A Textile Processing Traceability System Based on Identifier Resolution

By generating industrial internet identifiers and material coordinates for materials during textile processing, and combining them with equipment encoder pulses and quality inspection data, a material transformation diagram is constructed. This solves the problem of inaccurate material traceability in textile processing and enables rapid defect location and efficient quality problem analysis.

CN123117936APending Publication Date: 2026-11-17JIANGSU IND INTERNET DEV RES CENT
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611628936.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing textile processing traceability systems struggle to accurately describe the correspondence between materials at different processing stages. In particular, when local defects appear in finished products, it is difficult to determine the location and contribution of upstream materials in the defective area, leading to inaccurate quality analysis and defect localization.

Method used

By generating industrial internet identifiers for raw materials, yarns, fabric rolls, and cut pieces, and establishing material coordinates, combined with equipment encoder pulses and quality inspection data, a material transformation diagram is constructed, and coordinate interval mapping and verification are performed to achieve reverse positioning of defect coordinate intervals.

Benefits of technology

It improves the continuity and accuracy of material traceability, ensures the accurate correlation between processing data and specific material locations, and enhances the reliability of correlations and the efficiency of defect tracing in complex processing processes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123117936A_ABST
    Figure CN123117936A_ABST
Patent Text Reader

Abstract

This invention discloses a textile processing traceability system based on identifier resolution, belonging to the field of information management and quality traceability technology for textile processing. The system includes an identifier coordinate generation module, a process evidence collection module, a material transformation construction module, an identifier resolution verification module, and a defect reverse location module. By establishing industrial internet identifiers and material coordinates for raw materials, yarns, fabric rolls, and cut pieces, and combining this with equipment encoder pulses, process sensor data, and quality inspection data to form process evidence fragments, and establishing transformation records between material coordinate intervals at different processing stages, a material transformation diagram is formed after verification of continuity, contribution ratio conservation, and time overlap. Thus, based on the finished product defect coordinate interval, the corresponding upstream material coordinate interval, processing time window, and process evidence fragment are determined in reverse, achieving precise traceability and defect location in the textile processing process.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing and processing information management technology for textiles, and in particular to a textile processing traceability system based on identifier resolution. Background Technology

[0002] With the development of digitalization in textile production processes and industrial internet technology, textile enterprises are gradually adopting QR codes, RFID, barcodes, or industrial internet identifiers to identify raw materials, yarns, fabric rolls, cut pieces, and finished products. They also record equipment information, process parameters, quality inspection results, and material flow information at each processing station through production management systems. In processes such as blending, weaving, dyeing and finishing, slitting, and cutting, existing traceability systems typically use batch numbers, material numbers, or product identifiers as the basis for data association. Identification information is collected when materials enter or leave the processing station, and a production history is established according to production orders, processing batches, or upstream and downstream material correspondences. When quality traceability is required, the corresponding raw material batch, processing procedure, production equipment, processing time, and quality inspection records can be queried based on the finished product identifier, thus forming a textile processing information chain with products or batches as the basic units.

[0003] However, textiles undergo continuous material morphology changes during actual processing, including mixing, drafting, weaving, winding, slitting, and cutting. One upstream material may form multiple downstream materials, and multiple upstream materials may collectively form the same processed product. This makes it difficult to accurately describe the correspondence between local areas of materials at different processing stages by simply relying on upstream and downstream correlations based on batch numbers or identifiers. Especially when the finished product only exhibits localized color differences, yarn breaks, weaving defects, or other quality defects, existing technologies can usually only trace the defect to a specific roll, production batch, or processing step. It is difficult to further determine the actual location of the upstream material corresponding to the defective area, the contribution of different input materials to that area, and the corresponding process and quality inspection data during the defect formation period. At the same time, continuous process data generated by processing equipment is usually stored according to time, while material quality defects appear according to material location. The two types of data lack a stable positional correspondence. In cases of changes in equipment speed, material slitting, and combined processing of multiple materials, upstream and downstream traceability relationships are prone to shifts, breaks, or errors, affecting the accuracy of textile processing quality analysis and defect cause localization.

[0004] Therefore, it is necessary to propose a textile processing traceability technology that can improve the traceability accuracy and defect location accuracy of textile processing. Summary of the Invention

[0005] This application provides a textile processing traceability system based on identifier resolution to improve the accuracy of textile defect traceability.

