A whole-process quality monitoring and tracing system for production of rental bedding fabric

Abstract

This invention provides a full-process quality monitoring and traceability system for rental bedding fabric production, relating to the field of bedding fabric production technology. It includes a process parameter acquisition module, a digital twin monitoring module, an accelerated aging prediction module, a consortium blockchain traceability module, and a lifecycle management module. The process parameter acquisition module identifies and collects implicit key parameters affecting long-term washability throughout the entire rental bedding fabric production process. The digital twin monitoring module constructs a process digital twin based on the collected multi-source heterogeneous parameters. By locking down implicit parameters affecting long-term washability and combining the virtual-real fusion and real-time optimization characteristics of digital twin technology, this invention ensures the stability of the production process, significantly improves the fabric's washability and abrasion resistance, extends the average lifespan of the fabric, reduces quality defects during production, lowers production costs, and meets the core requirements of rental scenarios for high durability and low degradation of fabrics.

Description

Technical Field

[0001] This invention relates to the field of bedding fabric production technology, and in particular to a full-process quality monitoring and traceability system for the production of rental bedding fabrics. Background Technology

[0002] Rental bedding fabrics are widely used in hotels, hospitals, and guesthouses. Their core requirements are high durability and low degradation, needing to withstand frequent industrial washing while maintaining good physical properties and appearance, thus reducing replacement costs and wear rates during rental operations. Currently, quality monitoring in the production process of rental bedding fabrics mainly relies on conventional process parameter collection methods, focusing on monitoring basic parameters such as temperature and machine speed, and assessing fabric quality through manual or offline testing. Fabric lifespan prediction largely depends on traditional long-term washing tests. During rental operations, a passive maintenance and periodic replacement model is primarily employed. Conventional process parameter collection cannot capture the hidden key nodes and parameters affecting the long-term washability of fabrics, resulting in poor consistency in production processes and unstable washability and abrasion resistance of fabrics, making it difficult to meet the special needs of rental scenarios. Existing fabric life prediction is inefficient and lacks accuracy, unable to achieve quality grading, and difficult to match rental pricing requirements. Furthermore, it cannot quickly locate the root cause when quality abnormalities occur, resulting in low traceability efficiency. In addition, in existing technologies, centralized traceability systems have poor data security and insufficient traceability, failing to achieve effective retention and traceability of fabric lifecycle data, which easily leads to liability disputes. At the same time, the lack of lifecycle cost analysis and proactive life management mechanisms makes it difficult for rental companies to accurately identify high-loss items, resulting in unreasonable procurement and replacement strategies and high operating costs. Therefore, this invention proposes a full-process quality monitoring and traceability system for rental bedding fabric production to solve the problems existing in the prior art. Summary of the Invention

[0003] To address the aforementioned issues, this invention proposes a full-process quality monitoring and traceability system for the production of rental bedding fabrics. This system locks in the implicit parameters affecting long-term washability and combines the virtual-real fusion and real-time optimization characteristics of digital twin technology to ensure the stability of the production process, significantly improve the washability and abrasion resistance of the fabric, extend the average service life of the fabric, reduce quality defects in the production process, lower production costs, and meet the core requirements of rental scenarios for high durability and low degradation of fabrics.

[0004] To achieve the objectives of this invention, the following technical solution is provided: a full-process quality monitoring and traceability system for the production of rental bedding fabrics, comprising a process parameter acquisition module, a digital twin monitoring module, an accelerated aging prediction module, a consortium blockchain traceability module, and a lifecycle management module. The process parameter acquisition module is used to identify and collect implicit key parameters affecting long-term washability throughout the entire production process of rental bedding fabrics. The digital twin monitoring module constructs a process digital twin based on the collected multi-source heterogeneous parameters to achieve real-time monitoring and abnormal linkage adjustment of the production process. The accelerated aging prediction module is used to predict the number of times the fabric can be safely washed and to complete quality grading and coding. The consortium blockchain traceability module enables tamper-proof traceability of data throughout the entire lifecycle; the lifecycle management module is used to dynamically calculate the unit usage cost and output the optimal replacement threshold. All modules work together to monitor the quality, predict the lifespan, and trace the entire lifecycle of the rental bedding fabric production process.

