Automatic control system and method for polyester demolding cloth production line

Abstract

The invention relates to the field of automatic control, and particularly discloses an automatic control system and method for a polyester demolding cloth production line, and the method comprises the steps: collecting tension data in real time through a tension sensor installed at a key position in warping and weaving equipment, so as to obtain a time set of warping tension values; a recurrence plot of a time set of warping tension values is calculated by using a data analysis and coding mode based on deep learning, and time sequence spatial domain features are made, so that roller speed values are intelligently recommended according to representation after self-supervised sparsification of the time sequence spatial domain features of the warping tension. In this way, the tension data at all moments in the warping process can be accurately obtained, and the integrity and accuracy of the data are ensured. Meanwhile, the optimal roller speed value is automatically calculated and recommended according to real-time tension changes, and intelligent management and control over the warping tension are achieved.

Description

Technical Field

[0001] The present application relates to the field of automatic control, and more specifically, to an automatic control system and method for a polyester release cloth production line. Background Art

[0002] Polyester release cloth, also known as polyester isolation cloth or release fabric, is a special textile material, mainly used as a temporary interface layer in the composite manufacturing process to prevent direct contact between the mold and the molding material, ensuring easy separation after curing.

[0003] In the production process of polyester release cloth, in order to ensure the quality consistency and performance reliability of the final product, each production link must be strictly controlled. Especially in the warping and weaving stages, the stability of the tension directly affects the quality of the finished product, production efficiency and the degree of wear of the equipment. During the warping process, the rotation of the drum not only pulls the yarn to move, but also determines the tension state of the yarn. Therefore, precise control of the drum speed is crucial to maintaining the appropriate tension level, which not only ensures product quality, but also improves production efficiency and reduces equipment wear.

[0004] However, traditional technologies lack effective real-time tension monitoring methods and often rely on manual experience or simple mechanical devices to adjust the warping tension. They are unable to accurately obtain the tension data at each moment and key position during the warping process, which makes it difficult to maintain a stable and appropriate warping tension level. In addition, traditional roller speed control is usually based on fixed process parameters or manual judgment, and lacks the ability to make intelligent dynamic adjustments based on real-time tension changes. In the face of tension fluctuations during the warping process, it is impossible to quickly and accurately adjust the roller speed to an appropriate value, affecting the warping quality.

[0005] Therefore, an automatic control solution for a polyester release cloth production line is desired. Summary of the invention

[0006] In order to solve the above technical problems, this application is proposed.

[0007] According to one aspect of the present application, an automatic control system for a polyester release cloth production line is provided, which includes: the polyester release cloth production line includes a raw material selection device, a warping and weaving device, a heat setting device and a finished product performance testing device; wherein the raw material selection device is used to select polyester polyester industrial yarn with an appropriate fineness as a raw material according to the set product grammage and weaving density; the warping and weaving device is used to warp and weave the raw material using a preset weaving method to obtain a woven release cloth; the heat setting device is used to perform a heat setting treatment on the woven release cloth to eliminate the internal stress in the woven release cloth to obtain the polyester release cloth; the finished product performance testing device is used to perform a physical property test on the polyester release cloth to screen out polyester release cloths of qualified quality. Finished mold cloth product; wherein, the automatic control system of the polyester release cloth production line comprises: a warping tension acquisition module, which is used to collect tension data in real time through tension sensors installed at key positions in the warping and weaving equipment to obtain a time set of warping tension values; a warping tension time-series space extraction module, which is used to perform recursive graph calculation and tension time-series space feature extraction on the time set of the warping tension values ​​to obtain a warping tension time-series space domain feature map; a warping tension self-supervised sparseness module, which is used to perform self-supervised pseudo-element feature sparseness processing on the warping tension time-series space domain feature map to obtain a sparse warping tension time-series space domain feature map; a drum speed generation module, which is used to obtain a drum speed recommendation value based on the sparse warping tension time-series space domain feature map.

[0008] According to another aspect of the present application, there is provided an automatic control method for a polyester release cloth production line, comprising: collecting tension data in real time by means of tension sensors installed at key positions in the warping and weaving equipment to obtain a time set of warping tension values; performing recursive graph calculation and tension time-series spatial feature extraction on the time set of warping tension values ​​to obtain a warping tension time-series spatial domain feature graph; performing self-supervised pseudo-element feature sparse processing on the warping tension time-series spatial domain feature graph to obtain a sparse warping tension time-series spatial domain feature graph; and obtaining a recommended value for drum speed based on the sparse warping tension time-series spatial domain feature graph.

[0009] Compared with the prior art, the present application provides an automatic control system and method for a polyester release cloth production line, which collects tension data in real time through tension sensors installed at key positions in the warping and weaving equipment to obtain a time set of warping tension values, and uses a data analysis and encoding method based on deep learning to calculate the recursive graph of the time set of warping tension values ​​and perform temporal spatial domain features, thereby intelligently recommending the roller speed value based on the self-supervised sparse representation of the warping tension temporal spatial domain features. In this way, by installing tension sensors at key positions, the tension data at each moment in the warping process can be accurately obtained to ensure the integrity and accuracy of the data. At the same time, the optimal roller speed value is automatically calculated and recommended based on the real-time tension changes, avoiding the uncertainty caused by relying on fixed parameters or manual judgment, and realizing the intelligent management and control of warping tension. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0011] Figure 1 The block diagram is a block diagram of an automatic control system for a polyester release cloth production line according to an embodiment of the present application.

[0012] Figure 2 It is a block diagram of a warping tension time-series space extraction module in an automatic control system of a polyester release cloth production line according to an embodiment of the present application.

[0013] Figure 3 It is a block diagram of a warping tension self-supervision thinning module in an automatic control system of a polyester release cloth production line according to an embodiment of the present application.

[0014] Figure 4 The present invention is a flowchart of an automatic control method for a polyester release cloth production line according to an embodiment of the present application. DETAILED DESCRIPTION

[0015] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, but rather these embodiments are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0016] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0017] It should be understandable that wind turbine blades are costly and important. In recent years, product iterations have accelerated, unit prices have fallen, and epoxy resin prices have risen, so blade factories need to find alternative materials. Polyurethane has the advantage of low price and has become the first choice for epoxy resin replacement, and some blade factories have tried to use it. However, polyurethane contains isocyanate groups that react easily with water, resulting in reduced mechanical properties of the material, and nylon release cloth absorbs water quickly, which is not conducive to production. In contrast, polyester release cloth has fewer polar groups on the surface and is not easy to absorb water. A small amount of hydroxyl will not react with isocyanate in the polyurethane system, which is conducive to the production of polyurethane system blades. As more blade factories adopt the polyurethane system, its demand will increase significantly, and the development of related polyester release cloth can meet market demand.