[0006] This application provides a textile processing traceability system based on identifier resolution, comprising: an identifier coordinate generation module, used to generate industrial internet identifiers for raw materials, yarns, fabric rolls, and cut pieces entering the textile processing chain, and to establish material coordinates associated with the corresponding material location for each industrial internet identifier; a process evidence collection module, used to read input and output industrial internet identifiers at the processing station, simultaneously collect equipment encoder pulses, process sensor data, and quality inspection data, and divide the process sensor data and quality inspection data into process evidence fragments corresponding to the material coordinate intervals based on the equipment encoder pulses; and a material transformation construction module, used to, during blending, weaving, slitting, or cutting processes, determine the material transformation based on the input material feed amount, output material formation amount, and equipment encoder pulses. The coder pulse establishes a transformation record between the input and output material coordinate intervals, including the coordinate interval mapping relationship and the input material contribution ratio. The identifier parsing and verification module connects the transformation records corresponding to continuous processing steps according to the parsing relationship between industrial internet identifiers, forming a material transformation diagram, and verifies the material transformation diagram based on the continuity of material coordinates, the conservation relationship of the input material contribution ratio, and the overlap relationship of processing time. The defect reverse location module obtains the finished product industrial internet identifier and defect coordinate interval, performs coordinate interval mapping in reverse along the verified material transformation diagram, determines the upstream material coordinate interval and processing time window corresponding to the defect coordinate interval, and extracts the process evidence fragments corresponding to the processing time window as the traceability result.

[0007] This application has the following beneficial technical effects: (1) By associating industrial internet identifiers with material coordinates, the unified positioning of raw materials, yarns, fabric rolls, and cut pieces at different processing stages can be achieved, improving the continuity and accuracy of material traceability. (2) By using equipment encoder pulses to correspond process sensing data, quality inspection data, and material coordinate intervals, the processing data can be accurately associated with the specific material location. (3) By establishing a material transformation diagram through coordinate interval mapping relationships and input material contribution ratios, and performing continuity, conservation relationships, and time overlap checks, the reliability of complex processing process correlation relationships can be improved. (4) By reverse mapping the defect coordinate interval to the upstream material coordinate interval and processing time window, the material source and process evidence corresponding to the defect can be quickly located, improving the efficiency of quality problem analysis and defect traceability. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of a textile processing traceability system based on identifier resolution provided in the first embodiment of this application.

[0009] Figure 2 This is a schematic diagram comparing the defect tracing and location effects of a textile processing traceability system based on identifier resolution, which is involved in the first embodiment of this application. Detailed Implementation

[0010] Many specific details are set forth in the following description to provide a full understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this application; therefore, this application is not limited to the specific embodiments disclosed below.

[0011] The first embodiment of this application provides a textile processing traceability system based on identifier resolution. Please refer to... Figure 1 This figure is a schematic diagram of the first embodiment of this application. The following is in conjunction with... Figure 1 The first embodiment of this application provides a textile processing traceability system based on identifier resolution.

[0012] The textile processing traceability system based on identifier resolution includes an identifier coordinate generation module 101, a process evidence collection module 102, a material transformation construction module 103, an identifier resolution verification module 104, and a defect reverse location module 105.

[0013] The identifier coordinate generation module 101 is used to generate industrial internet identifiers for raw materials, yarns, fabric rolls and cut pieces entering the textile processing chain, and to establish material coordinates associated with the corresponding material location for each industrial internet identifier.

[0014] The identifier coordinate generation module 101 is used to establish an Industrial Internet identifier for each material entering the textile processing chain, enabling it to be identified by subsequent processing stations, and to establish material coordinates based on the Industrial Internet identifier. An Industrial Internet identifier is a machine-identifiable code used to uniquely identify a material object. Each Industrial Internet identifier corresponds to a specific material object in the current textile processing chain and is not repeatedly assigned to other materials. Industrial Internet identifiers can be formed by combining enterprise node codes, material type codes, and material sequence codes, or they can be obtained through an existing Industrial Internet identifier resolution system. For example, a roll of yarn to be woven can generate the identifier "TX-YARN-20260823-00018," where different fields are used to distinguish textile business, material type, generation date, and material sequence, respectively. In actual implementation, the specific coding format mentioned above is not required; it is sufficient as long as the corresponding material can be uniquely identified through the identifier resolution service.

[0015] After the identifier is generated, the identifier coordinate generation module 101 establishes a binding relationship between the industrial internet identifier and the material coordinates. The material coordinates are relative positional parameters used to describe the specific location inside the material corresponding to the same industrial internet identifier, not the geographical location of the material in the factory. The material coordinates use the preset coordinate starting point of each material as a reference, so that the subsequent collected processing data, quality inspection data, and transformation relationships between different materials can be further mapped to the specific location inside the material, rather than just to the entire material.

