Disposable underpants manufacturing management system based on production equipment state analysis

Abstract

The invention discloses a disposable underpants manufacturing management system based on production equipment state analysis, relates to the technical field of manufacturing management, and solves the problem of insufficient real-time performance caused by network transmission delay due to the fact that a traditional equipment monitoring system generally adopts a centralized data processing architecture in the prior art. Yield, abnormal information, efficiency, shutdown duration, temperature, pressure, conveying speed and visual data are obtained in real time; the state analysis module analyzes the equipment state through machine learning and calculates the health degree; the dynamic control module adjusts operation parameters according to the health degree; and the quality association database is used for storing the mapping relation between the equipment state and the quality index and triggering quality tracing.

Description

Technical Field

[0001] The present invention relates to the technical field of manufacturing management, and in particular to a disposable underwear manufacturing management system based on production equipment status analysis. Background Art

[0002] Through the Internet of Things, big data analysis and artificial intelligence technologies, an intelligent management platform is established to monitor the operating status of production equipment in real time, provide fault warnings and optimize performance. The system focuses on the characteristics of disposable underwear production scenarios, such as high-load equipment operation, sensitive process parameters and strong quality traceability requirements, and realizes full-chain digital management and control from equipment health management to production efficiency improvement.

[0003] However, in existing technologies, traditional methods use single-point threshold alarms that cannot reflect the overall health status of the equipment, and existing systems usually shut down for maintenance directly after detecting an anomaly. At the same time, traditional equipment monitoring systems usually use a centralized data processing architecture, which has the problem of insufficient real-time performance due to network transmission delays. Conventional anomaly detection algorithms often ignore the periodic characteristics of operating conditions when processing periodic signals, resulting in an increased false alarm rate.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to solve the above-mentioned problems and to propose a disposable underwear manufacturing management system based on production equipment status analysis.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A disposable underwear manufacturing management system based on production equipment status analysis includes a management center connected to:

[0008] Data acquisition module, real-time acquisition of output, abnormal information, efficiency, downtime, temperature, pressure, conveying speed and visual data;

[0009] The status analysis module uses machine learning to analyze device status and calculate health;

[0010] Dynamic control module, adjusts operating parameters according to health;

[0011] The quality association database stores the mapping relationship between device status and quality indicators, triggering quality traceability.

[0012] As a preferred embodiment of the present invention, the data acquisition module is a multi-source sensor network deployed on the production line, and performs data acquisition through distributed sensors.

[0013] As a preferred embodiment of the present invention, the operating parameters of the production line equipment are transmitted to the edge computing node in real time through a distributed sensor network; the temperature sensor monitors the surface temperature distribution of the hot pressing plate, the pressure sensor records the closing pressure curve of the forming mold, and the image acquisition device captures the flatness characteristics of the material conveying; the collected data are normalized and then input into the convolutional neural network to extract the characteristic vector of the equipment operating status.

[0014] As a preferred embodiment of the present invention, the state analysis module includes a combined architecture of an edge computing unit, a cloud analysis unit and an anomaly detection algorithm; the edge computing unit refers to a data processing unit deployed on the edge of the network physically close to the production equipment; the cloud analysis unit refers to a time series data analysis module deployed on a remote server; the anomaly detection algorithm refers to a composite algorithm that integrates unsupervised learning and dynamic data analysis.

[0015] As a preferred embodiment of the present invention, the edge computing unit first performs noise reduction processing on the temperature, pressure, and vibration signals to form a lightweight feature vector and upload it to the cloud; after receiving the feature vector, the cloud analysis unit analyzes the equipment parameter change trend through a pre-trained LSTM network and calculates the probability distribution of the remaining service life; the anomaly detection algorithm simultaneously performs spectrum slicing on the periodic vibration signal, uses a sliding window to divide the working cycle, and constructs an isolated tree model in each window to identify the spectrum distortion characteristics; the collaborative work of the three components realizes the multi-dimensional evaluation of the equipment status.

