fiber weight; Hunh represents the fiber ratio abnormality coefficient; λ2 represents the second fiber weight; Neat represents the fiber neatness anomaly coefficient; λ3 represents the third fiber weight; Juny represents the fiber mixing anomaly coefficient; Thic represents the first thickness anomaly coefficient; μ1 represents the thickness weight; Tanc represents the second distribution anomaly coefficient; μ2 represents the distribution weight; exp represents the exponential function with the natural constant e as the base; S50. Identify the mixed fiber warning coefficient. When the mixed fiber warning coefficient does not meet the requirement, transport the mixed fiber corresponding to the fiber mixing image to a fiber mixing device for further processing.

2. The production management method of a high-elastic polyester fabric according to claim 1, characterized in that: The fiber area division model includes a first image recognition layer, a second image recognition layer and a fiber type recognition layer; The first image recognition layer recognizes the fiber mixed image and temperature data to obtain fiber feature data; the fiber feature data includes color features and texture features of the fiber; The second image recognition layer recognizes the fiber infrared image and temperature data by a clustering algorithm to obtain the temperature distribution data of the mixed fiber; The fiber type recognition layer recognizes the fiber mixed image and the fiber infrared image according to the fiber characteristic data, the temperature distribution data and the temperature data to obtain the first fiber distribution data, the second fiber distribution data and the impurity distribution data.

3. The production management method of high-elastic polyester fabric according to claim 1, characterized in that: Identify the impurity distribution data to obtain the distribution position, distribution area and distribution shape of the impurities in the fiber mixing image; According to the distribution position of impurities, the minimum circumscribed circle that can cover all impurities is obtained; according to the area ratio of the minimum circumscribed circle to the fiber mixed image, the distribution anomaly coefficient of impurities is obtained; According to the distribution area of ​​the impurities, the total area of ​​the impurities in the fiber mixing image is obtained; according to the ratio of the total area of ​​the impurities to the area of ​​the fiber mixing image, the area anomaly coefficient of the impurities is obtained.

4. The production management method of high-elastic polyester fabric according to claim 1, characterized in that: The calculation formula of the impurity warning coefficient is: WARF=β1*Dist+β2*Area+β3*Shape; Among them, WARF represents the impurity warning coefficient; β1 represents the first impurity weight; Dist represents the impurity distribution anomaly coefficient; β2 represents the second impurity weight; Area represents the impurity area anomaly coefficient; β3 represents the third impurity weight; Shape represents the impurity shape anomaly coefficient.

5. The production management method of high-elastic polyester fabric according to claim 1, characterized in that: The fiber uniformity recognition model includes a mixing ratio recognition layer, a fiber neatness recognition layer, a fiber mixing recognition layer and a fiber abnormality recognition layer; The mixing ratio identification layer identifies the first fiber distribution area and the second fiber distribution area based on the first fiber distribution data and the second fiber distribution data; and calculates the fiber ratio anomaly coefficient based on the first fiber distribution area, the second fiber distribution area, and the standard mixing ratio; the standard mixing ratio is the fiber ratio in the fiber mixing stage; The fiber regularity recognition layer recognizes the fiber texture features in the first fiber distribution data and the second fiber distribution data to obtain the fiber regularity abnormality area; and obtains the fiber regularity abnormality coefficient according to the area ratio of the regularity abnormality area; The fiber mixing identification layer identifies the fiber characteristics and temperature distribution in the first fiber distribution data and the second fiber distribution data to obtain first fiber thickness distribution data and second fiber position distribution data; obtains a first thickness anomaly coefficient based on the data difference between the first fiber thickness distribution data and the standard thickness; obtains distance data between second fiber distribution positions based on the second fiber position distribution data, and obtains a second distribution anomaly coefficient based on the variance of the distance data; and calculates a fiber mixing anomaly coefficient based on the first thickness anomaly coefficient and the second distribution anomaly coefficient; The fiber abnormality identification layer obtains a mixed fiber warning coefficient according to the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient and the fiber mixing abnormality coefficient.