[0005] Further improvements are made in the following aspects: The process parameter acquisition module includes an IoT sensor network, a near-infrared spectroscopy detection unit, and a parameter preprocessing unit; the IoT sensor network is deployed throughout the entire production process of spinning, weaving, dyeing and finishing, and is used to collect implicit key parameters in real time; the near-infrared spectroscopy detection unit is deployed in the dyeing and finishing section to detect the uniformity of resin crosslinking agent penetration online; the parameter preprocessing unit is used to denoise and normalize the collected multi-source heterogeneous parameters, and couple and correlate them through timestamps and fabric spatial coordinates to form a standardized parameter dataset.

[0006] Further improvements are made in the following aspects: the implicit key parameters include yarn twist unevenness and hairiness index in the spinning section, warp and weft weaving tension fluctuations and instantaneous frictional heat accumulation of the fabric surface during warp breaks in the weaving section, dispersion of the pretreatment alkali-oxygen concentration curve in the dyeing and finishing section, transverse / longitudinal temperature difference distribution in the setting process, and uniformity of resin crosslinking agent penetration, and real-time closed-loop deviation of the stretching overfeed in the finishing section; the parameter preprocessing unit uses a wavelet threshold denoising algorithm for noise reduction and a min-max normalization algorithm for normalization to ensure the consistency and validity of the parameter data.

[0007] Further improvements are made in the following aspects: The digital twin monitoring module includes a twin model construction unit, an anomaly warning unit, and an equipment linkage unit; the twin model construction unit, based on a standardized parameter dataset after coupling and association, and combined with the physical properties of fabric washability, constructs a process digital twin covering the entire process of spinning, weaving, dyeing and finishing, and realizing real-time mapping between the physical production process and the virtual model; the anomaly warning unit uses machine learning algorithms to construct an anomaly warning model, compares the virtual model parameters with the actual collected parameters in real time, and automatically issues an alarm when the parameters deviate from a preset threshold; after receiving the alarm signal, the equipment linkage unit automatically links the production equipment to adjust the parameters, and synchronously records the adjustment process, adjustment time, and parameter changes before and after the adjustment to ensure the consistency of the production process.

[0008] Further improvements are made in the following aspects: The accelerated aging prediction module includes an accelerated aging testing unit, a prediction model training unit, a lifespan prediction unit, and a grading and coding unit; the accelerated aging testing unit conducts a rapid washing fatigue test on each batch of first-inspection sample fabric using a combination of enhanced washing medium and mechanical action, simulating 200 standard washing cycles within 48 hours, and measuring the breaking strength retention rate, anti-pilling grade, dimensional change rate, and whiteness reduction value as indicators of tolerance attenuation; the prediction model training unit uses implicit key parameters throughout the process as independent variables and tolerance attenuation indicators as dependent variables, employing an attention mechanism temporal network to train the prediction model, outputting the predicted number of safe washes N for each roll of fabric, with the prediction formula as follows: , Wherein, X is the implicit key parameter vector of the whole process, X is the parameter subset of the i-th process step, W is the model weight parameter, W is the LSTM layer weight, α is the attention weight, and n is the total number of process steps; the graded coding unit classifies according to the predicted number of safe washings, with grade A being ≥300 times, grade B being 200-300 times, and grade C being <200 times, and the code is assigned to the fabric edge or electronic tag to realize quality graded rental pricing.

[0009] Further improvements include: the accelerated aging prediction module also includes a root cause localization unit, which automatically identifies the top 5 process steps and corresponding parameters that contribute most to the number of safe washes of the fabric through an attention weight heatmap, clarifying the impact weight of each process parameter on lifespan, and providing root cause pre-localization support for subsequent quality traceability and process optimization; the enhanced washing medium is a mixed solution containing 0.3%-0.5% surfactant and 0.1%-0.2% alkaline additives, and the mechanical action adopts a stirring speed of 1200r / min to simulate the high-intensity washing environment in the rental scenario.

[0010] Further improvements include: the consortium blockchain traceability module adopts a consortium blockchain architecture, connecting five nodes: fabric suppliers, manufacturers, washing centers, leasing operators, and recycling organizations. Each node has data read / write permissions and verification permissions; all data is stored using a hash-based on-chain + timestamp method, including process parameters for the entire fabric production process, quality inspection data, accelerated aging test data, washing counts, wear records, repair history, and recycling records, ensuring that the data is unforgeable, non-repudiable, and traceable; the consortium blockchain uses the PBFT consensus mechanism to improve the efficiency and security of data on-chain.