[0016] As a preferred embodiment of the present invention, the operation process of the dynamic control module is as follows:

[0017] During the plastic sealing process, continuous temperature monitoring data triggers the PID controller to output a compensation signal, while the transmission mechanism executes a speed reduction command to form a dual adjustment mechanism to avoid temperature overshoot and insufficient sealing strength. When the health of the slitting equipment decreases, the control system automatically switches to the spare tool and generates a maintenance task to prevent tool failure from causing a decrease in material cutting accuracy. By collecting production data in real time, the raw material supply system dynamically matches the consumption rate to ensure that the production rhythm and material supply are synchronized to prevent raw material accumulation or shortage.

[0018] As a preferred embodiment of the present invention, the process of using the quality association database is as follows:

[0019] The correlation model between equipment vibration amplitude and joint fracture strength establishes a negative correlation between vibration amplitude and joint strength by real-time collection of vibration sensor data and combining it with destructive tensile test results. When the vibration amplitude exceeds the threshold, the equipment vibration reduction measures are automatically triggered; the quantitative relationship table between the temperature and humidity parameters of the sterilization process and the residual microorganisms is based on the results of microbial culture experiments under different temperature and humidity combinations to form a visual control standard for the sterilization effect. When an abnormal microbial residue is detected, the temperature and humidity control deviation of the sterilization equipment can be traced back; the regression analysis function of the degree of slitting tool wear on the material loss rate establishes a quadratic function relationship by regularly measuring the tool edge wear and combining it with the weighing data of raw material scraps. When the function prediction value exceeds the upper limit of material loss, a tool replacement instruction is generated.

[0020] As a preferred embodiment of the present invention, after the data acquisition module completes data acquisition, it performs a status analysis on the production equipment and uses the production output as a status evaluation parameter of the production equipment. When the status evaluation parameter is within the set threshold range, the normal state of the production equipment is set, and the Internet of Things module continues production according to the current production line operation setting; on the contrary, when the status evaluation parameter is not within the set threshold range, the abnormal state of the production equipment is set, the Internet of Things module adjusts the current production line operation setting, and marks the current output and production speed as abnormal information; after completing the abnormal information calibration, the production line capacity is tested, and the span of the production line capacity decrease during the production adjustment increase stage during the production stage is collected. , where the production capacity is evaluated by the ratio of the production speed of the current production cycle to the production speed of the set workload. If the span of the production line capacity drop does not exceed the set span threshold, it is inferred that the production line capacity is qualified; conversely, if the span of the production line capacity drop exceeds the set span threshold, it is inferred that the production line capacity is unqualified; the Internet of Things module analyzes the downtime distribution of the production line equipment. If the downtime distribution is irregular, there is a fluctuation in the equipment operation cycle. The Internet of Things module resets the equipment operation cycle of the production equipment. If the downtime distribution is regular, it is inferred that the abnormal information of the production line affects the operation of the production line. That is, after the Internet of Things module completes the production line adjustment, it predicts the abnormal information in advance and makes advance production line adjustments after the prediction.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] 1. This application provides timely warnings when equipment performance deteriorates slightly, preventing batch defects caused by equipment anomalies. Dynamic parameter adjustments maintain production process stability, reducing capacity losses caused by unplanned downtime. A correlation model between equipment status and quality data is established, significantly shortening the time required to analyze the root causes of quality anomalies. This achieves closed-loop management of the entire process, from equipment monitoring to quality control, improving the yield rate of disposable underwear production.