[0016] For yarns and fabric rolls with continuous length characteristics, the material coordinates preferably adopt a length coordinate system that increases along the material conveying direction. The starting end of the yarn or fabric roll is set as the zero point of the material coordinates, and the coordinates increase along the unwinding or feeding direction according to the actual length. For example, after generating an Industrial Internet identifier for a 1200-meter-long fabric roll, a material coordinate system from 0 meters to 1200 meters can be established, where coordinate 320 meters represents the position 320 meters from the starting end of the fabric roll along the feeding direction, and 320 meters to 325 meters represents a material coordinate interval on the fabric roll. The length can be obtained through an encoder, meter wheel, or existing length measuring device on the production equipment. The identifier coordinate generation module 101 stores the correspondence between the Industrial Internet identifier, coordinate starting point, coordinate direction, coordinate unit, and the currently valid coordinate range.

[0017] For cut pieces, the preferred material coordinate system is a two-dimensional positional coordinate system based on the cut piece itself. The lower left corner of the bounding rectangle of the cut piece can be set as the coordinate starting point, with the length and width directions of the cut piece as the two coordinate directions, and millimeters as the coordinate unit. For example, a cut piece approximately 600 mm × 400 mm in size can be located at a position 250 mm long and 120 mm wide from the coordinate starting point. When the actual outline of the cut piece is not rectangular, coordinates outside the outline are not considered valid material coordinates. Therefore, local defects on the same cut piece can be located using the corresponding material coordinates.

[0018] For raw materials entering the processing chain, when the raw material has a continuous form, material coordinates in length form can be established according to the actual feeding direction. When the raw material is loose fiber or other discrete raw material that cannot be directly measured continuously by length, material coordinates can be established according to the actual feeding sequence and cumulative feeding amount. The cumulative feeding amount is the amount of raw material input into the processing equipment from the beginning of the current processing, and can be expressed in grams or kilograms as the coordinate unit. For example, a batch of 100 kilograms of fiber raw material generates an Industrial Internet identifier, with the starting point of feeding determined as the zero point. When 25 kilograms are cumulatively fed into the equipment, the corresponding material coordinate is 25 kilograms, and a material coordinate interval is formed from 25 kilograms to 30 kilograms. This method allows raw materials that cannot be directly represented by length to also have a continuous and calculable position reference.

[0019] Industrial Internet identifiers can be carried by QR codes, RFID tags, electronic tags, or electronic identifier records in a database. After generating the industrial internet identifier, the identifier coordinate generation module 101 records the industrial internet identifier, material type, coordinate starting point, coordinate direction, coordinate unit, effective coordinate range, and generation time, and saves this record to the identifier resolution database. When materials enter the processing station, the processing equipment or identifier reading device reads the industrial internet identifier to obtain the definition method of the corresponding material coordinates and the current effective coordinate range, providing a unified position reference for subsequent process evidence collection.

[0020] In one specific implementation process, before a 1000-meter-long roll of fabric enters the slitting process, the identification coordinate generation module 101 generates an Industrial Internet identifier for it, setting the starting end of the roll to material coordinate 0 meters and the ending end to 1000 meters. If subsequent quality inspection reveals a weaving defect in the 425-428 meter range of the roll, this area can be directly represented as the 425-428 meter material coordinate range under this Industrial Internet identifier. In this way, data generated by different processing equipment can all use the Industrial Internet identifier and material coordinates as a common location reference, providing a clear data foundation for subsequently establishing material change records and performing reverse defect localization.

[0021] In this embodiment, the identifier coordinate generation module establishes material coordinates that are continuously updated as the material moves before the yarn or fabric roll enters the continuous processing process. First, it reads the feeding direction, starting position, and effective length of the yarn or fabric roll and generates a coordinate definition record. The coordinate definition record specifies a unified interpretation method for the current material coordinates. The coordinate origin represents the position where the material begins to be measured; preferably, the starting end of the material actually entering the processing equipment is determined as the coordinate origin. The positive coordinate direction indicates the direction in which the material coordinates increase as the material moves. The coordinate unit can be millimeters or meters. The effective coordinate range represents the start and end range within which the material corresponding to the current industrial internet identifier can be effectively located. For example, for a fabric roll with an effective length of 800 meters, with the end entering the feeding mechanism as the coordinate origin and the direction of movement of the fabric roll towards the processing equipment as the positive coordinate direction, and millimeters as the coordinate unit, the effective coordinate range is 0 millimeters to 800,000 millimeters.

