An intelligent textile production line control method and system

Abstract

The present invention discloses an intelligent textile production line control method and system, which relates to the field of industrial automation technology, and includes: collecting operation data in the textile production process in real time and performing data preprocessing; dynamically adjusting the process parameters of the textile production line based on the preprocessed data; real-time detecting the quality status in the textile production process and adjusting the production line operation based on the detection results; constructing a multi-feature fusion fault prediction model to predict the operation status of the production line equipment; identifying potential faults in advance according to the prediction results and adjusting the operation mode of the production line. By combining multi-feature fusion analysis, the present invention dynamically optimizes the operation status of equipment, identifies and adjusts potential faults in advance, improves production quality and efficiency, reduces maintenance costs and downtime risks, is significantly superior to traditional methods, and provides strong technical support for the intelligent upgrading of the textile industry.

Description

Technical Field

[0001] The present invention relates to the field of industrial automation technology, and particularly to an intelligent control method and system for a textile production line. Background Art

[0002] With the rapid advancement of Industry 4.0, textile production lines are gradually developing towards intelligence and automation. Traditional textile production lines are mainly mechanized and rely on manual monitoring and empirical adjustment of production parameters. However, with the continuous growth of the market demand for high-quality textiles, the limitations of the traditional method are gradually emerging. In recent years, the widespread application of Internet of Things (IoT) technology, artificial intelligence (AI) algorithms, and big data analysis technology has made it possible to upgrade the intelligence of textile production lines. For example, real-time collection and analysis of the operation data during the production process can optimize the production process; the quality inspection technology based on machine vision significantly improves the efficiency of defect identification; the prediction model combined with the equipment operation status helps to reduce the downtime caused by equipment failures. Nevertheless, there are still many deficiencies in the current textile production line control technology, making it difficult to achieve comprehensive optimization and intelligent control of the entire production process.

[0003] Although the existing textile production line control technology has made progress in data collection, process adjustment, quality inspection, etc., there are still significant deficiencies and it is difficult to meet the requirements of efficient and intelligent production. First, in terms of data processing, most technologies only stay in the simple sensor data collection stage and fail to achieve deep integration and feature extraction of multi-source data, resulting in poor real-time performance and accuracy of production parameter adjustment. Second, in terms of dynamically adjusting the production line process parameters, the existing technology often relies on a control logic based on fixed rules and lacks the intelligent adjustment ability combined with real-time data, making it difficult to cope with complex production environment changes. Third, in the field of quality inspection, most of the existing inspection systems are based on static rule settings and cannot efficiently identify the dynamic quality problems of products, such as surface defects and color uniformity changes. In addition, there are also obvious deficiencies in the existing technology for equipment failure prediction, mainly reflected in: the monitoring means of the equipment operation status are single, and it is impossible to effectively predict potential multi-feature failures; there is a lack of a comprehensive health scoring mechanism, making it difficult to quantify the equipment status and guide the production line adjustment. Finally, in terms of the adjustment of the operation mode, the existing methods mostly rely on manual intervention and it is difficult to form a real-time closed-loop optimization control. In summary, the existing technology cannot achieve the comprehensive intelligent control of the textile production line. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the technical problem to be solved by the present invention is that the existing control methods for textile production lines have insufficient data processing capabilities, lack of real-time dynamic process adjustment, limited equipment failure prediction capabilities, and the problem of how to achieve comprehensive intelligent control of the production line.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, an embodiment of the present invention provides an intelligent control method for a textile production line, including: collecting operation data during the textile production process in real time and performing data preprocessing; dynamically adjusting the process parameters of the textile production line based on the preprocessed data; real-time detecting the quality status during the textile production process and adjusting the operation of the production line based on the detection results; constructing a multi-feature fusion fault prediction model to predict the operation status of the production line equipment; and identifying potential faults in advance according to the prediction results and adjusting the operation mode of the production line.

[0008] As a preferred solution of the intelligent control method for a textile production line according to the present invention, wherein: the real-time collection of operation data during the textile production process includes operation parameter data of production equipment, environmental parameter data, and product quality data;

[0009] The operation parameter data of the production equipment includes motor speed, equipment load, bearing temperature, and vibration signal;

[0010] The environmental parameter data includes temperature, humidity, and air flow rate in the production area;

[0011] The product quality data includes the tension of the textile, weft density, and surface color uniformity.

[0012] As a preferred solution of the intelligent control method for a textile production line according to the present invention, wherein: the data preprocessing includes noise filtering and time-domain feature extraction for the operation parameter data of the production equipment, wherein the vibration signal is processed by the wavelet denoising method, and the motor speed and equipment load data are smoothed by filtering to eliminate instantaneous fluctuations;

[0013] Performing time series smoothing and multi-dimensional data fusion on the environmental parameter data, processing the temperature and humidity data by the moving average method, and performing trend analysis on the air flow rate to evaluate the stability of the production environment;

[0014] Performing feature extraction and anomaly detection on the product quality data, wherein the textile tension and weft density data are used to eliminate outliers by threshold judgment, and the surface color uniformity data is optimized by the histogram equalization method to optimize the data distribution.

[0015] As a preferred solution of the intelligent textile production line control method described in the present invention, wherein: the dynamic adjustment of the process parameters of the textile production line includes calculating the real-time change trend of the motor speed using a moving average algorithm based on the operation parameter data of the production equipment, and dynamically adjusting the spinning speed and loom tension adjustment parameters in combination with the equipment load data;

[0016] Based on the environmental parameter data, the PID control method is used to adjust the operation state of the humidifying device or air conditioning system, and the output power is adjusted in real time to maintain the temperature and humidity in the production area within the target range;

[0017] Based on the product quality data, combined with the correlation between the tension and the weft density, an optimization model is constructed to dynamically adjust the weft density, and adjustment parameters for the dye liquor concentration and coating speed are generated based on the color deviation analysis method for optimizing the dyeing and finishing process.