[0023] 2. This application achieves millisecond-level response speeds for production equipment status analysis, solving the latency issues of traditional cloud-based processing architectures. It accurately predicts the remaining useful life of equipment through deep learning models, providing data support for maintenance plan formulation. It employs a composite anomaly detection mechanism to effectively improve the accuracy of fault identification for periodic production equipment, providing reliable status monitoring assurance for manufacturing process quality control. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] To facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0025] Figure 1 It is a principle block diagram of the system of the present invention;

[0026] Figure 2 This is a flow chart of the method of embodiment 2 of the present invention. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0028] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0029] In existing technologies, the disconnect between equipment status monitoring and quality control is a common problem in the production of disposable underwear. Traditional methods rely primarily on manual inspections and offline quality spot checks, making it difficult to detect the correlation between equipment anomalies and product defects in real time. Existing monitoring systems typically use a single sensor for threshold alarms, are unable to conduct comprehensive analysis of multi-dimensional equipment data, and lack dynamic control mechanisms based on equipment health status. When the heat sealer's temperature drifts or the slitting tool wears, it is often not discovered until a batch of defective products has been produced, resulting in wasteful raw materials and rework costs.

[0030] To address these issues, R&D personnel discovered a nonlinear relationship between equipment status and product quality, necessitating the establishment of a multi-parameter integrated monitoring system. Long-term observation of key production processes such as sterilization, hot pressing, and slitting revealed that changes in the vibration spectrum precede visible product defects. This led to the idea of ​​building an equipment health assessment model, which uses machine learning algorithms to map sensor data to equipment status levels. They further realized the need for a dynamic compensation mechanism to automatically adjust process parameters at the initial stages of equipment performance degradation, rather than passively shutting down the machine for maintenance. This ultimately led to a closed-loop management approach, encompassing data collection, status analysis, and dynamic control.

[0031] Therefore, see Figure 1 As shown, this application proposes a disposable underwear manufacturing management system based on production equipment status analysis, including a management center connected to a data acquisition module, a status analysis module, a dynamic control module and a quality association database;

[0032] The data acquisition module acquires production, abnormal information, efficiency, downtime, temperature, pressure, conveying speed and visual data in real time;

[0033] The status analysis module uses machine learning to analyze device status and calculate health;

[0034] The dynamic control module adjusts operating parameters according to health;

[0035] The quality association database stores the mapping relationship between device status and quality indicators, triggering quality traceability.

[0036] Among them, the data acquisition module refers to the multi-source sensor network deployed on the production line. Specifically, it can be implemented by distributed sensor nodes connected by RS485 bus, which is used to synchronously collect equipment physical parameters and visual data.

[0037] The state analysis module refers to a computing unit capable of processing time series data. It can be implemented using an edge computing device equipped with the TensorFlow Lite framework, and identifies device abnormal patterns through a trained classification model.

[0038] The dynamic control module refers to a controller with real-time feedback adjustment function, which can be implemented by using a PLC and inverter linkage system to optimize process parameters based on the health assessment results.

[0039] A quality-related database refers to a data set that stores the correlation between equipment operating conditions and product quality inspection results. It can be implemented using a hybrid architecture of a time-series database and a relational database to establish a correspondence between equipment abnormal events and quality defects.

[0040] Specifically, the operating parameters of production line equipment are transmitted to edge computing nodes in real time through a distributed sensor network. Temperature sensors monitor the surface temperature distribution of the hot press plate, pressure sensors record the closing pressure curve of the forming mold, and image acquisition devices capture the flatness characteristics of material conveying. A distributed sensor network refers to a monitoring system composed of multiple types of sensors deployed at different physical locations on the production line. Specifically, it can be implemented using a combination of temperature sensors, pressure sensors, and vibration sensors, which can simultaneously collect multi-dimensional parameters of equipment operation.

[0041] These multi-dimensional data are normalized and then input into the convolutional neural network to extract the feature vector of the equipment operation status;

[0042] The status analysis module calculates the health index of the current device by comparing historical health status data, and triggers an early warning when the index falls below a preset threshold.

[0043] After receiving the health degradation signal, the dynamic control module automatically adjusts the equipment operating parameters, such as compensating for hot pressing temperature deviation or switching to a spare slitting tool.