[0022] After obtaining the coordinate definition record, continuous coordinate units are established based on the equipment encoder corresponding to the material feeding motion. The equipment encoder can be installed on the drive feed roller, traction roller, or a rotating shaft that maintains a fixed transmission relationship with it. A coordinate unit is the smallest continuous position interval in the material coordinates that can be distinguished by a single equipment encoder pulse. Adjacent coordinate units are connected end-to-end and do not overlap. After initial calibration, each pulse generated by the equipment encoder corresponds to an initial material displacement, and this initial material displacement is used as the initial length of the coordinate unit. For example, if the initial calibration shows that each equipment encoder pulse corresponds to a 0.5 mm material displacement, then 0 mm to 0.5 mm constitutes the first coordinate unit, 0.5 mm to 1.0 mm constitutes the second coordinate unit, and so on, covering the effective coordinate range.

[0023] Each coordinate unit generates a unique coordinate unit identifier. The coordinate unit identifier must at least determine its associated Industrial Internet identifier and its order within the current material. For example, the Industrial Internet identifier can be combined with a coordinate unit number that increments in the feeding direction to form the coordinate unit identifier. The coordinate unit identifier does not require a physical label; it can serve as a logical identifier stored in the system, used to accurately associate subsequently collected equipment encoder pulses, process sensor data, and quality inspection data with the corresponding material location.

[0024] To avoid cumulative deviations in the pre-set single-pulse material displacement caused by roller processing errors, transmission gaps, fabric tension variations, or yarn expansion and contraction, this embodiment further utilizes the actual material position to perform segmented calibration of the equipment encoder. The actual material position is the true feeding position of the material obtained using a length reference independent of the current equipment encoder. It can be obtained by a metering wheel that has been calibrated, a laser displacement detection device, a visual mark detection device, or an existing independent length detection device on the processing equipment, thereby avoiding the use of the same equipment encoder as both the calculated value and the calibration value.

[0025] After the equipment starts feeding, it reads the actual material position and cumulative encoder pulses at the boundary of the first coordinate unit, and then reads the actual material position and cumulative encoder pulses again at the boundary of the 100th coordinate unit. The pulse displacement conversion parameter represents the actual material movement distance corresponding to one encoder pulse in the current feeding state. It is determined by evenly distributing the difference in actual material position between two boundaries to the cumulative increase in encoder pulses between those two boundaries. For example, if the actual position of the first boundary is 0.50 mm, corresponding to 1 cumulative pulse; and the actual position of the 100th boundary is 50.30 mm, corresponding to 100 cumulative pulses, then the actual material movement between the two boundaries is 49.80 mm, and the encoder pulses increase by 99. Therefore, the material displacement corresponding to each encoder pulse is approximately 0.503 mm. The coordinate generation module then uses 0.503 mm as the pulse displacement conversion parameter for the current segment, instead of continuing to use the initial 0.5 mm.

[0026] The obtained pulse displacement conversion parameters are used to sequentially generate the boundaries of subsequent coordinate units. Each additional encoder pulse adds the current pulse displacement conversion parameter to the previous coordinate unit boundary, resulting in the next coordinate unit boundary. The length of the resulting coordinate units can be adjusted according to the actual feeding status, without requiring the entire yarn or fabric roll to always use the same fixed conversion value.

[0027] Every 100 coordinate units, the current actual material position is read again using an independent length reference, and compared with the corresponding coordinate unit boundary generated according to the current pulse displacement conversion parameters. The difference between the two is taken as the current coordinate deviation. When the coordinate deviation is not greater than the material displacement corresponding to one device encoder pulse, it is considered that the current deviation is still within a minimum coordinate unit, and the current coordinate unit sequence continues to be used. When the coordinate deviation is greater than the material displacement corresponding to one device encoder pulse, the current actual material position and the cumulative device encoder pulses read at the same time are established as a calibration breakpoint.

[0028] A calibration breakpoint is a defined positional reference that separates the previous and subsequent cumulative positioning segments. After establishing a calibration breakpoint, the subsequent position is no longer calculated using all cumulative encoder pulses from the material coordinate origin to the current position. Instead, the actual material position recorded at the calibration breakpoint is used as the new local coordinate reference, and subsequent coordinate units are generated only using the encoder pulses added after the calibration breakpoint. In this way, the cumulative deviation formed before the calibration breakpoint will not continue to be transmitted to subsequent material positions.