[0018] As a preferred solution of the intelligent textile production line control method described in the present invention, wherein: the adjustment of the production line operation includes obtaining the surface image of the textile through a machine vision system, and analyzing the type and position of the defects using a convolutional neural network algorithm to generate an adjustment signal to control the loom operation mode;

[0019] Obtain real-time tension data through a tension sensor, calculate the tension deviation value, and generate a signal for adjusting the loom running speed and weft input speed;

[0020] Real-time detection of the weft density data, and optimization of the feeding speed and operation mode of the spinning equipment in combination with the dynamic adjustment model;

[0021] Based on the surface color uniformity detection result, use the dynamic color difference analysis model to calculate the operation parameters of the dyeing and finishing equipment, and adjust the dye liquor concentration and coating speed;

[0022] Through the above detection and adjustment, a closed-loop feedback mechanism is formed to continuously update the equipment operation parameters to adapt to the real-time quality state changes.

[0023] As a preferred solution of the intelligent textile production line control method described in the present invention, wherein: the construction of the multi-feature fusion fault prediction model includes, based on the vibration signal of the production line equipment operation, the collected vibration time series data Perform a fast Fourier transform, calculate the frequency domain energy eigenvalue, and the calculation formula is:

[0024] ;

[0025] Wherein, is the vibration signal eigenvalue, and are the lower and upper limits of the vibration signal frequency respectively, is the frequency domain eigenvalue of the vibration signal; is the attenuation factor of the high-frequency signal, which is used to reduce the influence of high-frequency noise on the result; is the frequency variable of the current integration; is the small increment of frequency, which is the basic unit of frequency domain integration and is used to gradually accumulate the signal characteristics in the frequency domain range;

[0026] Based on the bearing temperature data, perform second derivative calculation and integral trend analysis on the time series to extract the temperature change eigenvalue, and the calculation formula is:

[0027] ;

[0028] where, is the temperature eigenvalue, is the time series function of the bearing temperature; represents the second derivative of the temperature change, which is used to evaluate the acceleration trend of the temperature; is the partial derivative; is the attenuation integral of the temperature trend, is the temperature fluctuation suppression factor; is the time series 's sine transform, which reflects the periodic fluctuation characteristics of the temperature change; is the time, which is the basic variable of the time series data;

[0029] Based on the motor current data, perform time series analysis on the collected current signal to extract the abnormal characteristics of the current signal through the autoregressive moving average model, and the eigenvalue calculation formula is:

[0030] ;

[0031] where, is the current eigenvalue, is the th current sampling value, is the non-linear amplification factor; represents the summation processing of current sampling points;

[0032] The prediction of the operation state of the production line equipment includes generating an equipment health score by synthesizing the eigenvalues , and its calculation formula is:

[0033] ;

[0034] where, are the normalization results of the vibration, temperature and current eigenvalues respectively; is the total number of feature values, indicating the number of features considered in the calculation of the health score; For the A characteristic value represents the operating status of the device in a certain dimension; For the The reference value of the characteristic value is used to compare the deviation between the actual characteristic value and the target state; It is a characteristic difference inhibitor;

[0035] The normalization formula is:

[0036] ;

[0037] in, is the normalized eigenvalue, is the original eigenvalue, is the maximum eigenvalue, is the minimum eigenvalue.

[0038] As a preferred solution of the intelligent textile production line control method of the present invention, the method of identifying potential faults in advance according to the prediction results includes: The potential failure risks are divided into multiple levels according to different value ranges, and corresponding production line adjustment strategies are adopted for different levels of risks;

[0039] When health score In the high-risk range, When the load is too low, the equipment with the abnormality will be stopped and the standby equipment will be switched to ensure the continuity of the production line. The overall speed of the production line will be reduced and the spinning load will be redistributed to the low-load equipment. The adjusted load distribution formula is:

[0040] ;

[0041] in, For the The load distribution value of each device, is the total load, Weight the device health score;

[0042] When health score In the medium risk range, When the equipment operating parameters are adjusted dynamically, including reducing the spinning speed, tension adjustment value and coating speed of the dyeing and finishing process, the dynamic adjustment formula is:

[0043] ;

[0044] in, are the adjusted operating parameters, is the current running parameter value, is the reference health score; it monitors the running status of the device in real time and detects whether there is a new downward trend in the health score;

[0045] When the health score is in the low-risk range, then maintain the current operating mode, regularly generate a device health report, and evaluate possible long-term hidden dangers;

[0046] The adjustment process is achieved through a closed-loop feedback mechanism, that is, after adjusting the operating mode, recalculate the device health score and continuously optimize the running status of the production line until the health scores of all devices are higher than the safety threshold.

[0047] In a second aspect, an embodiment of the present invention provides an intelligent textile production line control system, including:

[0048] A data acquisition module: It collects the running data in the textile production process in real time and performs data preprocessing;

[0049] A parameter adjustment module: Based on the preprocessed data, dynamically adjust the process parameters of the textile production line;

[0050] An operation adjustment module: It detects the quality status in the textile production process in real time and adjusts the production line operation based on the detection results;

[0051] A fault prediction module: Construct a multi-feature fusion fault prediction model to predict the running status of the production line equipment;

[0052] An operation mode optimization module: Identify potential faults in advance according to the prediction results and adjust the operation mode of the production line.

[0053] In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements the steps of the method described above.

[0054] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the method described above.