[0011] Further improvements include: the lifecycle management module includes a cost calculation unit and a replacement threshold output unit; the cost calculation unit has a built-in lifecycle cost analysis engine that dynamically calculates the unit usage cost C of each bedding item, and the calculation formula is: , Where C is the purchase cost, C is the washing cost, C is the loss cost, C is the labor cost, C is the equipment depreciation cost, and N is the predicted number of safe washing cycles. The replacement threshold output unit combines the AI ​​prediction model and actual operating data to output the optimal replacement threshold. When the actual number of fabric washing cycles is greater than or equal to the threshold and the defect rate is greater than the preset defect rate, a retirement recommendation is automatically issued.

[0012] Further improvements are made in that: the optimal replacement threshold is customized according to the operational needs of the rental operator, and the default threshold is ≥280 washing times and defect rate >1.2%; the defect rate is calculated as the ratio of the number of defects such as breakage, pilling, size deviation and whiteness reduction of a single bedding item to the total number of inspection items, and is updated in real time and synchronized to the blockchain.

[0013] Further improvements include a data visualization module and a historical data query module. The data visualization module displays production process parameters, digital twin mapping results, fabric life prediction data, traceability information, and unit usage costs in real time in the form of charts, allowing staff to monitor intuitively. The historical data query module supports querying full life cycle data by keywords such as fabric batch, production date, rental period, and number of washes.

[0014] The beneficial effects of this invention are as follows: 1. This invention locks in the implicit parameters that affect long-term washability, and combines the virtual-real fusion and real-time optimization characteristics of digital twin technology to ensure the stability of the production process, significantly improve the washability and abrasion resistance of the fabric, extend the average service life of the fabric, reduce quality defects in the production process, lower production costs, and meet the core requirements of rental scenarios for high resistance and low degradation of fabrics.

[0015] 2. This invention can simulate 200 standard washing cycles within 48 hours, quickly outputting the number of times each roll of fabric can be safely washed. It achieves quality-based rental pricing through graded coding, improving the rationality of rental pricing. At the same time, it achieves root cause pre-location through attention weight heatmap. When the fabric quality is abnormal, it can quickly locate the key process steps and parameters that affect the lifespan, providing strong support for process optimization and quality traceability, improving traceability efficiency and accuracy, and reducing quality disputes.

[0016] 3. This invention employs a five-node data hashing and blockchain-based time stamping system to ensure that the data throughout the entire lifecycle is unforgeable and non-repudiable, enabling full traceability of fabrics from production, washing, leasing to recycling. The lifecycle cost analysis engine accurately calculates the unit usage cost, and the optimal replacement threshold suggestion helps leasing companies accurately identify high-loss items, optimize procurement and replacement strategies, reduce linen loss rates, lower operating costs, and achieve a leap from "passive maintenance" to "proactive lifecycle management" in the operating model, thereby improving the operational efficiency and profitability of leasing companies. Attached Figure Description

[0017] Figure 1 This is the front view of the present invention. Detailed Implementation

[0018] To enhance understanding of the present invention, the present invention will be further described in detail below with reference to embodiments. These embodiments are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention. Example 1

[0019] according to Figure 1 As shown, this embodiment proposes a full-process quality monitoring and traceability system for rental bedding fabric production, suitable for small and medium-sized rental bedding fabric manufacturers. Process parameter acquisition module: Deploy a basic IoT sensor network, install twist meters and hairiness meters in the spinning section to collect yarn twist unevenness (control range 3.5%-4.5%) and hairiness index (H value control range 3.2-4.0); install tension sensors and temperature sensors in the weaving section to collect warp and weft weaving tension fluctuations (fluctuation range ≤ ±0.8N) and instantaneous frictional heat accumulation of warp stops on the fabric surface (control range ≤ 50℃); install concentration sensors, temperature sensors, and a basic near-infrared spectrometer in the dyeing and finishing section to collect the dispersion of the pretreatment alkali-oxygen concentration curve (control range ≤ 0.10), the transverse / longitudinal temperature difference distribution in the setting process (temperature difference ≤ 3℃), and the penetration uniformity of the resin crosslinking agent (penetration uniformity ≥ 90%); install displacement sensors in the finishing section to collect the real-time closed-loop deviation of the stretching overfeed (deviation ≤ ±0.5cm). The parameter preprocessing unit uses a wavelet threshold denoising algorithm to remove interference signals, and a min-max normalization algorithm to normalize the parameters to the [0,1] interval. The parameters are coupled and correlated with the timestamp and the fabric width direction partition coordinates (each 15cm is a partition) to form a standardized parameter dataset.