[0029] For example, the calculated calibration position based on the current pulse displacement conversion parameters is 100.60 mm, while the actual material position measured by the independent length detection device is 101.20 mm. The current single device encoder pulse corresponds to 0.503 mm. Since the coordinate deviation of 0.60 mm is greater than 0.503 mm, 101.20 mm and the corresponding cumulative device encoder pulses are established as the calibration breakpoint. Assuming the cumulative pulses corresponding to the calibration breakpoint are 200, when the cumulative pulses reach 260, only the newly added 60 device encoder pulses are used to determine the material displacement after the calibration breakpoint, and subsequent coordinate units are generated based on 101.20 mm, instead of re-accumulating the position from the coordinate origin using all 260 pulses.

[0030] As material feeding proceeds, multiple calibration breakpoints can be established. Each calibration breakpoint stores the corresponding actual material position, accumulated device encoder pulses, and pulse displacement conversion parameters used after the calibration breakpoint. The identifier coordinate generation module binds the coordinate definition record, all coordinate unit identifiers, calibration breakpoints, and corresponding accumulated device encoder pulses to the current material's Industrial Internet identifier, forming a material coordinate mapping table. The material coordinate mapping table is used to recover the specific coordinate position of the current material based on the Industrial Internet identifier and device encoder pulses, and serves as the data foundation for subsequent process evidence collection and material transformation construction.

[0031] When the yarn or fabric roll arrives at the processing station, the station reads the Industrial Internet identifier. The identifier coordinate generation module then obtains the corresponding material coordinate mapping table and reads the current cumulative device encoder pulse. The system searches the material coordinate mapping table for the calibration breakpoint where the cumulative device encoder pulse is no greater than the current cumulative device encoder pulse and whose value is closest to the current cumulative device encoder pulse. This calibration breakpoint is then designated as the most recent calibration breakpoint. Subsequently, the number of pulses that the current cumulative device encoder pulse has increased relative to the most recent calibration breakpoint is calculated, and the current position is determined according to the actual material position and pulse displacement conversion parameters corresponding to the most recent calibration breakpoint.

[0032] For example, if the most recent calibration breakpoint is located at the actual material position of 300.00 meters, corresponding to a cumulative encoder pulse count of 600,000, and the current cumulative encoder pulse count is 600,120, and each encoder pulse in the current segment corresponds to 0.5 millimeters, then after the calibration breakpoint, the material will move 60 millimeters, and the current coordinate boundary will be 300.060 meters. The corresponding coordinate unit is defined by this boundary and the boundary corresponding to the next adjacent pulse. The start and end positions of the coordinate unit obtained in this way are used as the current material coordinate interval output. The subsequent process evidence acquisition module can directly bind the data generated during the current processing to this material coordinate interval, thereby avoiding the problem of continuous expansion of the cumulative error of the equipment encoder during long-distance continuous feeding of textile materials, and maintaining a stable and reproducible correspondence between the industrial internet identifier and the actual material position.

[0033] The process evidence acquisition module 102 is used to read the input and output industrial internet identifiers at the processing station, simultaneously acquire equipment encoder pulses, process sensor data and quality inspection data, and divide the process sensor data and quality inspection data into process evidence segments corresponding to the material coordinate range based on the equipment encoder pulses.

[0034] The process evidence acquisition module 102 is used to establish a stable association between the process operation information and quality inspection information generated by the processing station and the specific location of the corresponding material during the actual processing of textiles. This allows for the direct retrieval of the actual process state experienced by a certain material range during processing based on the material coordinates. A processing station refers to the equipment location capable of processing, conveying, inspecting, or changing the state of raw materials, yarns, fabric rolls, or cut pieces. This includes blending equipment, spinning equipment, looms, dyeing and finishing equipment, setting equipment, slitting equipment, cutting equipment, and online quality inspection equipment.

[0035] When materials enter the processing station, the process evidence acquisition module 102 first reads the input Industrial Internet identifier; when processing is completed and a new material object is formed, it reads or receives the output Industrial Internet identifier. The input Industrial Internet identifier represents the material currently entering the processing station, and the output Industrial Internet identifier represents the material formed after passing through the current processing station and continued to be used in subsequent processing processes. The generation method of the Industrial Internet identifier and the correspondence between the Industrial Internet identifier and the material coordinates are established as described above and will not be repeated here.

[0036] The process evidence acquisition module 102 synchronously acquires equipment encoder pulses, process sensor data, and quality inspection data. Synchronous acquisition means that the above data are recorded under a unified time reference, ensuring that each piece of process sensor data and quality inspection data can determine its corresponding equipment encoder pulse range, without requiring all types of data to have exactly the same sampling frequency. In actual implementation, an industrial computer, programmable controller, or edge acquisition terminal can attach equipment timestamps to various types of data and establish a correspondence based on the same equipment clock.