6. A production management system for high-elastic polyester fabric, characterized in that: include: The fiber data acquisition module sets a heating device at the discharge port of the fiber combing equipment to heat the output mixed fiber according to the standard temperature; A camera device is set to shoot the mixed fibers at the discharge port to obtain a fiber mixed image; an infrared imaging device is set to identify the mixed fibers at the discharge port to obtain a fiber infrared image and fiber temperature; The fiber data recognition module constructs a fiber area division model to identify the fiber mixed image, fiber infrared image and temperature data; the mixed fibers are divided according to the fiber characteristic differences and temperature distribution differences to obtain the first fiber distribution data, the second fiber distribution data and the impurity distribution data; the temperature data includes the standard temperature and the fiber temperature; The first fiber is a polyester fiber; the second fiber is an elastic fiber; the first fiber distribution data includes distribution data of the first fiber in a fiber mixture image and distribution data of the first fiber in a fiber infrared image; the second fiber distribution data includes distribution data of the second fiber in a fiber mixture image and distribution data of the second fiber in a fiber infrared image; the impurity distribution data includes distribution data of impurities in a fiber mixture image and a fiber infrared image; The impurity anomaly processing module identifies the distribution position, distribution area, and distribution shape of impurities based on the impurity distribution data, and obtains the circularity of the impurities based on the distribution shape of the impurities; obtains the average circularity of all impurities, and obtains the impurity shape anomaly coefficient by subtracting the average circularity from 1; obtains the impurity early warning coefficient based on the impurity distribution anomaly coefficient, the impurity area anomaly coefficient, and the impurity shape anomaly coefficient; when the impurity early warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixing image is transported to the fiber combing equipment for reprocessing; A fiber abnormality recognition module constructs a fiber uniformity recognition model, identifies the first fiber distribution data and the second fiber distribution data, and obtains a fiber ratio abnormality coefficient, a fiber neatness abnormality coefficient, and a fiber uniformity abnormality coefficient; and then obtains a mixed fiber warning coefficient based on the fiber ratio abnormality coefficient, the fiber neatness abnormality coefficient, and the fiber uniformity abnormality coefficient; The calculation formula of the mixed fiber warning coefficient is: UNIF=λ1*Hunh+λ2*Neat+λ3*Juny; Juny=exp(μ1*Thic+μ2*Tanc); Among them, UNIF represents the mixed fiber warning coefficient; λ1 represents the first fiber weight; Hunh represents the fiber ratio abnormal coefficient; λ2 represents the second fiber weight; Neat represents the fiber neatness abnormal coefficient; λ3 represents the third fiber weight; Juny represents the fiber mixing abnormal coefficient; Thic represents the first thickness abnormal coefficient; μ1 represents the thickness weight; Tanc represents the second distribution abnormal coefficient; μ2 represents the distribution weight; exp represents the exponential function with the natural constant e as the base; The fiber abnormality processing module identifies the mixed fiber warning coefficient. When the mixed fiber warning coefficient does not meet the requirements, the mixed fiber corresponding to the fiber mixing image is transported to the fiber mixing equipment for further processing.

7. The production management system for high-elastic polyester fabric according to claim 6, characterized in that: In the impurity abnormality processing module, the calculation process of the impurity warning coefficient is as follows: Identify the impurity distribution data to obtain the distribution position, distribution area and distribution shape of the impurities in the fiber mixing image; According to the distribution position of impurities, the minimum circumscribed circle that can cover all impurities is obtained; according to the area ratio of the minimum circumscribed circle to the fiber mixed image, the distribution anomaly coefficient of impurities is obtained; According to the distribution area of ​​the impurities, the total area of ​​the impurities in the fiber mixing image is obtained; according to the ratio of the total area of ​​the impurities to the area of ​​the fiber mixing image, the area anomaly coefficient of the impurities is obtained.

8. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by a processor, implements the production management method for high-elastic polyester fabric according to any one of claims 1 to 5.

Images