[0020] Digital Twin Monitoring Module: Based on a standardized parameter dataset and combined with the washability properties of cotton-polyester blended fabrics (commonly used in rental bedding), a simplified digital twin of the process is constructed to achieve real-time mapping of the entire process, including spinning, weaving, dyeing, and finishing. The anomaly warning model is trained using a logistic regression algorithm, with preset normal thresholds for each parameter. It compares the virtual model parameters with the actual collected parameters in real time. When a parameter deviates from the threshold by ≥15%, an audible and visual warning is automatically issued. After receiving the warning signal, the equipment linkage unit issues an adjustment prompt, allowing staff to manually link the corresponding production equipment to adjust the parameters. The adjustment process, adjustment time, and parameter changes before and after the adjustment are recorded simultaneously to ensure process consistency.

[0021] Accelerated Aging Prediction Module: For each batch of first-inspection sample fabric (3 pieces selected from each batch, each 30cm×30cm), a rapid washing fatigue test is conducted using an enhanced washing medium (containing 0.4% surfactant and 0.15% alkaline additive) and a stirring speed of 1000r / min, simulating 200 standard washing cycles within 48 hours. The breaking strength retention rate (≥80%), anti-pilling grade (≥3.5), dimensional change rate (≤3%), and whiteness reduction value (≤6%) are measured as indicators of tolerance attenuation. Using implicit key parameters throughout the process as independent variables and tolerance attenuation indicators as dependent variables, an attention-based LSTM model is used to train the prediction model, outputting the predicted number of safe washes for each roll of fabric. According to the standards of Grade A ≥300 times, Grade B 200-300 times, and Grade C <200 times, inkjet coding is applied to the fabric edge. Through attention weight heatmap, the 5 process steps and parameters that contribute the most to the lifespan are identified, providing root cause pre-location for traceability.

[0022] Consortium blockchain traceability module: Adopting a basic consortium blockchain architecture, it connects four core nodes: fabric suppliers, manufacturers, washing centers, and leasing operators. It uses the PBFT consensus mechanism, with consensus latency controlled within 500ms. All data is encrypted using the SHA-256 hash algorithm and stored with timestamps. It focuses on storing process parameters, quality inspection data, accelerated aging test data, and washing counts for the entire fabric production process, ensuring that the data is tamper-proof and non-repudiable. It supports each node to query relevant data according to its permissions, realizing traceability of the core life cycle of the fabric.

[0023] Lifecycle Management Module: The cost calculation unit dynamically calculates the unit usage cost of each bedding item, where C (purchase cost) is 75 yuan / item, C (washing cost) is 2.2 yuan / wash, C (damage cost) is 4.5 yuan / item, C (labor cost) is 2.8 yuan / item, and C (equipment depreciation cost) is 1.8 yuan / item. If the predicted safe washing count N=280 times, then the unit usage cost C=(75+2.2×280+4.5+2.8+1.8) / 280≈3.05 yuan / wash. The replacement threshold output unit has a default threshold of ≥280 washes and a defect rate >1.5%. When the actual number of washes reaches 280 and the defect rate is 1.6%, a retirement recommendation is automatically issued.

[0024] This embodiment is suitable for the needs of small and medium-sized enterprises, with controllable costs. The accuracy of process parameter collection meets the requirements of basic leasing scenarios. Life prediction is fast and efficient, and traceability data is safe and reliable. It can realize proactive life management and significantly reduce operating costs compared with existing technologies. Example 2