[0055] Advantages of the present invention: By collecting the operation data in the textile production process in real time and performing multi-source fusion and deep feature extraction, the efficient utilization and standardization of data are achieved; by dynamically adjusting the process parameters of the textile production line, the real-time performance and accuracy of the production process are improved; by performing real-time quality inspection and dynamically adjusting the production line operation based on the inspection results, the product qualification rate and the stability of the production line are significantly improved; by constructing a multi-feature fusion fault prediction model to predict the operation state of the equipment, potential faults are identified in advance, reducing the equipment downtime and maintenance costs; by optimizing and adjusting the production line operation mode based on the prediction results, a real-time closed-loop intelligent control system is formed, realizing the comprehensive intelligent, automated and efficient operation of the textile production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings, where:

[0057] Figure 1 is the overall flowchart of an intelligent textile production line control method provided by the first embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0058] In order to make the above objects, features and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention will be described in detail below with reference to the drawings of the specification. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0059] Embodiment 1, referring to Figure 1 , which is an embodiment of the present invention, provides an intelligent textile production line control method, including:

[0060] S1: Collect the operation data in the textile production process in real time and perform data preprocessing.

[0061] Collecting the operation data in the textile production process in real time includes the operation parameter data of the production equipment, the environmental parameter data and the product quality data.

[0062] The operating parameter data of the production equipment include motor speed, equipment load, bearing temperature, and vibration signal. Motor speed: It is used to monitor the operating speed of the textile production line equipment, reflects the dynamic performance of the production line equipment, and is an important input variable for the dynamic adjustment of process parameters. Equipment load: It represents the current load status of the textile production line equipment. By monitoring the load changes, the stability and efficiency of the equipment operation can be evaluated. Bearing temperature: The bearing temperature is an important indicator parameter for the healthy state of the equipment operation. Abnormal temperature usually indicates the potential risk of equipment failure. Vibration signal: The vibration signal can characterize the dynamic changes during the operation of mechanical equipment and is one of the core parameters for detecting equipment failures.

[0063] The environmental parameter data include the temperature, humidity, and air velocity in the production area. These data are used to evaluate the stability of the textile production environment. Especially during the production of high-quality textiles, minor changes in the environment may directly affect the physical properties of the textiles (such as weft density and tension).

[0064] The product quality data include the tension, weft density, and surface color uniformity of the textiles. Tension of textiles: It is the core index to measure the stress state of the yarn during the textile process. Tension changes will directly affect the weft density and the uniformity of the textiles. Weft density: It characterizes the number of weft yarns per unit area of the textiles and is an important index for evaluating the quality of the textiles and the production process parameters. Surface color uniformity: It is the core parameter to quantify the appearance quality of the textiles through optical characteristics and is usually optimized and evaluated by the color histogram analysis method.

[0065] Data preprocessing includes noise filtering and time-domain feature extraction for the operating parameter data of the production equipment. Among them, the vibration signal is processed by the wavelet denoising method, and the instantaneous fluctuations of the motor speed and equipment load data are eliminated through smoothing filtering;

[0066] Time series smoothing and multi-dimensional data fusion are performed on the environmental parameter data. The sliding average method is used to process the temperature and humidity data, and the trend analysis of the air velocity is carried out to evaluate the stability of the production environment;

[0067] Feature extraction and anomaly detection are performed on the product quality data. Among them, the abnormal values of the tension and weft density data of the textiles are removed through threshold judgment, and the surface color uniformity data is optimized by the histogram equalization method for the data distribution.

[0068] It should be noted that for the non-stationary characteristics of vibration signals, the wavelet denoising method can effectively extract key frequency domain features, while filtering out high-frequency noise and enhancing the signal quality. Smoothing filtering is used to eliminate the instantaneous fluctuations in the motor speed and equipment load to ensure that the data is more stable and reliable. The moving average method is used to process the temperature and humidity data, eliminating the short-term fluctuation trend and retaining the long-term change trend, improving the timeliness and accuracy of environmental data. Histogram equalization is used to optimize the distribution of color uniformity data, and by enhancing the contrast, it can more accurately reflect the surface quality of textiles.

[0069] S2: Based on the preprocessed data, dynamically adjust the process parameters of the textile production line.

[0070] Based on the operating parameter data of the production equipment, use the moving average algorithm to calculate the real-time change trend of the motor speed, and dynamically adjust the spinning speed and loom tension adjustment parameters in combination with the equipment load data.

[0071] It should be noted that in the textile production line, the acquisition of motor speed data is a continuous process. However, real-time motor speed data is often affected by instantaneous fluctuations, such as mechanical vibration, electrical noise, or short-term interference during the production process. Although these fluctuations do not significantly change the long-term operating state of the motor, if these raw data are directly used, it will lead to instability in the production control process, especially when adjusting the spinning speed and loom tension, unnecessary fluctuations may be introduced.

[0072] The moving average algorithm is a commonly used method for smoothing data. By calculating the average value of a set of recent sampled data, it removes short-term fluctuations and retains the long-term change trend of the data. In the production line, the moving average method can, according to the set time window, such as the most recent 5 to 10 sampling points, average these speed values to provide a more stable speed trend. This smooth speed change curve provides a stable input for subsequent dynamic adjustment.

[0073] In textile production, the spinning speed needs to be coordinated with the motor speed and must also adapt to the real-time changes in the equipment load. The equipment load represents the working load state of the equipment. Excessive load may lead to increased equipment wear, while too low a load will affect production efficiency. The goal of dynamically adjusting the spinning speed is to ensure that the equipment operates efficiently within a safe range.