[0037] The encoder pulses are pulse signals output by an encoder mounted on a feed roller, traction roller, winding shaft, loom spindle, or other rotating component with a defined transmission relationship to the material movement. Since the encoder generates a defined number of pulses per rotation angle, the cumulative number of pulses reflects the distance the material travels within the processing equipment. The encoder is preferably installed on an active traction part where there is no significant slippage; when a mechanical transmission ratio exists, the number of pulses per unit material length is predetermined based on the encoder's pulse count per revolution, the roller circumference, and the transmission ratio.

[0038] For example, if the feed roller has a circumference of 500 mm and the encoder outputs 1000 pulses per revolution, and the encoder is directly connected to the feed roller, every 1000 pulses correspond to a material movement of 500 mm, meaning each millimeter corresponds to 2 pulses. If 4000 pulses are accumulated within a certain time period, it can be determined that the material movement during that time period is 2000 mm. If the current input material coordinates start at 35 meters, the corresponding material coordinate range for that time period is 35 meters to 37 meters. In this way, equipment operation data originally collected according to time can be converted into data that directly corresponds to a material coordinate range.

[0039] Process sensing data refers to measurement data reflecting the operating status of equipment or the processing conditions of materials during the manufacturing process. Its specific type depends on the processing station. For example, at the weaving station, it may include loom speed, warp tension, weft tension, and ambient temperature and humidity; at the dyeing and finishing station, it may include dye liquor temperature, level, pH value, operating speed, and roller pressure; at the setting station, it may include oven temperature, fabric speed, and width; and at the slitting station, it may include feeding speed, traction tension, and tool status. The process evidence acquisition module 102 does not require all processing stations to collect identical data; it only needs to collect process sensing data that characterizes the actual operating status of the current processing process.

[0040] Quality inspection data refers to data generated during the processing by online inspection devices or manual inspection terminals to characterize the quality status of materials. This includes defect type, defect location, defect level, color deviation, width deviation, weight measurement results, and other information reflecting the processing quality of textiles. When online visual inspection equipment detects fabric defects, it can input the defect detection time, defect category, and detection location as quality inspection data into the process evidence collection module 102.

[0041] A process evidence segment refers to a data set formed by combining equipment encoder pulses, process sensor data, and quality inspection data according to their corresponding relationships within a defined material coordinate range. Each process evidence segment corresponds to at least one Industrial Internet identifier and one material coordinate range, and records the actual process sensor data and quality inspection data collected during the passage of that material coordinate range through the current processing station. The division length of the process evidence segment can be preset according to the processing technology and data sampling density, for example, forming a process evidence segment every 0.5 meters, 1 meter, or 5 meters, or a new process evidence segment can be formed when the encoder accumulates a preset number of pulses.

[0042] In one specific implementation, the encoder of a loom transmits 2000 pulses corresponding to a fabric movement of 1 meter, and the system is configured to form a process evidence segment every 1 meter. When the coordinates of a fabric roll are between 120 meters and 121 meters, the module records the corresponding 2000 encoder pulses and collects the warp tension, weft tension, loom speed, and online visual inspection results within this pulse range. If a warp tension anomaly occurs near 120.4 meters, and the visual inspection device detects a weaving defect at 120.6 meters, both the tension anomaly information and the weaving defect information are stored in the process evidence segment corresponding to 120 meters to 121 meters. Subsequent queries of this material coordinate range can directly obtain the actual processing status experienced within that range.

[0043] When the sampling frequency of process sensor data is higher than the frequency of process evidence segment division, all sampled values ​​falling within the corresponding encoder pulse range are saved in the corresponding process evidence segment, or the maximum, minimum, average, and abnormal sampled values ​​are saved simultaneously. When the quality inspection data is a discrete event, the material coordinate range to which it belongs is determined based on the cumulative encoder pulses corresponding to the quality inspection event, and then stored in the corresponding process evidence segment. This avoids material position shifts caused by changes in equipment speed when directly matching based solely on timestamps.

[0044] After the process evidence acquisition module 102 completes the acquisition, it saves the input industrial internet identifier, output industrial internet identifier, processing station identifier, material coordinate range, encoder pulse range, processing start and end time, process sensor data, and quality inspection data. Thus, each piece of material can obtain process evidence corresponding to its actual processing position, providing a verifiable processing data foundation for subsequently establishing material change records between different processing stages and for defect reverse location.

[0045] In this embodiment, the process evidence acquisition module is also used to uniformly convert the process sensing data and quality inspection data generated at the processing station according to different sampling cycles into the material movement reference represented by the equipment encoder pulse, and further convert it into continuous material coordinates, so that the processing data and the actual material position passing through the processing station form a definite correspondence.