[0018] The polyester release cloth production line includes raw material selection equipment, warping and weaving equipment, heat setting equipment and finished product performance testing equipment; wherein the raw material selection equipment is used to select polyester polyester industrial yarn with suitable fineness as raw material according to the set product weight and weaving density; the warping and weaving equipment is used to warp and weave the raw material using a preset weaving method to obtain a woven release cloth; the heat setting equipment is used to perform heat setting treatment on the woven release cloth to eliminate the internal stress in the woven release cloth to obtain the polyester release cloth; the finished product performance testing equipment is used to perform physical property testing on the polyester release cloth to screen polyester release cloth finished products with qualified quality.

[0019] Specifically, the raw material selection device is used to select polyester industrial yarn with a suitable fineness as the raw material according to the set product weight and weaving density. It should be understood that in the production process of polyester release cloth, the raw material selection device plays a vital role.

[0020] First, to ensure that the selected raw materials meet the requirements of a specific finished product, the material selection equipment must be able to accurately understand and implement product specifications. Product weight refers to the mass per unit area, usually measured in grams per square meter (g / m²), while weave density refers to the ratio of the number of warp and weft yarns and their arrangement. Together, these two parameters define the basic structure and feel of the fabric. Therefore, before any actual operation begins, the ideal weight and weave density of the target product are determined based on the technical specifications provided by the customer or the results of market research, and this information is entered into the control system of the material selection equipment.

[0021] Next, the raw material selection equipment will obtain a series of polyester industrial yarn samples of different deniers from inventory or suppliers. The "denier" here refers to the fineness of the fiber, which is usually expressed in denier (D), that is, the weight in grams of 9,000 meters of fiber under standard conditions. The choice of denier is crucial because it directly affects key properties of the fabric, such as feel, breathability, and overall strength. Coarser fibers can provide higher strength but may sacrifice softness; conversely, finer fibers feel better but may not provide sufficient support in some applications. Therefore, finding a balance is very necessary to achieve the established product characteristics and functions.

[0022] When all the necessary parameters are ready, the raw material selection equipment starts its core algorithm and begins to analyze the data of each candidate sample. Modern raw material selection equipment is often equipped with advanced sensing technology and computing power, which can quickly evaluate a large number of samples in a short time. They detect the specific physical properties of each fiber, such as diameter, elastic modulus, elongation at break, etc., and calculate the matching degree based on the pre-set target values. For example, if a fabric that is both light and tough is required, the system will give priority to fibers with a smaller diameter and good elasticity. In addition, the equipment will also take into account various factors that may arise in the actual production process, such as the friction coefficient between yarns, the tightness of winding, etc., which are important variables affecting the quality of the final product.

[0023] It is worth noting that in addition to relying on hardware facilities, software algorithms also play an indispensable role in the selection of raw materials. By learning and accumulating historical data, intelligent algorithms can help identify which types of fiber combinations have performed best in the past, thus providing a reference for current decision-making.

[0024] Once the polyester industrial yarn that best meets the requirements is found, the raw material selection equipment will mark it and notify the relevant staff to prepare it for the next warping and weaving process. During the whole process, all operation records will be saved in detail for future traceability and verification.

[0025] Specifically, the warping and weaving equipment is used to warp and weave the raw material using a preset weaving method to obtain a woven release cloth. Accordingly, the warping and weaving equipment in the polyester release cloth production line undertakes the important task of processing the selected polyester polyester industrial yarn into a woven release cloth through a preset weaving method.

[0026] When polyester industrial yarns of suitable denier are selected and ready, they are sent to the warping equipment. Here, the yarns are first cleaned and combed to remove possible impurities or irregular parts to ensure smooth subsequent operations. Subsequently, these preliminarily treated yarns are evenly distributed on a wide reel according to predetermined product specifications such as width, length, and arrangement density. The key to this stage is to maintain the appropriate tension level, because too loose will cause the yarn to be loose and disordered, affecting the compactness of the fabric; while too tight may cause fiber breakage or deformation.

[0027] After the warping step, the yarn is then put into the weaving process. The core of the weaving equipment is the loom, which interweaves the warp and weft according to the preset weaving method. The so-called "preset weaving method" refers to the specific structure designed according to the characteristics of the target product, such as plain, twill, satin or more complex composite patterns. The loom is equipped with a precise guide device to guide the warp yarn (longitudinal yarn) and the weft yarn (transverse yarn) to interweave according to the established pattern. In this process, each yarn must fall into the specified position accurately, otherwise it will cause defects in the fabric. In order to achieve such high precision, modern weaving equipment usually uses electronic control systems to drive various moving parts, including but not limited to shuttles, needle plates, and weft beating mechanisms. At the same time, high-speed photography technology and online detection systems are also widely used to monitor the weaving process and detect and correct any possible problems in time. Once weaving is completed, what is obtained is the so-called "post-weaving demoulding cloth", which has formed the basic fabric form.

[0028] Specifically, the heat setting device is used to perform heat setting treatment on the release cloth after weaving to eliminate the internal stress in the release cloth after weaving to obtain the polyester release cloth. It should be understood that heat setting treatment is one of the key steps in the production process of polyester release cloth, which is intended to eliminate the internal stress accumulated in the release cloth after weaving during the weaving process, thereby ensuring that the finished product has excellent dimensional stability and mechanical properties. This process needs to be carried out in a specially designed heat setting device, which determines the optimal heat setting process by the influence of parameters such as setting temperature, vehicle speed, tension, overfeed rate, etc. on the setting effect.

[0029] Specifically, when the demoulding cloth is output from the weaving equipment after weaving, it is immediately sent to the heat setting equipment. Before entering the equipment, the surface of the fabric is usually cleaned once to remove possible oil or impurities to ensure the effect of subsequent treatment. Subsequently, the fabric is guided to the preheating zone, where the fabric is gradually heated to a temperature close to but below its glass transition temperature (Tg). The main purpose of this stage is to make the fiber molecular chain begin to become soft and have a certain fluidity without melting and deformation. In order to achieve the purpose of uniform heating, multiple groups of infrared heaters or hot air circulation devices are configured in the preheating zone, which can flexibly adjust the power output according to the thickness of the fabric and the characteristics of the material to ensure that every place can obtain sufficient heat.