[0074] Combined with the stable rotational speed trend calculated by the moving average method, the adjustment of the spinning speed follows the following logic: When the stable trend of the motor speed shows a relatively high speed, the spinning speed can be appropriately increased to make full use of the operating capacity of the equipment and improve production efficiency; if the equipment load shows a relatively high working pressure, the spinning speed needs to be appropriately reduced to avoid equipment overload and operating failures; when the motor speed is stable but the load is low, the utilization rate of the equipment can be increased by increasing the spinning speed, while preventing spinning quality problems caused by too low a load.

[0075] This dynamic adjustment strategy ensures the stability of equipment operation and optimizes the spinning speed to match the real-time equipment status.

[0076] The loom tension is an important parameter affecting the quality of textiles and is closely related to the spinning speed, motor speed, and equipment load. During the dynamic adjustment process, it is necessary to adjust the tension in a timely manner according to the real-time equipment operating status, especially the changes in load and speed, to ensure the consistency of spinning quality and weft density.

[0077] The adjustment logic of the loom tension includes the following points: When the equipment load increases, the loom tension should be appropriately increased to ensure the stability of the yarn under high-load operation and prevent problems such as uneven weft density or broken yarn caused by yarn relaxation; when the equipment load decreases, the loom tension should be appropriately reduced to avoid excessive tensile force on the yarn, resulting in yarn deformation or breakage; if the motor speed changes smoothly, the tension adjustment should be prioritized according to the load change; if the speed fluctuates greatly, the speed and load data should be comprehensively used for dynamic optimization.

[0078] Through this dynamic adjustment, the loom tension can always adapt to the current equipment operating status, ensuring the quality and consistency of textiles.

[0079] Furthermore, although the moving average algorithm is an efficient and simple method, in specific scenarios, the following alternative solutions can be considered:

[0080] Exponentially Weighted Moving Average Algorithm (EWMA): EWMA assigns higher weights to the latest data points and gradually reduces the weights of historical data, enabling it to respond more quickly to the changing trend of the motor speed and being suitable for production scenarios that require rapid adjustment.

[0081] Median Filtering: If there are significant outliers (abnormal data) in the motor speed data, median filtering can effectively eliminate the interference of these anomalies on the adjustment process and provide a more reliable trend judgment.

[0082] Kalman Filtering: In complex dynamic systems, the Kalman filter provides optimized speed data through prediction and update loops, which can not only smooth the data but also anticipate trend changes in advance.

[0083] Furthermore, the application of the moving average algorithm significantly reduces the interference of short-term fluctuations in motor speed on the spinning speed and loom tension adjustment, providing a more reliable input for subsequent dynamic adjustment. Dynamically adjusting the spinning speed can avoid equipment overload while maintaining high efficiency, improving the overall stability of the production line. The dynamic adjustment of loom tension ensures the uniformity of yarn tension, reduces the occurrence probability of uneven weft density and broken yarn, and greatly improves the finished product quality of textiles.

[0084] Based on the environmental parameter data, the PID control method is used to adjust the operating state of the humidifying device or air conditioning system, and the output power is adjusted in real time to maintain the temperature and humidity in the production area within the target range; the proportional integral derivative (PID) control method calculates the error in real time and dynamically adjusts the output power to ensure that the temperature and humidity are maintained within the target range, which is a classic method in industrial control.

[0085] It should be noted that PID control is an automated adjustment method widely used in industrial control systems. By the collaborative work of three parts: proportional (P), integral (I), and derivative (D), the output power of the humidifying device or air conditioning system is adjusted in real time: Proportional control: Directly adjusts the output power according to the magnitude of the error. The larger the error, the greater the adjustment amplitude, which is used to quickly respond to changes in temperature and humidity; Integral control: Considers the accumulation of errors and gradually corrects the environmental deviation to prevent the system from deviating from the target value for a long time; Derivative control: Adjusts in advance according to the trend of error change to avoid overshoot or under-adjustment.

[0086] The core of PID control is to dynamically balance the roles of the three parts, and achieve precise adjustment by coordinating the contributions of proportional, integral, and derivative.

[0087] According to the calculation results of PID control, the system adjusts the operating state of the humidifying device or air conditioning system in real time:

[0088] Humidifying device: When the humidity is lower than the target value, the humidifying device increases the steam output to gradually increase the air humidity; when the humidity is higher than the target value, it reduces or suspends the steam output to avoid excessive humidity affecting the spinning quality;

[0089] Air conditioning system: If the temperature is higher than the target value, the air conditioning system increases the cooling power to lower the temperature in the production area; if the temperature is lower than the target value, the air conditioning system switches to the heating mode to increase the temperature.

[0090] The adjustment process is always based on real-time data and continuously updates the adjustment actions as the environmental conditions change, ensuring that the temperature and humidity in the production environment are always maintained within the target range.

[0091] After each adjustment of the output power is completed, the system continuously collects the latest temperature and humidity data and calculates the updated error. These errors are then input back into the PID control system to correct the adjustment actions. This closed-loop feedback mechanism ensures that: the system can quickly respond to new environmental changes; avoid over-adjustment or under-adjustment; and maintain the stability of temperature and humidity in the long term.

[0092] Furthermore, PID control can adjust the output power in real time. Even when environmental parameters change rapidly, it can quickly correct the temperature and humidity deviation, preventing environmental fluctuations from affecting the quality of textiles. Through the PID control method, the system can adjust the operating states of the humidifying device and the air conditioner in real time according to small changes in the environment, significantly improving the control accuracy compared with traditional time-based control or manual adjustment methods. A stable temperature and humidity environment helps to improve the consistency of yarn tension, reduce the unevenness of weft density, and avoid the negative impacts of excessive dryness or moisture on textile performance.