[0025] according to Figure 1 As shown, this embodiment proposes a full-process quality monitoring and traceability system for rental bedding fabric production, suitable for medium and large-sized rental bedding fabric manufacturers and large-scale rental operators: Process parameter acquisition module: Deploy a high-precision IoT sensor network. In the spinning section, install high-precision twist meters and hairiness meters to collect yarn twist unevenness (control range 3.2%-4.5%) and hairiness index (H value control range 3.0-4.0). In the weaving section, install high-precision tension sensors and temperature sensors to collect warp and weft weaving tension fluctuations (fluctuation range ≤ ±0.5N) and instantaneous frictional heat accumulation of warp stops on the fabric surface (control range ≤ 45℃). In the dyeing and finishing section, install high-precision concentration sensors, temperature sensors, and a high-precision near-infrared spectrometer to collect the dispersion of the pretreatment alkali-oxygen concentration curve (control range ≤ 0.08), the transverse / longitudinal temperature difference distribution in the setting process (temperature difference ≤ 2℃), and the penetration uniformity of the resin crosslinking agent (penetration uniformity ≥ 92%). In the finishing section, install high-precision displacement sensors to collect the real-time closed-loop deviation of the stretching overfeed (deviation ≤ ±0.3cm). The parameter preprocessing unit uses a wavelet threshold denoising algorithm to remove interference signals, and a min-max normalization algorithm to normalize the parameters to the [0,1] interval. The parameters are coupled and correlated with the timestamp and the fabric width direction partition coordinates (each 10cm is a partition) to form a standardized parameter dataset.

[0026] Digital Twin Monitoring Module: Based on a standardized parameter dataset and combined with the washability properties of cotton-polyester blended fabrics (commonly used in rental bedding), a complete digital twin of the process is constructed to achieve real-time mapping of the entire process, including spinning, weaving, dyeing and finishing, and supporting process parameter simulation and optimization. The anomaly warning model is trained using a random forest algorithm, with preset normal thresholds for each parameter. It compares the virtual model parameters with the actual collected parameters in real time. When a parameter deviates from the threshold by ≥10%, an audible and visual warning is automatically issued, and warning details are pushed to the staff's terminal. After receiving the warning signal, the equipment linkage unit automatically links the corresponding production equipment to adjust the parameters. For example, when the alkali-oxygen concentration dispersion in the dyeing and finishing section exceeds the standard, the speed of the liquid supply pump is automatically adjusted, and the adjustment time, parameters before and after the adjustment, and the personnel involved in the adjustment are recorded simultaneously to ensure process consistency.

[0027] Accelerated Aging Prediction Module: For each batch of first-inspection sample fabric (5 pieces randomly selected from each batch, each 30cm x 30cm), a rapid washing fatigue test is conducted using an enhanced washing medium (containing 0.4% surfactant and 0.15% alkaline additive) and a stirring speed of 1200 rpm, simulating 200 standard washing cycles within 48 hours. The breaking strength retention rate (≥85%), pilling resistance grade (≥4), dimensional change rate (≤2%), and whiteness reduction value (≤5%) are measured as indicators of tolerance degradation. Using implicit key parameters throughout the process as independent variables and tolerance degradation indicators as dependent variables, an attention-based method is employed. An LSTM model is used to train a prediction model, which outputs the predicted number of safe washes for each roll of fabric. According to the standards of Grade A ≥ 300 times, Grade B 200-300 times, and Grade C < 200 times, laser coding is applied to the fabric edge, and the grading information and full-process parameters are stored in the electronic tag. Through attention weight heatmap, the five process steps and parameters that contribute the most to the lifespan are identified (in order: transverse temperature difference in the setting process, uniformity of resin crosslinking agent penetration, yarn twist unevenness, stretching overfeed deviation, and warp and weft weaving tension fluctuation), providing root cause pre-location for traceability.

[0028] The consortium blockchain traceability module adopts an advanced consortium blockchain architecture, connecting five nodes: fabric suppliers (providing fabric raw material parameters), manufacturers (providing production process and quality data), washing centers (providing washing times and parameters), leasing operators (providing leasing records and wear records), and recycling organizations (providing recycling records). It employs the PBFT consensus mechanism, with consensus latency controlled within 450ms. All data is encrypted using the SHA-256 hash algorithm and stored with timestamps, including process parameters throughout the entire fabric production process, quality inspection data, accelerated aging test data, washing times, wear records, maintenance history, and recycling records, ensuring data is tamper-proof and non-repudiable. Each node can query relevant data according to its permissions, enabling full lifecycle traceability of the fabric, and data can also be exported for compliance checks.