[0044] The quality-related database continuously records changes in equipment status and product sampling results. When an abnormality in a specific equipment parameter is detected, the production batches within that period are automatically traced and quality reviews are performed.

[0045] Compared to existing technologies, traditional methods using single-point threshold alarms fail to reflect the overall health of the equipment. This solution achieves precise assessment of equipment status through multi-dimensional data fusion and machine learning models. Existing systems typically shut down for maintenance upon detecting an anomaly, while this solution maintains production continuity through dynamic parameter adjustments. Conventional quality management systems operate independently of the equipment monitoring system, while this solution establishes a mapping relationship between equipment status and quality indicators, enabling rapid quality traceability of abnormal operating conditions.

[0046] Through the above technical solution, this application can provide timely warnings when equipment performance deteriorates slightly, avoiding batch defects caused by equipment anomalies. Dynamic parameter adjustment maintains production process stability and reduces production capacity losses caused by unplanned downtime. A correlation model between equipment status and quality data is established, significantly shortening the time required to analyze the root causes of quality anomalies. This achieves closed-loop management of the entire process, from equipment monitoring to quality control, improving the yield rate of disposable underwear production.

[0047] Example 2

[0048] See also Figure 2 As shown, the present application further proposes a specific implementation of the state analysis module, including a combined architecture of an edge computing unit, a cloud analysis unit and an anomaly detection algorithm.

[0049] The edge computing unit is a data processing unit deployed at the edge of the network, physically close to production equipment. It can be implemented using an embedded industrial computer coupled with a data cleaning algorithm. It calculates the time-domain statistics and frequency-domain energy distribution of equipment vibration signals to perform feature compression. This unit reduces the amount of raw data transmitted and ensures real-time status analysis.

[0050] The cloud-based analysis unit is a time-series data analysis module deployed on a remote server. It utilizes a bidirectional LSTM neural network built on the TensorFlow framework to establish a degradation model by analyzing the time-series trends of equipment operating parameters. This unit leverages cloud computing resources to process complex models and achieve accurate lifespan predictions.

[0051] Anomaly detection algorithms are a composite of unsupervised learning and dynamic data analysis. Specifically, they use a sliding window mechanism to divide equipment operating cycles and apply the isolation forest algorithm within each window to detect outliers in spectral features. This algorithm adapts to the cyclical operating characteristics of production equipment and improves the accuracy of identifying anomaly patterns.

[0052] Specifically, the edge computing unit first performs noise reduction on the temperature, pressure, and vibration signals, extracts characteristic parameters such as peak factor and kurtosis coefficient, and forms a lightweight feature vector to upload to the cloud.

[0053] After receiving the feature vector, the cloud analysis unit analyzes the trend of device parameter changes through the pre-trained LSTM network and calculates the probability distribution of the remaining service life.

[0054] The anomaly detection algorithm synchronously slices the spectrum of the periodic vibration signal, divides the working cycle using a sliding window, and constructs an isolation tree model in each window to identify the spectrum distortion features.

[0055] The collaborative work of the three components enables a multi-dimensional assessment of equipment status.

[0056] Compared with existing technologies, traditional equipment monitoring systems typically use a centralized data processing architecture, which suffers from insufficient real-time performance due to network transmission delays. Existing life prediction methods often rely on empirical formulas, which make it difficult to capture the nonlinear characteristics of equipment degradation. Conventional anomaly detection algorithms often ignore the cyclical characteristics of operating conditions when processing periodic signals, resulting in increased false alarm rates. This solution uses a collaborative architecture of edge computing and the cloud to reduce response latency while ensuring computational accuracy. It uses an LSTM network to establish an equipment degradation model, which can effectively learn complex time series features. Combined with an anomaly detection method using a sliding window mechanism, it significantly improves recognition accuracy under periodic conditions.