[0030] As the fabric temperature rises, it continues to the main setting area, where the fabric will undergo the core heat setting process. A constant and uniform high temperature environment is maintained inside the main setting area, which is generally set to be slightly higher than the glass transition temperature of the fabric material, but much lower than its melting point. This temperature condition allows the macromolecules within the fiber to rearrange and release the residual internal stress caused by the weaving process. At the same time, the heat setting equipment is also equipped with a sophisticated tension control system to keep the fabric in a moderately stretched state throughout the process. Appropriate tension helps prevent the fabric from shrinking too much and can help it form a more regular and flat surface structure. It is worth noting that different types of polyester materials may have different specific requirements for heat setting. Therefore, the equipment will automatically match the most suitable operating parameters, such as temperature curve, residence time, and tension size, according to the input product specifications.

[0031] After a period of high temperature treatment, the fabric enters the cooling zone, where it is rapidly cooled and solidified. Rapid cooling can lock in the new arrangement of the fibers, giving the fabric long-lasting and stable physical properties. There are many ways to cool, including water cooling, air cooling, and even liquid nitrogen rapid freezing. The specific choice depends on factors such as the speed of the actual production line and the type of product. Regardless of the method used, the ultimate goal is to ensure that the fabric is cooled to near room temperature in a short period of time to avoid the generation of new internal stresses due to slow cooling. In addition, special attention should be paid to preventing problems such as condensation on the surface of the fabric during the cooling process, so as not to affect product quality.

[0032] Specifically, the finished product performance testing equipment is used to test the physical properties of the polyester release cloth to screen the polyester release cloth finished products with qualified quality. Accordingly, the finished product performance testing equipment plays a vital role at the end of the polyester release cloth production line. It is responsible for conducting a comprehensive and detailed physical performance test on the fabric after heat setting treatment to ensure that the final product delivered to the customer fully meets the established quality standards.

[0033] When the heat-set polyester release cloth enters the finished product performance testing stage, the first step is a preliminary appearance inspection. This inspection uses high-speed cameras and image recognition technology to capture subtle features on the fabric surface, including but not limited to color uniformity, texture consistency, and the presence of defects or foreign matter. These visual data will be transmitted to the computer system in real time for rapid analysis. Any areas that do not meet the requirements will be marked and the operator will be notified for further review. This non-contact inspection method not only improves efficiency, but also avoids secondary contamination or damage caused by manual touch.

[0034] Next comes the dimensional stability test. Since changes in temperature and tension during the heat setting process may cause slight deformation of the fabric, it is necessary to verify whether its actual size remains within the allowable range. To this end, the finished product performance testing equipment is equipped with a high-precision laser rangefinder or multi-point displacement sensor, which can perform multi-point measurements along the length and width of the fabric and record the specific values ​​of each position. The system then automatically compares these measured data with the preset standard parameters and calculates the deviation value. If it is found that the situation exceeds the allowable range, the equipment will immediately issue an alarm to remind the relevant personnel to take corrective measures.

[0035] Mechanical performance evaluation is also an indispensable part. In order to ensure that the polyester release cloth can withstand various stresses without damage in actual applications, it must be tested for multiple mechanical indicators such as tensile strength, tear strength, and abrasion resistance. This is usually done on a dedicated material testing machine, which applies the corresponding type of load according to different test requirements and accurately records the fabric's response in the process through built-in sensors, such as maximum load-bearing capacity, elongation at break, etc. At the same time, for some special-purpose products, additional test items such as bending stiffness and compression recovery rate may be added. All the data obtained will be stored as an important basis for subsequent quality judgment.

[0036] In addition to the above-mentioned static tests, the performance under dynamic conditions is equally important. For example, in the durability test under simulated use environment, the samples will be placed in a specific climate chamber and experience periodic temperature and humidity changes, ultraviolet radiation and other conditions to test their long-term stability and anti-aging ability. Throughout the process, the monitoring instrument continuously tracks and records the changing trends of various key parameters until the predetermined aging cycle is reached. The purpose of this is to ensure that even in extreme environments, the polyester release cloth can maintain good working conditions and meet the long-term use needs of users.

[0037] Finally, all qualified products that pass strict testing will be marked with a unique identification code and attached with a detailed performance report for traceability verification. Unqualified products will be stored separately and the corresponding rework or scrapping procedures will be initiated.

[0038] It should be understood that in the textile industry, a drum (also called a take-up drum, guide drum or traction drum) refers to a rotating mechanical component used to pull, guide and control the movement of yarn or fiber materials. The drum is usually made of metal or other wear-resistant materials, and the surface is specially treated to ensure proper friction with the yarn. In the warping process, the drum plays a vital role. The speed of the drum directly affects the tension level of the yarn during the entire transmission process. The appropriate tension is essential to keep the yarn neatly arranged and prevent entanglement and breakage. By accurately adjusting the drum speed, the yarn tension can be dynamically controlled to ensure that it can be maintained in a stable and suitable state under different working conditions.

[0039] However, traditional technologies lack effective real-time tension monitoring methods and often rely on manual experience or simple mechanical devices to adjust the warping tension. They are unable to accurately obtain the tension data at each moment and key position during the warping process, which makes it difficult to maintain a stable and appropriate warping tension level. In addition, traditional roller speed control is usually based on fixed process parameters or manual judgment, and lacks the ability to make intelligent dynamic adjustments based on real-time tension changes. In the face of tension fluctuations during the warping process, it is impossible to quickly and accurately adjust the roller speed to an appropriate value, affecting the warping quality.

[0040] Based on this, during the warping and weaving process, the technical concept of the present application is to collect tension data in real time through tension sensors installed at key positions in the warping and weaving equipment to obtain the time set of the warping tension value, and use data analysis and encoding methods based on deep learning to calculate the recursive graph of the time set of the warping tension value and perform temporal spatial domain features, so as to intelligently recommend the roller speed value based on the self-supervised sparse representation of the warping tension temporal spatial domain features. In this way, by installing tension sensors at key positions, the tension data at each moment in the warping process can be accurately obtained to ensure the integrity and accuracy of the data. At the same time, the optimal roller speed value is automatically calculated and recommended based on the real-time tension changes, avoiding the uncertainty caused by relying on fixed parameters or manual judgment, and realizing the intelligent management and control of the warping tension.