[0029] Lifecycle Management Module: The cost calculation unit dynamically calculates the unit usage cost of each bedding item, where C (purchase cost) is 80 yuan / item, C (washing cost) is 2 yuan / wash, C (damage cost) is 5 yuan / item, C (labor cost) is 3 yuan / item, and C (equipment depreciation cost) is 2 yuan / item. If the predicted number of safe washes is N=300, then the unit usage cost C=(80+2×300+5+3+2) / 300≈2.93 yuan / wash. The replacement threshold output unit has a default threshold of ≥280 washes and a defect rate >1.2%. When the actual number of washes reaches 280 and the defect rate is 1.5%, a retirement suggestion is automatically issued, and the leasing operator is coordinated to adjust the operation plan.

[0030] The data visualization module and the historical data query module: The data visualization module uses line charts to show the trend of process parameter changes, bar charts to show the distribution of fabric quality grading, and tables to show life cycle cost data, and supports the annotation of abnormal data; The historical data query module supports keyword queries by batch, production date, number of washes, rental period, etc., with a query response time of 0.8s, and supports historical data comparison and analysis, providing data support for process optimization and operational decision-making.

[0031] This embodiment features precise acquisition of process parameters, good process consistency, and stable fabric washability; high accuracy and speed in lifespan prediction, reasonable quality grading, and accurate and comprehensive traceability; secure and reliable traceability data; and intelligent operation management, which can significantly reduce operating costs and enhance enterprise competitiveness.

[0032] Comparative example: The existing methods for monitoring and managing the production quality of rental bedding fabrics are as follows: During production, only conventional process parameters such as temperature and machine speed are collected, and fabric quality is monitored through manual, timed inspections; fabric lifespan prediction uses traditional washing tests, requiring 15 days to perform standard washing cycles on sample fabrics, accumulating 200 washes, which cannot achieve rapid batch prediction and does not include quality grading; traceability uses a centralized database, recording only fabric production batches and basic testing data, excluding data such as washing frequency and wear records, making the data susceptible to tampering and loss; during operation, a periodic replacement model is adopted, with all bedding replaced every 6 months, without calculating lifecycle costs or proactive lifespan management.

[0033] The shortcomings of this embodiment are: inaccurate process parameter collection, inability to capture hidden key parameters, unstable fabric washability, and large fluctuations in lifespan; long lifespan prediction time-consuming, unable to meet the needs of mass production, lack of quality grading, and unreasonable rental pricing; incomplete and unreliable traceability data, making full traceability impossible; unreasonable operation strategy, high linen loss rate, high replacement cost, and inability to achieve proactive lifespan management.

[0034] Validation data: To verify the effectiveness of this invention, cotton-polyester blend rental bedding fabrics of the same batch and specifications (specifications: 40s×40s, 133×72) were selected. Production, monitoring, and operation were conducted using comparative examples, Example 1, and Example 2, respectively. Performance indicators were compared and tested, and the verification data are shown in the table below: Average number of safe washes 220 times 275 times 290 times An increase of 31.8% Fabric quality defect rate 8.5% 6.8% 6.2% Decrease of 27.1% Lifetime prediction time 15 days 48 hours 48 hours Shortened by 94.4% Linen loss rate 10.2% 6.5% 5.9% Reduced by 42.2% Unit usage cost 3.85 yuan / time 3.05 yuan / time 2.93 yuan / time Decreased by 23.9% Source tracing response time 3.5s 1.0s 0.8s Shortened by 77.1% Stability of process parameters (deviation rate) 12.3% 7.5% 4.8% Reduced by 61.0% As can be seen from the above verification data, compared with the prior art, the present invention has significantly improved the fabric washability, quality stability, life prediction efficiency, traceability speed, and operating cost control. It can effectively solve the defects of the prior art and meet the rental scenario's demand for fabrics with "high resistance and low degradation".