[0057] Through the above technical solutions, this application achieves millisecond-level response speeds for production equipment status analysis, solving the latency issues of traditional cloud-based processing architectures. It accurately predicts the remaining useful life of equipment through deep learning models, providing data support for maintenance plan formulation. It also employs a composite anomaly detection mechanism to effectively improve the accuracy of fault identification for periodic production equipment, providing reliable status monitoring assurance for manufacturing process quality control.

[0058] Example 3

[0059] The present application further proposes that the dynamic control module performs at least one of the following operations: in the plastic sealing process, when the temperature sensor deviates from the set value by ±5% for three consecutive sampling cycles, the PID closed-loop adjustment is started and the conveyor belt speed is reduced by 10%-15%; when the health index of the elastic material slitting equipment is lower than 0.7, it automatically switches to the spare tool and generates a maintenance work order; according to the real-time production fluctuation data, the supply rate of non-woven fabric raw materials is dynamically adjusted to keep the inventory turnover rate in the range of 85%-92%.

[0060] Among them, PID closed-loop regulation refers to the real-time correction of temperature deviation through the proportional integral differential algorithm. It can be implemented using the PID module built into the industrial controller to quickly eliminate the impact of temperature fluctuations on sealing quality.

[0061] The health index refers to the equipment operating status score calculated through sensor data. It can be implemented using a weighted scoring model to quantitatively evaluate the degree of equipment performance degradation.

[0062] Inventory turnover rate refers to the ratio of raw material usage to total inventory per unit time. It can be achieved through a model matching material consumption rate with replenishment cycle, and is used to maintain production continuity and reduce warehousing costs.

[0063] Specifically, in the plastic sealing process, continuous temperature monitoring data triggers the PID controller to output a compensation signal, while the transmission mechanism executes a speed reduction instruction to form a dual adjustment mechanism to avoid temperature overshoot and insufficient sealing strength.

[0064] When the health of the slitting equipment declines, the control system automatically switches to a spare tool and generates a maintenance task to prevent tool failure from causing a decrease in material cutting accuracy. By collecting real-time production data, the raw material supply system dynamically matches the consumption rate to ensure that the production rhythm and material supply are synchronized, preventing raw material accumulation or shortages.

[0065] Compared to existing technologies, existing production systems typically rely on single-parameter threshold alarms, preventing the formation of multi-level coordinated control. This solution establishes a linkage mechanism between temperature regulation, equipment switching, and raw material supply, enabling simultaneous compensation actions when abnormal operating conditions are detected. This approach offers faster response times and more comprehensive parameter coverage than traditional single-point control methods.

[0066] Through the above technical solution, this application effectively addresses the issue of product quality fluctuations under abnormal equipment operating conditions. Closed-loop control maintains the stability of key process parameters, reducing product defects caused by equipment performance degradation. The automatic switching mechanism reduces the probability of unplanned downtime, and the dynamic raw material matching strategy avoids the risk of production interruptions, forming a comprehensive quality control system covering equipment maintenance and material management.

[0067] Example 4

[0068] This application further proposes a quality correlation database that includes a correlation model between equipment vibration amplitude and underwear seam breaking strength, a quantitative relationship table between sterilization process temperature and humidity parameters and microbial residues, and a regression analysis function of cutting tool wear degree on material loss rate.

[0069] Among them, the correlation model between the equipment vibration amplitude and the underwear seam breaking strength refers to a mathematical relationship model between the equipment vibration parameters and the product seam strength established through statistical analysis methods. Specifically, it can be implemented using the multivariate linear regression method. This model is used to guide the adjustment of the equipment vibration control parameters.

[0070] The quantitative relationship table between the temperature and humidity parameters of the sterilization process and the residual microorganisms refers to a mapping table between the temperature and humidity combinations and the microbial detection results established based on experimental data. Specifically, the orthogonal test method can be used to generate data samples. This table provides a standardized basis for setting the sterilization process parameters.