[0041] Figure 1 : is a block diagram of an automatic control system for a polyester release cloth production line according to an embodiment of the present application. Specifically, Figure 1As shown, according to the automatic control system 100 of the polyester release cloth production line of the embodiment of the present application, it includes: a warping tension acquisition module 110, which is used to collect tension data in real time through tension sensors installed at key positions in the warping and weaving equipment to obtain a time set of warping tension values; a warping tension time-series space extraction module 120, which is used to perform recursive graph calculation and tension time-series space feature extraction on the time set of the warping tension values ​​to obtain a warping tension time-series space domain feature map; a warping tension self-supervised sparseness module 130, which is used to perform self-supervised pseudo-element feature sparseness processing on the warping tension time-series space domain feature map to obtain a sparse warping tension time-series space domain feature map; a drum speed generation module 140, which is used to obtain a drum speed recommendation value based on the sparse warping tension time-series space domain feature map.

[0042] In the embodiment of the present application, the warping tension acquisition module 110 is used to collect tension data in real time through tension sensors installed at key positions in the warping and weaving equipment to obtain a time set of warping tension values. It should be understood that warping tension refers to the tensile force that the yarn bears during the weaving preparation process, that is, when the fibers or yarns are arranged in a parallel and tight state (warping process). Maintaining appropriate warping tension is crucial to ensuring the quality of the fabric. If the tension is too small, the yarn may be loose and disordered, resulting in an unstable fabric structure; and if the tension is too large, it may cause the yarn to break or the fabric to deform. Therefore, in order to ensure the uniformity and stability of the yarn during the warping and weaving process to improve the quality of the final product, in the technical solution of the present application, the tension data is collected in real time by the tension sensors installed at key positions in the warping and weaving equipment to obtain a time set of warping tension values, so that it can be analyzed and processed to analyze the optimal production parameter settings, such as drum speed, etc., to achieve an efficient and stable production process.

[0043] First of all, the rational configuration of tension sensors at key locations of warping and weaving equipment is the basis for effective monitoring. These sensors are usually placed at important nodes on the yarn path, such as yarn guides, drum surfaces, and near the sley. The sensors have high sensitivity and fast response capabilities, and can capture small changes in the yarn throughout the entire movement process. In order to adapt to different types of yarns and working environments, the tension sensor itself also has a certain degree of flexibility, and can adjust the measurement range and output signal type over a wide range. For example, for yarns of varying thicknesses, different sensing elements can be selected; for applications under high-speed operation or high temperature conditions, products with stronger wear resistance and anti-interference performance are required. In addition, the number and layout of sensors also need to be optimized according to specific production needs, covering all key areas that may affect tension changes, while avoiding excessive redundancy and waste of resources.

[0044] When the tension sensor is in place, it will continuously sense the actual tension on the yarn and convert this information into electrical signals for transmission to the data acquisition system. Modern tension monitoring systems are often equipped with high-performance analog-to-digital converters (ADCs) to accurately convert continuous analog signals from sensors into discrete digital forms for subsequent processing. Considering that there may be a large number of interference sources in the warping and weaving processes, such as motor noise, mechanical vibration, etc., the data acquisition system also integrates filtering circuits and noise suppression algorithms to ensure that the collected data is pure and free of noise. At the same time, in order to increase the sampling frequency and storage capacity, a large-capacity cache mechanism is used inside the system, which can record the tension state at every moment without affecting normal production. The advantage of this is that even at high speeds, the data of each instantaneous point can be fully saved, providing detailed information for subsequent analysis.

[0045] In the embodiment of the present application, the warping tension time series space extraction module 120 is used to perform recursive graph calculation and tension time series space feature extraction on the time set of the warping tension value to obtain a warping tension time series space domain feature graph. Figure 2 FIG. 1 is a block diagram of a warping tension time sequence space extraction module in an automatic control system of a polyester release fabric production line according to an embodiment of the present application. Specifically, Figure 2 As shown, the warping tension time-series space extraction module 120 includes: a warping tension recursive graph calculation unit 121, which is used to calculate the recursive graph of the time set of the warping tension values ​​to obtain the warping tension recursive graph; and a warping tension feature extraction unit 122, which is used to pass the warping tension recursive graph through a CAN-based tension space domain feature extractor to obtain the warping tension time-series space domain feature graph.

[0046] Specifically, the warping tension recursion graph calculation unit 121 is used to calculate the recursion graph of the time set of the warping tension value to obtain the warping tension recursion graph. Accordingly, considering that the warping tension value will change continuously over time, it is a dynamic data series. During the warping process, the change of tension may show complex nonlinear behavior, and the original warping tension value time set data dimension is relatively single, and traditional statistical methods are difficult to capture these characteristics. Therefore, in order to effectively reveal the repetitive patterns and structures in the time series, in the technical solution of the present application, the recursion graph of the time set of the warping tension value is calculated to convert the time series data into a two-dimensional graphic structure to obtain the warping tension recursion graph. In this way, the complex patterns of the warping tension being interrelated at different times, such as periodic fluctuations, similarities of changes in different time periods, etc., can be displayed, making the relationship originally hidden in the high-dimensional space become intuitive and visible, showing rich texture, morphology and other information, and providing a basis for the subsequent extraction of temporal spatial domain features from the spatial domain.

[0047] In particular, the specific process of calculating the recursion diagram of the time set of the warping tension values ​​to obtain the warping tension recursion diagram is as follows: Before calculating the recursion diagram, the collected time set of the warping tension values ​​must be preprocessed as necessary. This step ensures that the subsequent analysis is based on high-quality data, thereby improving the validity and reliability of the recursion diagram. The preprocessed warping tension values ​​are organized into a list of values ​​sampled at uniform intervals, providing a solid foundation for the construction of the recursion diagram.

[0048] Next is the core part of recursive graph construction. The basic principle of recursive graph is to map the relationship to a set of points in a two-dimensional plane by comparing the similarities between different points in the time series. Specifically, for a given time series , the recursive graph defines a matrix , where each element represents the and The two states are "close". The "closeness" here is usually judged based on the Euclidean distance or other metrics. When the distance between the two states is less than a set threshold When , they are considered "close" and a point is marked at the corresponding position in the recursive graph; otherwise, it is left blank. In this way, as the length of the time series increases, the recursive graph will gradually show complex patterns, intuitively showing the structural characteristics of the time series.