[0046] After the material enters the processing station, the input and output Industrial Internet of Things (IIoT) identifiers are read first, and the previously established material coordinate mapping table is obtained based on the input IIoT identifier. Simultaneously, the encoders of the equipment with a defined transmission relationship to the material movement are continuously monitored. Each time an encoder pulse is detected, its sequence number is recorded, and the pulse time is recorded using the unified clock at the processing station. The pulse times arranged according to their sequence numbers constitute a pulse time sequence. This pulse time sequence reflects the actual time taken for each distinguishable position to pass through the processing station during the material's movement. Therefore, even if the equipment accelerates or decelerates, the correspondence between time and material movement position can be maintained.

[0047] Process sensing data can include data characterizing the current processing status, such as tension, temperature, pressure, operating speed, and humidity. Quality inspection data can include defect type, defect level, color difference, width deviation, and online visual inspection results. Different data can use different sampling frequencies, and the actual sampling time is recorded for each data item. The process evidence acquisition module, based on the data sampling time, finds the nearest device encoder pulse before the sampling time and the nearest device encoder pulse after the sampling time in the pulse time sequence, and determines the corresponding pulse position based on the position of the sampling time between the two pulse times. Here, the pulse position is the position of the data in the continuous device encoder pulse sequence, which can be an integer pulse or a decimal position between two integer pulses.

[0048] For example, pulse time 10:00:00:100 for pulse sequence number 1200, pulse time 10:00:00:110 for pulse sequence number 1201, and sampling time for a tension data point 10:00:00:106 for a sampling time, then this sampling time is located at the 60% mark of the time interval between the two pulses. Therefore, the pulse position corresponding to this tension data point can be determined as 1200.6. Using this method, various types of data originally recorded according to time are converted to the same device encoder pulse reference. All process sensing data, quality inspection data, and their corresponding pulse positions are arranged in pulse order to form a pulse domain data sequence. The "domain" in the pulse domain data sequence indicates that the data position index uses the device encoder pulse, rather than simply the sampling time.

[0049] Then, the material coordinate mapping table is invoked to convert each pulse position in the pulse domain data sequence into its corresponding material coordinates. For integer pulses, their corresponding material coordinates can be directly determined according to the material coordinate mapping table; for fractional pulse positions between two integer pulses, the positions can be determined based on the material coordinates corresponding to the encoder pulses of two adjacent devices, using the same position ratio. The data arranged in ascending order of material coordinates form a material coordinate sequence. Thus, even if the processing equipment speed changes, and the same time interval corresponds to different lengths of material, the process sensing data is still arranged according to the actual material position passing through the equipment.

[0050] After the material coordinate sequence is formed, segment boundaries are inserted based on definite events during the processing. Segment boundaries are the device encoder pulse positions and corresponding material coordinates used to determine the boundary between two adjacent process evidence segments. When quality inspection data is generated, segment boundaries are inserted at the pulse positions corresponding to the quality inspection data, ensuring that the location of a quality anomaly can independently form an evidence boundary. When a change in the input industrial internet identifier is detected, segment boundaries are inserted at the device encoder pulse positions corresponding to the identifier switch, preventing data from two different input materials from being grouped into the same segment. When process sensing data changes from within the preset process allowable range to outside the preset process allowable range, or recovers from outside the preset process allowable range to within the preset process allowable range, segment boundaries are inserted at the pulse positions corresponding to the first sampled data crossing the range.

[0051] For example, in a weaving station, the allowable range for warp tension is 90 N to 110 N. Adjacent sampled data are 106 N, 108 N, 114 N, and 116 N. The position corresponding to 114 N is the first point where the allowable range is exceeded, and a segment boundary is inserted at the pulse position corresponding to 114 N. When the tension subsequently decreases from 113 N to 108 N, a new segment boundary is inserted at the position corresponding to 108 N. This ensures that the data before, during, and after the process anomaly are in different segments.

[0052] According to the feeding direction, the boundaries of two adjacent segments are obtained, and all equipment encoder pulses, process sensor data, and quality inspection data between the two segment boundaries are extracted to form candidate evidence segments. Each candidate evidence segment records the first pulse, the last pulse, and the first and last material coordinates obtained from the material coordinate mapping table. Candidate evidence segments represent a set of data that has not yet undergone physical position consistency verification and are not directly used as the final process evidence segments.

[0053] A closure check is performed on each candidate evidence segment. The closure check verifies the material movement distance reflected by the device encoder against the material position change reflected in the material coordinate mapping table. Specifically, the cumulative material displacement is determined based on the number of pulses between the first and last pulses of the candidate evidence segment and the pulse displacement conversion parameters of the corresponding segment; simultaneously, the coordinate span is determined by subtracting the first material coordinate from the last material coordinate. Under normal circumstances, both should be consistent or within the error range that the device encoder pulses can distinguish.