[0035] This end-to-end quality monitoring and traceability system for rental bedding fabric production locks in hidden parameters affecting long-term washability. By combining the virtual-real fusion and real-time optimization capabilities of digital twin technology, it ensures the stability of the production process, significantly improves the fabric's washability and abrasion resistance, extends its average lifespan, reduces quality defects during production, lowers production costs, and meets the core requirements of rental scenarios for high-resilience and low-deterioration fabrics. This invention can simulate 200 standard washing cycles within 48 hours, quickly outputting the safe washing count for each roll of fabric. Graded coding enables quality-based rental pricing, improving the rationality of rental pricing. Simultaneously, attention weight heatmaps enable root cause pre-location. When fabric quality abnormalities occur, key process steps and parameters affecting lifespan can be quickly identified, providing strong support for process optimization and quality traceability, improving traceability efficiency and accuracy, and reducing quality disputes. This invention employs a five-node data hashing and blockchain-based time stamping system to ensure that the data throughout the entire lifecycle is unforgeable and non-repudiable, enabling full traceability of fabrics from production, washing, leasing to recycling. The lifecycle cost analysis engine accurately calculates the unit usage cost, and the optimal replacement threshold suggestion helps leasing companies accurately identify high-loss items, optimize procurement and replacement strategies, reduce linen loss rates, lower operating costs, and achieve a leap from "passive maintenance" to "proactive lifecycle management" in the operating model, thereby improving the operational efficiency and profitability of leasing companies.

[0036] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A full-process quality monitoring and traceability system for the production of rental bedding fabrics, comprising a process parameter acquisition module, a digital twin monitoring module, an accelerated aging prediction module, a consortium blockchain traceability module, and a lifecycle management module, characterized in that: The process parameter acquisition module is used to identify and collect hidden key parameters that affect long-term washability in the entire production process of rental bedding fabrics. The digital twin monitoring module constructs a process digital twin based on the collected multi-source heterogeneous parameters to realize real-time monitoring and abnormal linkage adjustment of the production process. The accelerated aging prediction module is used to predict the number of times the fabric can be safely washed and complete the quality grading and coding. The consortium blockchain traceability module enables tamper-proof traceability of data throughout the entire lifecycle; the lifecycle management module is used to dynamically calculate the unit usage cost and output the optimal replacement threshold. All modules work together to monitor the quality, predict the lifespan, and trace the entire lifecycle of the rental bedding fabric production process.

2. The whole-process quality monitoring and traceability system for rental bedding fabric production according to claim 1, characterized in that: The process parameter acquisition module includes an IoT sensor network, a near-infrared spectroscopy detection unit, and a parameter preprocessing unit. The IoT sensor network is deployed throughout the entire production process, including spinning, weaving, dyeing and finishing, to collect implicit key parameters in real time. The near-infrared spectroscopy detection unit is deployed in the dyeing and finishing section to detect the uniformity of resin crosslinking agent penetration online. The parameter preprocessing unit is used to denoise and normalize the collected multi-source heterogeneous parameters, and couple and correlate them with timestamps and fabric spatial coordinates to form a standardized parameter dataset.

3. The whole-process quality monitoring and traceability system for rental bedding fabric production according to claim 2, characterized in that: The implicit key parameters include yarn twist unevenness and hairiness index in the spinning section, warp and weft weaving tension fluctuations and instantaneous frictional heat accumulation of the fabric surface during warp breaks in the weaving section, dispersion of the pretreatment alkali-oxygen concentration curve in the dyeing and finishing section, transverse / longitudinal temperature difference distribution in the setting process, and uniformity of resin crosslinking agent penetration, and real-time closed-loop deviation of the stretching overfeed in the finishing section. The parameter preprocessing unit uses a wavelet threshold denoising algorithm for noise reduction and a min-max normalization algorithm for normalization to ensure the consistency and validity of the parameter data.

4. The whole-process quality monitoring and traceability system for the production of rental bedding fabrics according to claim 1, characterized in that: The digital twin monitoring module includes a twin model construction unit, an anomaly early warning unit, and an equipment linkage unit. The twin model construction unit, based on a standardized parameter dataset after coupling and association, and combined with the fabric's washability and physical properties, constructs a digital twin of the entire process, covering spinning, weaving, dyeing, and finishing, achieving real-time mapping between the physical production process and the virtual model. The anomaly early warning unit uses machine learning algorithms to construct an anomaly early warning model, comparing the virtual model parameters with the actual collected parameters in real time, and automatically issuing an early warning when the parameters deviate from a preset threshold. Upon receiving the early warning signal, the equipment linkage unit automatically links the production equipment to adjust parameters, and simultaneously records the adjustment process, adjustment time, and parameter changes before and after the adjustment, ensuring consistency in the production process.