[0049] The key to building a recursive graph is to choose three main parameters properly: delay time , Embedding Dimension and threshold Delay time Controlling the time interval between adjacent data points determines the spatial resolution of the recurrence graph; while the embedding dimension specifies how many consecutive data points are considered as a whole for comparison. These two parameters together affect whether the recurrence graph can accurately capture the dynamic characteristics of the time series. It can be estimated by autocorrelation function or mutual information method, The optimal solution is found based on empirical rules or by trying different combinations. , which is directly related to the sparsity of the recursive graph. If it is too small, the image will be too scattered, and if it is too large, the image will be blurred. Common methods are to dynamically adjust by percentile method or fixed ratio scaling. , to ensure that the recurrence diagram can clearly express the essential characteristics of the time series and have good visual effects.

[0050] Once the above parameters are determined, the calculation of the recurrence graph can be formally started. First, the time series is reconstructed into trajectory points in the high-dimensional phase space according to the selected delay time and embedding dimension. Then, all possible point pairs are traversed. , calculate the distance between them and compare it with the threshold Compare. If the conditions are met, a point is set at the corresponding position of the recursive graph; otherwise, no mark is made. As the calculation proceeds, a recursive graph composed of dense points is gradually formed, and each point represents the reappearance of a state at a certain moment in the past or the occurrence of a similar state. The final warping tension recursive graph can not only intuitively show the change law of warping tension over time, but also contains rich spatiotemporal correlation information, laying the foundation for further feature extraction and intelligent control.

[0051] It is worth noting that recurrence graphs are not just static visualization tools, but can also be combined with other mathematical and technical means for deeper analysis. For example, using dimensionality reduction algorithms such as principal component analysis (PCA) or local linear embedding (LLE), the most representative feature vectors can be extracted from the recurrence graph, simplifying the data structure while retaining key information.

[0052] Specifically, the warping tension feature extraction unit 122 is used to pass the warping tension recursive graph through a tension spatial domain feature extractor based on CAN to obtain the warping tension time-series spatial domain feature graph. It should be understood that, considering that the warping tension has obvious spatiotemporal characteristics, in space, the yarn tensions at different positions of the warping machine affect each other; in time, the tension will change dynamically as the warping process progresses. Therefore, in order to focus on the key warping tension features and explore the correlation between the time-series features of each spatial position, in the technical solution of the present application, the warping tension recursive graph is passed through a tension spatial domain feature extractor based on CAN to obtain the warping tension time-series spatial domain feature graph. It should be understood that the CAN network, i.e., the convolutional attention network, is a deep learning model that combines a convolutional neural network (CNN) with an attention mechanism. Specifically, CNN has a powerful feature extraction capability, and performs convolution operations by sliding the convolution kernel in the convolution layer on the input data to automatically extract local features in the data, such as extracting features such as edges and textures in images, and extracting local patterns and trends in time series. As the number of network layers increases, more abstract and advanced features can be extracted. The pooling layer can reduce the dimension of features, reduce the amount of data and calculation, while retaining key features. The attention mechanism enables the network to automatically focus on the key parts of the input data when processing data. It calculates the importance weights of different parts of the input data, and then performs weighted summation or other methods of fusion on these parts, so that the network pays more attention to information useful for the task and improves the performance and efficiency of the model. Therefore, CAN (convolutional attention network) can focus on key spatial domain features in the recursive graph. For example, different areas in the recursive graph may represent the tension change characteristics of different time periods or the tension correlation of yarns at different positions. At the same time, it can handle this spatiotemporal data relationship well. It can associate the tension characteristics of adjacent yarns in space, and consider the changes of these associations in time series, so as to accurately mine the inherent laws of warping tension in both time and space dimensions. For example, by processing the recursive graph, the CAN network can discover the potential relationship between the tension change at a certain position at a certain moment and the subsequent tension change at other positions, as well as the propagation law of tension on the entire warping machine in different time periods.

[0053] In an embodiment of the present application, the warping tension self-supervised sparse module 130 is used to perform self-supervised pseudo-element feature sparse processing on the warping tension time-series space domain feature map to obtain a sparse warping tension time-series space domain feature map. Furthermore, considering that in the warping process, key tension features are crucial to ensuring product quality and optimizing production control. The original time-series space domain feature map is often mixed with a large amount of redundant information and noise, which not only increases the complexity of data processing, but also may cover up the key factors that really affect the quality. Therefore, in order to remove unimportant features and simplify the data structure, while retaining those features that are of great significance to the change of warping tension, the final feature map is more focused on reflecting meaningful changes in the actual production process. In the technical solution of the present application, the warping tension time-series space domain feature map is subjected to self-supervised pseudo-element feature sparse processing to obtain a sparse warping tension time-series space domain feature map. That is, the self-supervised learning method can learn useful representations from unlabeled data, and by constructing pseudo meta-features to evaluate the importance of each feature, it can focus on those features that are closely related to the key changes in warping tension (such as tension mutations, tension unevenness, and other issues). This can highlight features that are valuable for intelligent control and quality assessment, making subsequent analysis and decision-making more effective.

[0054] Figure 3 FIG. 1 is a block diagram of a warping tension self-supervision thinning module in an automatic control system of a polyester release fabric production line according to an embodiment of the present application. Specifically, Figure 3 As shown, the warping tension self-supervised sparseness module 130 includes: a feature discretization unit 131, which is used to discretize the features of the warping tension time-series space domain feature map to obtain a set of local feature matrices of the warping tension time-series space domain; a warping tension information pruning calculation unit 132, which is used to perform information pruning processing on each warping tension time-series space domain local feature matrix in the set of the warping tension time-series space domain local feature matrices to obtain a set of warping tension time-series space domain information pruning matrices; a warping tension information aggregation unit 133, which is used to perform feature-oriented pruning and aggregation on the set of the warping tension time-series space domain local feature matrices based on the set of the warping tension time-series space domain information pruning matrices to obtain the sparse warping tension time-series space domain feature map.