[0071] The regression analysis function of the slitting tool wear degree to the material loss rate refers to the fitting function of the tool wear monitoring data and the raw material loss data, which can be specifically implemented by a polynomial regression algorithm. This function is used to predict the tool replacement cycle.

[0072] Specifically, the correlation model between equipment vibration amplitude and seam fracture strength establishes a negative correlation between vibration amplitude and seam strength by collecting vibration sensor data in real time and combining it with destructive tensile test results. When the vibration amplitude exceeds the threshold, the equipment vibration reduction measures are automatically triggered. The quantitative relationship table between the temperature and humidity parameters of the sterilization process and the residual microorganisms is based on the results of microbial culture experiments under different temperature and humidity combinations to form a visual control standard for the sterilization effect. When an abnormal microbial residue is detected, the temperature and humidity control deviation of the sterilization equipment can be traced back. The regression analysis function of the degree of slitting tool wear on the material loss rate establishes a quadratic function relationship by regularly measuring the wear of the tool edge and combining it with the weighing data of raw material scraps. When the function prediction value exceeds the upper limit of material loss, a tool replacement instruction is generated.

[0073] Compared to existing technologies, traditional quality management systems only record equipment operating parameters and quality inspection results without establishing a mathematical model linking these parameters, requiring manual empirical analysis to trace defects. This solution automatically maps equipment status parameters to quality indicators by constructing quantitative models for three dimensions: equipment vibration and joint strength; temperature and humidity and microbial residue; and tool wear and material loss. This reduces defect root cause locating time by approximately 60%.

[0074] Through the above technical solution, this application solves the technical problem of unclear correlation between equipment status and product quality. It can quickly determine the risk of seam strength defects based on equipment vibration data, match microbial control standards in real time through temperature and humidity parameters, and accurately predict raw material loss trends based on the degree of tool wear, thereby realizing automated traceability and preventive control of quality problems.

[0075] Example 5

[0076] After the data acquisition module completes data acquisition, it analyzes the status of the production equipment and uses the production output as the status evaluation parameter of the production equipment. If the status evaluation parameter is within the set threshold range, the normal status of the production equipment is set, and the IoT module continues production according to the current production line operation settings. On the contrary, if the status evaluation parameter is not within the set threshold range, the abnormal status of the production equipment is set, and the IoT module adjusts the current production line operation settings and marks the current output and production speed as abnormal information. After the abnormal information calibration is completed, the production line capacity is tested, and the span of the production line capacity reduction during the production adjustment increase stage is collected, where the capacity is based on the current The ratio of the production speed of the production cycle to the production speed of the set workload is used as an evaluation parameter. If the span of the production line capacity reduction does not exceed the set span threshold, it is inferred that the production line capacity is qualified; conversely, if the span of the production line capacity reduction exceeds the set span threshold, it is inferred that the production line capacity is unqualified; the Internet of Things module analyzes the downtime distribution of the production line equipment. If the downtime distribution is irregular, there is a fluctuation in the equipment operation cycle. The Internet of Things module resets the equipment operation cycle of the production equipment. If the downtime distribution is regular, it is inferred that the abnormal information of the production line affects the operation of the production line. That is, after the Internet of Things module completes the production line adjustment, it predicts the abnormal information in advance and makes advance production line adjustments after the prediction.

[0077] When the present invention is in use, the data acquisition module acquires production, abnormal information, efficiency, downtime, temperature, pressure, conveying speed and visual data in real time;

[0078] The status analysis module uses machine learning to analyze device status and calculate health. The dynamic control module adjusts operating parameters based on health. The quality association database stores the mapping relationship between device status and quality indicators, triggering quality traceability.

[0079] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A disposable underwear manufacturing management system based on production equipment status analysis, characterized in that: Including the management center, the management center is connected to: Data acquisition module, real-time acquisition of output, abnormal information, efficiency, downtime, temperature, conveying speed and visual data; The status analysis module uses machine learning to analyze device status and calculate health; Dynamic control module, adjusts operating parameters according to health; The quality association database stores the mapping relationship between device status and quality indicators, triggering quality traceability.