[0093] Based on the product quality data, combined with the correlation between tension and weft density, an optimization model is constructed to dynamically adjust the weft density, and adjustment parameters for dye liquor concentration and coating speed are generated based on the color deviation analysis method for optimizing the dyeing and finishing process.

[0094] It should be noted that fluctuations in spinning tension usually lead to changes in weft density. For example, excessive tension may cause too high weft density, while too small tension may result in uneven weft density.

[0095] By analyzing historical data, a correlation model between tension and weft density is constructed. This model directly correlates tension fluctuations with changes in weft density and is used to predict the real-time change trend of weft density.

[0096] Based on the correlation model between tension and weft density, the optimized adjustment parameters of weft density are calculated in real time. For example, when excessive tension is detected, the system automatically reduces the speed of the weft density adjustment device to keep the density within the target range.

[0097] The adjustment goal of weft density is to ensure its uniformity while meeting the physical performance requirements of the fabric. Traditional weft density adjustment methods usually rely on manual experience and are difficult to quickly respond to the dynamic changes in tension during the production process. The present invention realizes the automatic dynamic adjustment of weft density through the correlation model, significantly improving the accuracy of adjustment.

[0098] Color deviation analysis: By comparing the actual color with the target color, a color deviation value is generated to guide the adjustment of dye liquor concentration and coating speed.

[0099] When the deviation is large, increase the dye liquor concentration to enhance color saturation.

[0100] When the deviation is small, appropriately reduce the concentration to avoid over-dyeing.

[0101] Coating speed optimization: Color deviation is also used to adjust the coating speed:

[0102] When the color deviation is large, slow down the coating speed to increase the penetration time of the dye solution.

[0103] When the color deviation is small, speed up the coating speed to improve production efficiency.

[0104] Furthermore, based on the comprehensive color deviation analysis results, the system automatically generates adjustment parameters for the dye solution concentration and coating speed, and transmits them to the dyeing and finishing equipment in real time for execution.

[0105] This process forms a closed-loop feedback mechanism, continuously optimizing the operating state of the dyeing and finishing process by continuously monitoring the color deviation. Traditional dyeing and finishing process control mostly uses fixed parameters and cannot respond to the dynamic changes of color deviation in real time. Through color deviation analysis, the present invention dynamically adjusts the dye solution concentration and coating speed, realizing the intelligent optimization of the dyeing and finishing process.

[0106] S3: Real-time detect the quality status during the textile production process, and adjust the production line operation based on the detection results.

[0107] Obtain the surface image of the textile through a machine vision system, and use a convolutional neural network algorithm to analyze the type and location of defects, generating adjustment signals to control the loom operation mode. Adjustment signals generated according to the type and location of defects, for example: if it is identified as a warp break, control the loom to pause and start a maintenance signal. If uneven weft distribution is detected, adjust the weft input speed to improve uniformity.

[0108] Obtain real-time tension data through a tension sensor, calculate the tension deviation value, and generate signals for adjusting the loom running speed and weft input speed. For example: collect spinning tension data, compare it with the target tension, and calculate the tension deviation value. When the tension is too high, reduce the loom running speed to reduce the tension. When the tension is too low, increase the weft input speed to maintain the yarn stability.

[0109] Real-time detect the weft density data, and optimize the feeding speed and operation mode of the spinning equipment in combination with a dynamic adjustment model. For example: when the weft density is too high, the system slows down the feeding speed to avoid yarn accumulation; when the weft density is too low, the system speeds up the feeding speed to increase the density.

[0110] Based on the surface color uniformity detection results, use a dynamic color difference analysis model to calculate the operating parameters of the dyeing and finishing equipment, and adjust the dye solution concentration and coating speed. For example: when the color difference is too large, the system increases the dye solution concentration or slows down the coating speed to increase the dyeing saturation; when the color difference is too small, the system reduces the dye solution concentration or speeds up the coating speed to improve production efficiency.

[0111] A closed-loop feedback mechanism is formed through the above detections and adjustments to continuously update the equipment operation parameters to adapt to real-time quality status changes.

[0112] It should be noted that during the production process, when the machine vision system detects spots and defects on the surface of textiles, the system immediately reduces the weft input speed and adjusts the operation mode to prevent the expansion of defects. When the tension sensor detects excessive tension, the system reduces the loom operation speed and appropriately reduces the yarn tension, and finally returns to the normal state. In the dyeing and finishing process, when the machine vision detects an increase in color deviation, the system automatically increases the dye liquor concentration and slows down the coating speed to ensure uniform dyeing.

[0113] S4: Construct a multi-feature fusion fault prediction model to predict the operation status of production line equipment.

[0114] Based on the vibration signals of the production line equipment operation, for the collected vibration time series data perform a fast Fourier transform to calculate the frequency domain energy eigenvalue, and the calculation formula is:

[0115] ;

[0116] where is the vibration signal eigenvalue, and are respectively the lower limit and upper limit of the frequency of the vibration signal, is the frequency domain eigenvalue of the vibration signal; is the attenuation factor of the high-frequency signal, used to reduce the influence of high-frequency noise on the result; is the frequency variable of the current integration; is the small increment of frequency, which is the basic unit of frequency domain integration and is used to gradually accumulate the signal characteristics within the frequency domain range.

[0117] The vibration signal reflects the dynamic state during equipment operation, and its frequency domain eigenvalue is calculated through the fast Fourier transform (FFT) and represents the energy distribution of the signal within a specific frequency range. The lower limit and upper limit of the frequency of the vibration signal are used to define the frequency range of interest and avoid the influence of interference frequency bands. The introduction of the high-frequency attenuation factor suppresses the interference of high-frequency noise on the eigenvalue calculation and ensures that the eigenvalue is more reliable.