[0055] Specifically, in the embodiment of the present application, the feature discretization unit 131 is used to discretize the warping tension time series space domain feature map along the channel dimension of the warping tension time series space domain feature map to obtain a set of local feature matrices of the warping tension time series space domain. The process can be expressed by the formula: ;in, is the spatial domain characteristic diagram of the warping tension time series, is the feature discretization operation, and is the first, second, and third local feature matrix in the set of warping tension time series space domain. and A local feature matrix in the spatial domain of the warping tension time series.

[0056] It should be understood that the warping tension time series spatial domain feature diagram is feature discretized to obtain a set of warping tension time series spatial domain local feature matrices to obtain a set of warping tension time series spatial domain local feature matrices representing tension variation patterns at different time points and spatial positions.

[0057] Specifically, in the embodiment of the present application, the warping tension information pruning calculation unit 132 includes: a self-supervised pseudo-element feature extraction subunit, which is used to input the set of the warping tension temporal space domain local feature matrix into the self-supervised pseudo-element feature extraction network to obtain the warping tension temporal space domain self-supervised pseudo-element feature matrix. The process can be expressed by the formula: ;in, It is the first in the set of local characteristic matrices in the warping tension time series space domain. A local feature matrix of warping tension time series in space domain, For calculation The one-norm of yes The corresponding local norm value of the warping tension time series space domain, is a masking operation, yes The corresponding local mask value of the warping tension time series in the spatial domain, is the number of characteristic matrices in the set of local characteristic matrices in the warping tension time series space domain, It is the self-supervised pseudo-element feature matrix in the time series space domain of warping tension.

[0058] The warping tension time series space domain pruning subunit is used to calculate the information pruning matrix between the warping tension time series space domain self-supervised pseudo-element feature matrix and each warping tension time series space domain local feature matrix in the set of the warping tension time series space domain local feature matrices to obtain the set of the warping tension time series space domain information pruning matrices.

[0059] More specifically, in an embodiment of the present application, the warping tension time series space domain pruning subunit is used to: use a feature local interaction module based on a softmax function to locally prune and interact the warping tension time series space domain self-supervised pseudo-element feature matrix with each warping tension time series space domain local feature matrix in the set of the warping tension time series space domain local feature matrix to obtain the set of the warping tension time series space domain information pruning matrix. The process can be expressed by the formula: ;in, It is the first in the set of local characteristic matrices in the warping tension time series space domain. A local feature matrix of warping tension time series in space domain, is the pseudo-element feature matrix of self-supervisory warping tension temporal space domain, yes The transposed matrix of is matrix multiplication, yes The scale is the width of the matrix multiplied by the height of the matrix, is the normalization function, is the first in the set of warping tension time series spatial domain information pruning matrices A pruning matrix of the spatial domain information of the warping tension time series.

[0060] Accordingly, the set of local feature matrices of warping tension in time-series space domain is input into the self-supervised pseudo-meta feature extraction network to mine pseudo-meta features in warping tension features by means of self-supervised learning. Pseudo-meta features are an abstract representation that can represent the essential characteristics of data. Through self-supervised learning, the network can automatically find the most critical feature combinations for describing the variation law of warping tension from the local feature matrix, thereby learning the key features formed by the intrinsic correlation between warping tension at different time and space positions, and obtaining the self-supervised pseudo-meta feature matrix of warping tension in time-series space domain that can accurately reflect the core variation pattern of warping tension.

[0061] It should be understood that in order to determine which features in each local feature matrix are closely related to the key pseudo-meta features and which are relatively unimportant, the information pruning matrix between the self-supervised pseudo-meta feature matrix in the time-series space domain of the warping tension and each local feature matrix in the time-series space domain of the warping tension is calculated to construct a standard that can measure the contribution of each local feature to the overall key features, so as to dynamically remove those features that contribute less to describing the change of the warping tension in a targeted manner, retain the most valuable features, and obtain the set of information pruning matrices in the time-series space domain of the warping tension. Each element in these matrices represents the degree of correlation between the features in the corresponding local feature matrix and the self-supervised pseudo-meta features.

[0062] Specifically, in the embodiment of the present application, the warping tension information aggregation unit 133 is used to: perform feature-oriented pruning on each corresponding group of warping tension time series space domain information pruning matrices and warping tension time series space domain local feature matrices in the set of the warping tension time series space domain information pruning matrices and the set of the warping tension time series space domain local feature matrices by position point multiplication to obtain a set of sparse warping tension time series space domain local feature matrices. The process can be expressed by the formula: ;in, It is the first in the set of local characteristic matrices in the warping tension time series space domain. A local feature matrix of warping tension time series in space domain, is the first in the set of warping tension time series spatial domain information pruning matrices A warping tension time series spatial domain information pruning matrix, It is the point multiplication by position. is the first local feature matrix in the set of sparse warping tension time series space domain. A sparse local feature matrix in the spatial domain of the warping tension time series.

[0063] The set of the sparse warping tension time series spatial domain local feature matrices is subjected to feature aggregation to obtain the sparse warping tension time series spatial domain feature map. The process can be expressed by the formula: ;in, and is the first, second, and third local feature matrices in the set of sparse warping tension time series space domain. and A sparse local feature matrix of warping tension time series in spatial domain, is the feature aggregation along the channel dimension, It is the spatial domain feature map of the sparse warping tension time series.

[0064] Accordingly, each group of corresponding warping tension time-series space domain information pruning matrix and warping tension time-series space domain local feature matrix is ​​point-multiplied by position to perform feature directional pruning to obtain a set of sparse warping tension time-series space domain local feature matrices. That is, by point-multiplying by position, those features with low correlation can be weakened or removed in the local feature matrix, thereby achieving directional pruning of features. In particular, before performing point-multiplication by position, a threshold is usually set to determine which features should be completely excluded. This threshold can be dynamically adjusted according to the needs of a specific task, or it can be automatically optimized using techniques such as cross-validation to ensure that the pruned features still have clear physical meanings and are easy to interpret and apply.

[0065] Finally, the set of local feature matrices of the sparse warping tension time series space domain is aggregated to restore the overall structure of the data, and a more concise sparse warping tension time series space domain feature map is obtained while retaining key features.