2. A disposable underwear manufacturing management system based on production equipment status analysis according to claim 1, characterized in that: The data acquisition module is a multi-source sensor network deployed on the production line, which collects data through distributed sensors.

3. A disposable underwear manufacturing management system based on production equipment status analysis according to claim 1, characterized in that: The operating parameters of the production line equipment are transmitted to the edge computing node in real time through a distributed sensor network; the temperature sensor monitors the surface temperature distribution of the hot press plate, and the image acquisition device captures the flatness characteristics of the material conveying; the collected data are normalized and input into the convolutional neural network to extract the feature vector of the equipment operating status.

4. A disposable underwear manufacturing management system based on production equipment status analysis according to claim 3, characterized in that: The state analysis module includes a combined architecture of an edge computing unit, a cloud analysis unit, and an anomaly detection algorithm; the edge computing unit refers to a data processing unit deployed at the edge of the network physically close to the production equipment; the cloud analysis unit refers to a time series data analysis module deployed on a remote server; the anomaly detection algorithm refers to a composite algorithm that integrates unsupervised learning and dynamic data analysis.

5. A disposable underwear manufacturing management system based on production equipment status analysis according to claim 4, characterized in that: The edge computing unit first performs noise reduction on the temperature and vibration signals to form a lightweight feature vector and upload it to the cloud. After receiving the feature vector, the cloud analysis unit uses a pre-trained LSTM network to analyze the trend of device parameter changes and calculate the probability distribution of the remaining service life. The anomaly detection algorithm simultaneously performs spectrum slicing on the periodic vibration signal, divides the working cycle using a sliding window, and constructs an isolation tree model in each window to identify spectrum distortion features. The collaborative work of the three components enables a multi-dimensional assessment of the device status.

6. A disposable underwear manufacturing management system based on production equipment status analysis according to claim 5, characterized in that: The operation process of the dynamic control module is as follows: During the plastic sealing process, continuous temperature monitoring data triggers the PID controller to output a compensation signal, while the transmission mechanism executes a speed reduction command to form a dual adjustment mechanism to avoid temperature overshoot and insufficient sealing strength. By collecting production data in real time, the raw material supply system dynamically matches the consumption rate to ensure that the production rhythm and material supply are synchronized to prevent raw material accumulation or shortage.

7. A disposable underwear manufacturing management system based on production equipment status analysis according to claim 6, characterized in that: After the data acquisition module completes data acquisition, it analyzes the status of the production equipment and uses the production output as the status evaluation parameter of the production equipment. If the status evaluation parameter is within the set threshold range, the normal status of the production equipment is set, and the IoT module continues production according to the current production line operation settings. On the contrary, if the status evaluation parameter is not within the set threshold range, the abnormal status of the production equipment is set, and the IoT module adjusts the current production line operation settings and marks the current output and production speed as abnormal information. After the abnormal information calibration is completed, the production line capacity is tested, and the span of the production line capacity reduction during the production adjustment increase stage is collected, where the capacity is based on the current The ratio of the production speed of the production cycle to the production speed of the set workload is used as an evaluation parameter. If the span of the production line capacity reduction does not exceed the set span threshold, it is inferred that the production line capacity is qualified; conversely, if the span of the production line capacity reduction exceeds the set span threshold, it is inferred that the production line capacity is unqualified; the Internet of Things module analyzes the downtime distribution of the production line equipment. If the downtime distribution is irregular, there is a fluctuation in the equipment operation cycle. The Internet of Things module resets the equipment operation cycle of the production equipment. If the downtime distribution is regular, it is inferred that the abnormal information of the production line affects the operation of the production line. That is, after the Internet of Things module completes the production line adjustment, it predicts the abnormal information in advance and makes advance production line adjustments after the prediction.

Images