[0118] Based on the bearing temperature data, perform a second-order derivative calculation and integral trend analysis on the time series to extract the temperature change eigenvalue, and the calculation formula is:

[0119] ;

[0120] Among them, is the temperature characteristic value, is the time series function of the bearing temperature; represents the second derivative of the temperature change, which is used to evaluate the acceleration trend of the temperature; is the partial derivative; is the decay integral of the temperature trend, is the temperature fluctuation suppression factor; is the time series The sine transform of, reflecting the periodic fluctuation characteristics of the temperature change; is time, which is the basic variable of the time series data.

[0121] The bearing temperature reflects the thermal state of the equipment, and its change trend is an important basis for judging equipment abnormalities. By performing second derivative and integral analysis on the temperature time series Acceleration trend and long-term fluctuation trend characteristics are extracted. The second derivative is used to evaluate the severity of the temperature change, and the integral trend is used to quantify the periodic fluctuation characteristics of the temperature change. The temperature fluctuation suppression factor suppresses the influence of violent fluctuations on the characteristic value and ensures the smoothness of feature extraction.

[0122] Based on the motor current data, the collected current signal is subjected to time series analysis, and the abnormal characteristics of the current signal are extracted through the autoregressive moving average model. The characteristic value calculation formula is:

[0123] ;

[0124] Among them, is the current characteristic value, is the th current sampling value, is the non-linear amplification factor; represents the summation processing of current sampling points.

[0125] The current signal is an important indicator of the electrical state of the equipment. Abnormal characteristics are extracted through time series analysis and the autoregressive moving average model (ARMA). The sampling point is the collected current signal value, and the summation operation reflects the overall trend of the signal. The non-linear amplification factor is used to amplify the influence of key feature points and enhance the sensitivity of the model to abnormal features.

[0126] Predicting the operating state of the production line equipment includes synthesizing characteristic values to generate an equipment health score , and its calculation formula is:

[0127] ;

[0128] wherein, are respectively the normalization results of vibration, temperature and current eigenvalue; is the total number of eigenvalues, representing the number of features considered in the health score calculation; is the th eigenvalue, representing the operating state of the device in a certain dimension; is the th reference value of the eigenvalue, used to compare the deviation between the actual eigenvalue and the target state; is the feature difference suppression factor.

[0129] The normalization formula is:

[0130] ;

[0131] wherein, is the normalized eigenvalue, is the original eigenvalue, is the maximum value of the eigenvalue, is the minimum value of the eigenvalue.

[0132] The health score is comprehensively generated through the normalization results of vibration, temperature and current eigenvalues, and is used to quantify the operating state of the device. The normalization formula ensures that the dimensions of the eigenvalues are consistent, and the calculation method of the lowest eigenvalue dominating the score result (using the suppression factor to dynamically adjust the weight) avoids the influence of high eigenvalues covering up key low eigenvalues and improves the accuracy of the prediction result.

[0133] S5: Identify potential faults in advance according to the prediction results and adjust the operating mode of the production line.

[0134] According to different value ranges of the device health score , the potential fault risks are divided into multiple levels, and corresponding production line adjustment strategies are adopted for different levels of risks.

[0135] When the health score is in the high-risk interval, , then stop the operation of the abnormal device, and at the same time switch to the standby device to ensure the continuity of the production line; reduce the overall operating speed of the production line, and reallocate the spinning load to the low-load devices. The adjusted load distribution formula is:

[0136] ;

[0137] wherein, is the The load distribution value of the device, is the total load, and is the weight of the device health score.

[0138] When the health score is in the medium-risk range, then the device operating parameters are dynamically adjusted, including reducing the spinning speed, the tension adjustment value, and the coating speed in the dyeing and finishing process. The dynamic adjustment formula is:

[0139] ;

[0140] where, is the adjusted operating parameter, is the current operating parameter value, and is the reference health score; The operating status of the device is monitored in real time to detect whether there is a new downward trend in the health score.

[0141] When the health score is in the low-risk range, then the current operating mode is maintained, and a device health report is generated regularly to evaluate possible long-term hidden dangers.

[0142] The adjustment process is achieved through a closed-loop feedback mechanism, that is, after adjusting the operating mode, the device health score is recalculated and the operating status of the production line is continuously optimized until the health scores of all devices are higher than the safety threshold.

[0143] It should be noted that dividing the health score into multiple risk levels (high, medium, low) can quickly identify different risk levels of the device status, avoid the deficiencies of a unified strategy, and improve the pertinence of the adjustment. Traditional fault management methods often cannot distinguish risk levels, resulting in overly conservative or radical adjustment strategies. The present invention realizes risk classification and differential adjustment by finely dividing the health score value range. When the device is in the high-risk range, it stops running and switches to a standby device, effectively preventing the chain effect of serious faults on the production line; When the device is in the medium-risk range, the operating parameters are dynamically adjusted to delay the deterioration of the device, significantly improving the device utilization rate.

[0144] Embodiment 2 is the second embodiment of the present invention. The difference from the previous embodiment is:

[0145] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the current technical solution, can be embodied in the form of a software product. The current computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs, etc., various media that can store program codes.

[0146] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a defined sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0147] More specific examples (non-exhaustive list) of computer-readable media include the following: electrical connection parts (electronic devices) having one or more wirings, portable computer disk cartridges (magnetic devices), random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memories), optical fiber devices, and portable compact disc read-only memories (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, then editing, interpreting, or processing it in other suitable ways when necessary, and then storing it in a computer memory.