[0066] In an embodiment of the present application, the drum speed generation module 140 is used to obtain a recommended drum speed value based on the sparse warping tension time-series spatial domain feature map. Specifically, in an embodiment of the present application, the drum speed generation module is used to: pass the sparse warping tension time-series spatial domain feature map through a decoder-based warping drum controller to obtain the recommended drum speed decoding value as the recommended drum speed value. That is, after the previous processing steps, the sparse warping tension time-series spatial domain feature map has extracted the tension feature information that is closely related to the warping process and is critical, but these feature information is still a relatively abstract data representation form and cannot be directly used to control the speed of the warping drum. The decoder-based warping drum controller can decode and convert this abstract feature map data into a numerical form that is compatible with the drum speed control, thereby establishing a bridge from the tension feature to the drum speed control signal. In this scenario, it can learn the complex mapping relationship between the tension feature in the sparse feature map and the optimal operating speed of the warping drum. Through a large amount of training data, the decoder can gradually adjust its own parameters to accurately convert the input feature map information into appropriate recommended drum speed decoding values, thereby achieving precise control of the warping process.

[0067] More specifically, in an embodiment of the present application, the drum speed generation module is used to: expand the sparse warping tension time-series space domain feature map into a sparse warping tension time-series space domain feature vector; multiply the decoding weight matrix of the decoder by the sparse warping tension time-series space domain feature vector to obtain the decoded sparse warping tension time-series space domain feature vector, and accumulate and sum all eigenvalues ​​of the decoded sparse warping tension time-series space domain feature vector to obtain the recommended decoding value of the drum speed.

[0068] It should be understood that here, when the warping tension time series spatial domain feature map represents the spatial domain correlation features of the warping tension recursive map, when performing feature directional pruning based on self-supervised pseudo-element features, the misalignment of the self-supervised pseudo-element of the spatial feature pattern of the time domain phase space trajectory will also cause the directional pruning distribution richness of the sparse warping tension time series spatial domain feature map, resulting in repetition of unstructured probability mapping, affecting the accuracy of the decoding results.

[0069] In a preferred example, the sparse warping tension time-series spatial domain feature map is passed through a decoder-based warping drum controller to obtain a recommended decoded value of the drum speed, including: determining the number of non-zero eigenvalues ​​in the sparse warping tension time-series spatial domain feature map, and using the width of the sparse warping tension time-series spatial domain feature map as the recommended decoded value. ,high With the number of channels The product of minus the number of non-zero eigenvalues ​​is used to obtain the zero-dimensional value of the sparse warping tension time series space domain .

[0070] Calculate the zero-dimensional value of the sparse warping tension time series space domain minus one of the zero-dimensional values ​​of the sparse warping tension time series space domain Multiply and divide respectively by the zero-dimensional value of the time series space domain of the sparse warping tension The first sparse warping tension time series spatial domain fitting value obtained after and the second sparse warping tension time series spatial domain field fitting value .

[0071] The square root of the sum of squares of all eigenvalues ​​of the sparse warping tension time series spatial domain feature map is calculated to obtain the sparse warping tension time series spatial domain pattern characterization value: ; With the said sparse warping tension time series space domain zero differentiable value Calculate each eigenvalue of the sparse warping tension time series spatial domain feature map for the index The power function , and multiplied by the first sparse warping tension time series spatial domain field fitting value To obtain the local representation of the time series of warping tension in the sparse space domain .

[0072] Calculate each eigenvalue of the sparse warping tension time series spatial domain feature map The second thinning warping tension time series spatial domain field fitting value and the spatial domain pattern representation value of the sparse warping tension time series The product of sparse warping tension time series space domain center direction value is obtained .

[0073] The sparse warping tension time series spatial domain local representation value The center direction value of the sparse warping tension time series space domain The weighted difference Composition-optimized sparse warping tension temporal spatial domain feature map.

[0074] The optimized sparse warping tension time-series spatial domain feature map is passed through a decoder-based warping drum controller to obtain a recommended decoded value for the drum speed.

[0075] Therefore, based on the vector field fitting law of high-dimensional manifolds, the isolated zeros in the sparse warping tension time series space domain feature map are used as the differentiable dimension of the high-dimensional manifold, so as to fix the normal direction pointing to the distribution center near the local coordinates represented by the eigenvalues ​​of the sparse warping tension time series space domain feature map. In this way, the invalid repetition in the probability mapping process caused by the unstructured characteristics of the rich distribution of the sparse warping tension time series space domain feature map is avoided by effectively aligning the characteristic distribution pattern of the sparse warping tension time series space domain feature map to the probability density differentiable field, thereby improving the accuracy of the decoding result obtained by the decoder-based warping drum controller input.

[0076] In summary, the automatic control system 100 of the polyester release cloth production line based on the embodiment of the present application is explained, which collects tension data in real time through tension sensors installed at key positions in the warping and weaving equipment to obtain the time set of the warping tension value, and uses data analysis and encoding based on deep learning to calculate the recursive graph of the time set of the warping tension value and perform time-series spatial domain features, so as to intelligently recommend the roller speed value based on the self-supervised sparse representation of the time-series spatial domain features of the warping tension. In this way, by installing tension sensors at key positions, the tension data at each moment in the warping process can be accurately obtained to ensure the integrity and accuracy of the data. At the same time, the optimal roller speed value is automatically calculated and recommended according to the real-time tension changes, avoiding the uncertainty caused by relying on fixed parameters or manual judgment, and realizing the intelligent management and control of the warping tension.

[0077] As described above, the automatic control system 100 of the polyester release cloth production line according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with an automatic control algorithm of the polyester release cloth production line. In a possible implementation, the automatic control system 100 of the polyester release cloth production line according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the automatic control system 100 of the polyester release cloth production line can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the automatic control system 100 of the polyester release cloth production line can also be one of the many hardware modules of the wireless terminal.

[0078] Alternatively, in another example, the automatic control system 100 of the polyester release cloth production line and the wireless terminal may also be separate devices, and the automatic control system 100 of the polyester release cloth production line may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0079] Figure 4FIG. 1 is a flow chart of an automatic control method for a polyester release cloth production line according to an embodiment of the present application. Figure 4 As shown, according to the automatic control method of the polyester release cloth production line of the embodiment of the present application, the method includes: S110, collecting tension data in real time through tension sensors installed at key positions in the warping and weaving equipment to obtain a time set of warping tension values; S120, performing recursive graph calculation and tension time-series spatial feature extraction on the time set of warping tension values ​​to obtain a warping tension time-series spatial domain feature graph; S130, performing self-supervised pseudo-element feature sparse processing on the warping tension time-series spatial domain feature graph to obtain a sparse warping tension time-series spatial domain feature graph; S140, obtaining a recommended value of the drum speed based on the sparse warping tension time-series spatial domain feature graph.