5. The whole-process quality monitoring and traceability system for the production of rental bedding fabrics according to claim 1, characterized in that: The accelerated aging prediction module includes an accelerated aging testing unit, a prediction model training unit, a lifespan prediction unit, and a grading and coding unit. The accelerated aging testing unit conducts a rapid washing fatigue test on each batch of first-inspection sample fabric using a combination of enhanced washing media and mechanical action, simulating 200 standard washing cycles within 48 hours. It measures the breaking strength retention rate, anti-pilling grade, dimensional change rate, and whiteness reduction value as indicators of tolerance degradation. The prediction model training unit uses implicit key parameters throughout the process as independent variables and tolerance degradation indicators as dependent variables, employing an attention mechanism temporal network to train the prediction model. It outputs the predicted number of safe washes N for each roll of fabric, using the following prediction formula: , Wherein, X is the implicit key parameter vector of the whole process, X is the parameter subset of the i-th process step, W is the model weight parameter, W is the LSTM layer weight, α is the attention weight, and n is the total number of process steps; the graded coding unit classifies according to the predicted number of safe washings, with grade A being ≥300 times, grade B being 200-300 times, and grade C being <200 times, and the code is assigned to the fabric edge or electronic tag to realize quality graded rental pricing.

6. The whole-process quality monitoring and traceability system for the production of rental bedding fabrics according to claim 5, characterized in that: The accelerated aging prediction module also includes a root cause localization unit, which automatically identifies the top 5 process steps and corresponding parameters that contribute the most to the number of safe washes of the fabric through an attention weight heatmap, clarifies the influence weight of each process parameter on the lifespan, and provides root cause pre-localization support for subsequent quality traceability and process optimization. The enhanced washing medium is a mixed solution containing 0.3%-0.5% surfactant and 0.1%-0.2% alkaline additives. The mechanical action is achieved by stirring at a speed of 1200 r / min to simulate the high-intensity washing environment in a rental scenario.

7. The whole-process quality monitoring and traceability system for the production of rental bedding fabrics according to claim 1, characterized in that: The consortium blockchain traceability module adopts a consortium blockchain architecture, connecting five nodes: fabric suppliers, manufacturers, washing centers, leasing operators, and recycling organizations. Each node has data read / write permissions and verification permissions. All data is stored using a hash-based on-chain and timestamp-based storage method, including process parameters for the entire fabric production process, quality inspection data, accelerated aging test data, number of washes, wear records, repair history, and recycling records, ensuring that the data is unforgeable, non-repudiable, and traceable. The consortium blockchain uses the PBFT consensus mechanism to improve the efficiency and security of data on-chain storage.

8. The whole-process quality monitoring and traceability system for the production of rental bedding fabrics according to claim 1, characterized in that: The lifecycle management module includes a cost calculation unit and a replacement threshold output unit; the cost calculation unit has a built-in lifecycle cost analysis engine that dynamically calculates the unit usage cost C of each bedding item, and the calculation formula is as follows: , Where C is the purchase cost, C is the washing cost, C is the loss cost, C is the labor cost, C is the equipment depreciation cost, and N is the predicted number of safe washing cycles. The replacement threshold output unit combines the AI ​​prediction model and actual operating data to output the optimal replacement threshold. When the actual number of fabric washing cycles is greater than or equal to the threshold and the defect rate is greater than the preset defect rate, a retirement recommendation is automatically issued.

9. A full-process quality monitoring and traceability system for the production of rental bedding fabrics according to claim 8, characterized in that: The optimal replacement threshold is customized according to the operational needs of the rental operator. The default threshold is ≥280 washes and a defect rate >1.2%. The defect rate is calculated as the ratio of the number of defects such as breakage, pilling, size deviation, and whiteness reduction of a single bedding item to the total number of inspection items. It is updated in real time and synchronized to the blockchain.

10. A full-process quality monitoring and traceability system for the production of rental bedding fabrics according to any one of claims 1-9, characterized in that: It also includes a data visualization module and a historical data query module; the data visualization module displays production process parameters, digital twin mapping results, fabric life prediction data, traceability information and unit usage cost in real time in the form of charts, for staff to monitor intuitively; the historical data query module supports querying full life cycle data by keywords such as fabric batch, production date, rental period, and number of washes.

Images