[0148] Embodiment 3, an embodiment of the present invention, provides an intelligent textile production line control system, including a data acquisition module, a parameter adjustment module, an operation adjustment module, a fault prediction module, and an operation mode optimization module.

[0149] Data acquisition module: Real-time collect the operation data during the textile production process and perform data preprocessing.

[0150] Parameter adjustment module: Dynamically adjust the process parameters of the textile production line based on the pre-processed data.

[0151] Operation adjustment module: Real-time detect the quality status during the textile production process and adjust the production line operation based on the detection results.

[0152] Fault prediction module: Construct a multi-feature fusion fault prediction model to predict the operation status of the production line equipment.

[0153] Operation mode optimization module: Identify potential faults in advance according to the prediction results and adjust the operation mode of the production line.

[0154] Example 4, an embodiment of the present invention, provides an intelligent textile production line control method. To verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation / contrast experiments.

[0155] The purpose of this experiment is to verify the effectiveness of the textile production line control method, specifically including real-time data collection and preprocessing, dynamic adjustment of process parameters, real-time quality status detection, application of a multi-feature fusion fault prediction model, and adjustment of the production line operation mode according to the prediction results. The test environment is the production workshop of a textile factory, and the test objects are 6 main production equipment (Equipment A to Equipment F).

[0156] Test preparation:

[0157] Install data collection devices on the production line to record the operation data of the equipment in real time, including tension, weft density, equipment health score, and running speed.

[0158] Equip with an intelligent control system to achieve the functions of dynamically adjusting the production line process parameters and real-time quality status detection.

[0159] Introduce a multi-feature fusion fault prediction model in the test scenario to analyze the equipment operation status and generate a health score.

[0160] Preset three health score intervals:

[0161] High risk (score < 80);

[0162] Medium risk (80 ≤ score ≤ 90);

[0163] Low risk (score > 90).

[0164] Implementation process:

[0165] Real-time data collection and preprocessing: Record the tension, weft density, and running speed data of the equipment every 5 minutes, and preprocess the data through noise filtering and moving average algorithms to exclude the interference of sudden fluctuations on the analysis.

[0166] Dynamic adjustment of process parameters:

[0167] Analysis of the operating data of equipment A and equipment B found that the weft density deviated from the target value (25 threads / cm); by dynamically adjusting the weft input speed and loom tension, the process parameters were optimized in real time;

[0168] The operating speed of equipment C was relatively high, increasing the tension fluctuation. The production process was stabilized by reducing the operating speed.

[0169] Real-time quality inspection:

[0170] The surface defects of the textiles of equipment D were detected by a machine vision system, and uneven weft distribution was found;

[0171] According to the detection results, the tension value was adjusted and the spinning tasks were redistributed to other equipment.

[0172] Fault prediction and health score update:

[0173] A multi-feature fusion model was used to comprehensively analyze the tension, weft density, and equipment operating speed to generate a health score;

[0174] The health scores of equipment D and equipment F were lower than the medium-risk range, triggering an adjustment strategy.

[0175] Adjust the operating mode:

[0176] Stop the operation of equipment D, switch to the standby equipment, and at the same time reduce the overall speed of the production line to relieve the load;

[0177] For equipment F, dynamically adjust the process parameters, reduce the operating speed, and redistribute the load.

[0178] For specific experimental data, please refer to Table 1.

[0179] Table 1 Reference table of experimental data

[0180]

[0181] Before adjustment, the weft density of equipment A and equipment B was low, and it returned to the target value of 25 threads / cm after adjustment.

[0182] After the adjustment of equipment D, the tension decreased from 10.9 N to 10.2 N, and the weft density increased from 24.8 to 25.2, indicating that the adjustment significantly improved the production quality.

[0183] The scores of equipment D and equipment F were relatively low, 78 and 80 respectively, both triggering the adjustment strategy. After stopping the operation of equipment D, the load distribution of the production line was optimized, and the health scores of the remaining equipment did not show a significant decrease.

[0184] Data shows that through real-time detection and dynamic adjustment, the operating parameters of the equipment quickly return to the target range, significantly reducing quality fluctuations. The multi-feature fusion model successfully identifies potential faults, optimizes the operating mode in advance, and avoids the impact of equipment damage on the production line. The health scores are all higher than 80 after adjustment, verifying the effectiveness of the closed-loop feedback mechanism and ensuring the continuity and stability of production.