[0080] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned automatic control method for the polyester release cloth production line have been described in the above reference. Figures 1 to 3 The description of the automatic control system of the polyester release cloth production line has been described in detail, and therefore, its repeated description will be omitted.

[0081] The above specific implementation methods further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific implementation methods of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An automatic control system for a polyester release cloth production line, wherein: The polyester release cloth production line includes raw material selection equipment, warping and weaving equipment, heat setting equipment and finished product performance testing equipment; wherein the raw material selection equipment is used to select polyester polyester industrial yarn with suitable fineness as raw material according to the set product weight and weaving density; the warping and weaving equipment is used to warp and weave the raw material using a preset weaving method to obtain a woven release cloth; the heat setting equipment is used to perform heat setting treatment on the woven release cloth to eliminate the internal stress in the woven release cloth to obtain the polyester release cloth; the finished product performance testing equipment is used to perform physical property testing on the polyester release cloth to screen out qualified polyester release cloth products; characterized in that the polyester release cloth production line The automatic control system of the production line includes: a warping tension acquisition module, which is used to collect tension data in real time through tension sensors installed at key positions in the warping and weaving equipment to obtain a time set of warping tension values; a warping tension time-series space extraction module, which is used to perform recursive graph calculation and tension time-series space feature extraction on the time set of the warping tension values ​​to obtain a warping tension time-series space domain feature map; a warping tension self-supervised sparseness module, which is used to perform self-supervised pseudo-element feature sparseness processing on the warping tension time-series space domain feature map to obtain a sparse warping tension time-series space domain feature map; and a drum speed generation module, which is used to obtain a drum speed recommendation value based on the sparse warping tension time-series space domain feature map.

2. The automatic control system of the polyester release cloth production line according to claim 1, characterized in that: The warping tension time-series space extraction module includes: a warping tension recursive graph calculation unit, which is used to calculate the recursive graph of the time set of the warping tension values ​​to obtain the warping tension recursive graph; and a warping tension feature extraction unit, which is used to pass the warping tension recursive graph through a CAN-based tension space domain feature extractor to obtain the warping tension time-series space domain feature graph.

3. The automatic control system of the polyester release cloth production line according to claim 2, characterized in that: The warping tension self-supervised sparseness module includes: a feature discretization unit, which is used to discretize the features of the warping tension time-series space domain feature map to obtain a set of local feature matrices in the warping tension time-series space domain; a warping tension information pruning calculation unit, which is used to perform information pruning processing on each warping tension time-series space domain local feature matrix in the set of local feature matrices in the warping tension time-series space domain to obtain a set of information pruning matrices in the warping tension time-series space domain; a warping tension information aggregation unit, which is used to perform feature-oriented pruning and aggregation on the set of local feature matrices in the warping tension time-series space domain based on the set of information pruning matrices in the warping tension time-series space domain to obtain the sparse warping tension time-series space domain feature map.

4. The automatic control system of the polyester release cloth production line according to claim 3, characterized in that: The feature discretization unit is used to discretize the warping tension time series space domain feature map along the channel dimension of the warping tension time series space domain feature map to obtain a set of local feature matrices of the warping tension time series space domain.

5. The automatic control system of the polyester release cloth production line according to claim 4, characterized in that: The warping tension information pruning calculation unit includes: a self-supervised pseudo-element feature extraction subunit, which is used to input the set of the warping tension time series space domain local feature matrices into the self-supervised pseudo-element feature extraction network to obtain the warping tension time series space domain self-supervised pseudo-element feature matrix; a warping tension time series space domain pruning subunit, which is used to calculate the information pruning matrix between the warping tension time series space domain self-supervised pseudo-element feature matrix and each warping tension time series space domain local feature matrix in the set of the warping tension time series space domain local feature matrix to obtain the set of the warping tension time series space domain information pruning matrix.

6. The automatic control system of the polyester release cloth production line according to claim 5, characterized in that: The warping tension time series space domain pruning subunit is used to: use a feature local interaction module based on a softmax function to perform local pruning and interaction association between the warping tension time series space domain self-supervised pseudo-element feature matrix and each warping tension time series space domain local feature matrix in the set of the warping tension time series space domain local feature matrices to obtain a set of the warping tension time series space domain information pruning matrices.

7. The automatic control system of the polyester release cloth production line according to claim 6, characterized in that: The warping tension information aggregation unit is used to: perform feature-oriented pruning on each corresponding group of warping tension time series space domain information pruning matrices and warping tension time series space domain local feature matrices in the set of the warping tension time series space domain information pruning matrices and the set of the warping tension time series space domain local feature matrices according to the position point multiplication to obtain a set of sparse warping tension time series space domain local feature matrices; and perform feature aggregation on the set of sparse warping tension time series space domain local feature matrices to obtain the sparse warping tension time series space domain feature map.

8. The automatic control system of the polyester release cloth production line according to claim 7, characterized in that: The drum speed generation module is used to: pass the sparse warping tension time-series spatial domain feature map through a decoder-based warping drum controller to obtain the drum speed recommended decoding value as the drum speed recommended value.

9. An automatic control method for a polyester release cloth production line according to the automatic control system of the polyester release cloth production line according to claim 1, characterized in that: include: The tension sensor installed at the key position of the warping and weaving equipment collects tension data in real time to obtain the time set of warping tension value; Performing recursive graph calculation and tension time series spatial feature extraction on the time set of the warping tension value to obtain a warping tension time series spatial domain feature graph; performing self-supervised pseudo-element feature sparse processing on the warping tension time series spatial domain feature graph to obtain a sparse warping tension time series spatial domain feature graph; Based on the sparse warping tension time series spatial domain characteristic diagram, a recommended value of the drum speed is obtained.

10. The automatic control method for a polyester release cloth production line according to claim 9, characterized in that: The recursive graph calculation and tension time-series space feature extraction are performed on the time set of the warping tension values ​​to obtain the warping tension time-series space domain feature graph, including: calculating the recursive graph of the time set of the warping tension values ​​to obtain the warping tension recursive graph; passing the warping tension recursive graph through a tension space domain feature extractor based on CAN to obtain the warping tension time-series space domain feature graph.

Images