[0185] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. An intelligent textile production line control method, characterized in that: include: Collect operation data in the textile production process in real time and perform data preprocessing; Dynamically adjust the process parameters of the textile production line based on preprocessed data; Real-time detection of the quality status of textile production and adjustment of production line operation based on the detection results; Construct a multi-feature fusion fault prediction model to predict the operating status of production line equipment; Identify potential failures in advance based on prediction results and adjust the operation mode of the production line; The data preprocessing includes noise filtering and time domain feature extraction of the operating parameter data of the production equipment, wherein the vibration signal is processed by wavelet denoising method, and the motor speed and equipment load data are smoothed to eliminate instantaneous fluctuations; Perform time series smoothing and multi-dimensional data fusion on environmental parameter data, process temperature and humidity data using the sliding average method, and perform trend analysis on air velocity to evaluate the stability of the production environment; Feature extraction and anomaly detection are performed on product quality data. The textile tension and weft density data are judged by threshold to eliminate outliers, and the surface color uniformity data is optimized by histogram equalization method. The dynamic adjustment of the process parameters of the textile production line includes calculating the real-time change trend of the motor speed using a sliding average algorithm based on the operating parameter data of the production equipment, and dynamically adjusting the spinning speed and loom tension adjustment parameters in combination with the equipment load data; Based on environmental parameter data, the PID control method is used to adjust the operating status of the humidification device or air conditioning system, and the temperature and humidity of the production area are maintained within the target range by adjusting the output power in real time; Based on product quality data and the relationship between tension and weft density, an optimization model is constructed to dynamically adjust the weft density, and based on the color deviation analysis method, adjustment parameters for dye concentration and coating speed are generated to optimize the dyeing and finishing process. The real-time collection of operating data during the textile production process includes operating parameter data of production equipment, environmental parameter data and product quality data; The operating parameter data of the production equipment include motor speed, equipment load, bearing temperature and vibration signal; The environmental parameter data include temperature, humidity and air velocity in the production area; The product quality data include the tension, weft yarn density and surface color uniformity of the textile; The multi-feature fusion fault prediction model is constructed by: based on the vibration signal of the production line equipment, the collected vibration time series data is Perform fast Fourier transform and calculate the frequency domain energy eigenvalue. The calculation formula is: ； in, is the vibration signal characteristic value, and are the lower and upper limits of the frequency of the vibration signal, respectively. is the frequency domain eigenvalue of the vibration signal; is the attenuation factor of the high-frequency signal, which is used to reduce the impact of high-frequency noise on the results; is the frequency variable of the current integration; It is a small increment of frequency and is the basic unit of frequency domain integration, used to gradually accumulate signal characteristics within the frequency domain; Based on the bearing temperature data, the time series Perform second-order derivative calculation and integral trend analysis to extract temperature change characteristic values. The calculation formula is: ； in, is the temperature characteristic value, is the time series function of bearing temperature; The second derivative of temperature change is used to evaluate the acceleration trend of temperature; is the partial derivative; is the decay integral of the temperature trend, is the temperature fluctuation suppression factor; For time series The sinusoidal transformation reflects the periodic fluctuation characteristics of temperature changes; is time, which is the basic variable of time series data; Based on the motor current data, the collected current signal Perform time series analysis and extract the abnormal characteristics of the current signal through the autoregressive moving average model. The characteristic value calculation formula is: ； in, is the current characteristic value, For the The current sampling value, is the nonlinear amplification factor; Express The summation process of current sampling points; The prediction of the operating status of the production line equipment includes: synthesizing the characteristic values ​​to generate the equipment health score , and its calculation formula is: ； in, They are the normalized results of vibration, temperature and current characteristic values ​​respectively; is the total number of feature values, indicating the number of features considered in the calculation of the health score; For the A characteristic value represents the operating status of the device in a certain dimension; For the The reference value of the characteristic value is used to compare the deviation between the actual characteristic value and the target state; It is a characteristic difference inhibitor; The normalization formula is: ； in, is the normalized eigenvalue, is the original eigenvalue, is the maximum eigenvalue, is the minimum eigenvalue.

2. The intelligent textile production line control method according to claim 1, characterized in that: The adjusting the operation of the production line includes acquiring a surface image of the textile through a machine vision system, analyzing the type and location of defects using a convolutional neural network algorithm, and generating an adjustment signal to control the operation mode of the loom; The real-time tension data is obtained through the tension sensor, and the tension deviation value is calculated to generate a signal for adjusting the loom running speed and the weft yarn input speed; Real-time detection of weft density data, and optimization of the feeding speed and operation mode of the spinning equipment in combination with the dynamic adjustment model; Based on the surface color uniformity test results, the dynamic color difference analysis model is used to calculate the operating parameters of the dyeing and finishing equipment, and the dye concentration and coating speed are adjusted; Through the above detection and adjustment, a closed-loop feedback mechanism is formed to continuously update the equipment operating parameters to adapt to real-time changes in quality status.

3. The intelligent textile production line control method according to claim 2, characterized in that: The identification of potential failures in advance based on the prediction results includes: The potential failure risks are divided into multiple levels according to different value ranges, and corresponding production line adjustment strategies are adopted for different levels of risks; When health score In the high-risk range, When the load is too low, the equipment with the abnormality will be stopped and the standby equipment will be switched to ensure the continuity of the production line. The overall speed of the production line will be reduced and the spinning load will be redistributed to the low-load equipment. The adjusted load distribution formula is: ； in, For the The load distribution value of each device, is the total load, Weight the device health score; When health score In the medium risk range, When the equipment operating parameters are adjusted dynamically, including reducing the spinning speed, tension adjustment value and coating speed of the dyeing and finishing process, the dynamic adjustment formula is: ； in, are the adjusted operating parameters, is the current running parameter value, For reference health scores; monitor the equipment operating status in real time to detect whether there is a new downward trend in health scores; When health score In the low risk range, When the device is in the current operating mode, it will generate equipment health reports regularly to assess possible long-term risks. The adjustment process is achieved through a closed-loop feedback mechanism, that is, after adjusting the operating mode, the equipment health score is recalculated and the operating status of the production line is continuously optimized until the health score of all equipment is higher than the safety threshold.

4. An intelligent textile production line control system, used to implement the intelligent textile production line control method according to any one of claims 1 to 3, characterized in that: include: Data acquisition module: real-time collection of operating data in the textile production process and data preprocessing; Parameter adjustment module: dynamically adjust the process parameters of the textile production line based on the pre-processed data; Operation adjustment module: Real-time detection of the quality status of the textile production process and adjustment of the production line operation based on the detection results; Fault prediction module: builds a multi-feature fusion fault prediction model to predict the operating status of production line equipment; Operation mode optimization module: Identify potential failures in advance based on prediction results and adjust the operation mode of the production line.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the intelligent textile production line control method described in any one of claims 1 to 3 are implemented.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the intelligent textile production line control method described in any one of claims 1 to 3 are implemented.

Images