[0054] For example, if a candidate evidence segment contains 2000 valid pulses, and each encoder pulse in the current segment corresponds to 0.5 mm, then the cumulative material displacement is 1000 mm. If the material coordinates at the beginning are 325.000 meters and the material coordinates at the end are 326.000 meters, then the coordinate span is also 1000 mm, and the closure check passes. However, if the material coordinates at the end are 326.004 meters, then the coordinate span is 1004 mm, which differs from the cumulative material displacement by 4 mm. This is significantly greater than the 0.5 mm corresponding to a single encoder pulse, requiring further splitting of the candidate evidence segment.

[0055] When the above situation occurs, starting from the beginning of the candidate evidence segment, the material displacement corresponding to the device encoder is accumulated step by step according to the pulse sequence, and the material coordinate span corresponding to the same pulse position is obtained simultaneously. The difference between the two is compared for each additional device encoder pulse. When the difference first exceeds the material displacement corresponding to the current single device encoder pulse, that pulse position is determined as the new segment boundary, and the original candidate evidence segment is divided into two candidate evidence segments, and closure verification is re-performed for each. If the split segment still fails the closure verification, the first out-of-tolerance position is determined and split in the same way. Through this process, the positional error caused by momentary slippage of the equipment, missed pulses, material expansion / contraction, or abnormal coordinate conversion can be limited to the local area where the abnormality actually occurs, without allowing the error to spread to the entire processing area.

[0056] After closure verification, the process sensor data and quality inspection data in the candidate evidence fragments are bound together with the input Industrial Internet identifier, the output Industrial Internet identifier, the device encoder pulse range consisting of the first pulse to the last pulse, and the material coordinate interval consisting of the first material coordinate to the last material coordinate to form a process evidence fragment. Each process evidence fragment thus formed can determine the specific material location, the actual processing state and quality state experienced at that location, and its material location has been verified by the closure of the device encoder cumulative displacement and the material coordinate span. The formed process evidence fragments are then output to the material transformation construction module to establish a transformation record between the input material coordinate interval and the output material coordinate interval, providing processing data with defined positional boundaries for subsequent material transformation map construction and defect reverse location.

[0057] The material transformation construction module 103 is used to establish a transformation record between the input material coordinate range and the output material coordinate range during the blending, weaving, slitting or cutting process, based on the input material feeding amount, the output material forming amount and the equipment encoder pulse. The transformation record includes the coordinate range mapping relationship and the input material contribution ratio.

[0058] The material transformation construction module 103 is used to describe the correspondence between the specific positions of the input material and the specific positions of the output material when the material changes from an input state to an output state during textile processing. Since blending, weaving, slitting, and cutting all cause changes in the shape, quantity, or combination of materials, simply recording the association between the input and output Industrial Internet identifiers cannot determine the actual origin of a specific area in the output material from a particular position in the input material. Therefore, this embodiment further establishes a transformation record between the coordinate ranges of the input and output materials based on the Industrial Internet identifier association.

[0059] A transformation record is a data record established for a single continuous processing step. It represents the specific correspondence between one or more input material coordinate ranges and one or more output material coordinate ranges after the current processing step. Each transformation record includes at least the processing station, input industrial internet identifier, input material coordinate range, output industrial internet identifier, output material coordinate range, coordinate range mapping relationship, input material contribution ratio, and the corresponding equipment encoder pulse range. The input and output material coordinates are established as described above and will not be repeated here.

[0060] The input material feed quantity is the actual amount of input material entering the processing equipment during the corresponding processing period in the current transformation record. For yarn and fabric rolls, it can be expressed as length; for raw materials such as loose fibers, it can be expressed as mass; for cases where multiple yarns participate in weaving simultaneously, the feed length of each input material can also be recorded separately. The output material formation quantity is the amount of output material formed during the same processing period, which can be expressed as length, area, mass, or a unit of measurement consistent with the coordinates of the output material. The equipment encoder pulses are used to determine the start and end positions of the feeding process and the output formation process, thereby avoiding errors caused by estimating the material position solely based on the processing start and end times.

[0061] The coordinate interval mapping relationship represents the positional correspondence between the input material coordinate interval and the output material coordinate interval. During establishment, the material transformation construction module 103 uses the equipment encoder pulse as the processing progress benchmark to determine the cumulative feeding position of the input material and the cumulative formation position of the output material at the beginning and end of the current processing stage, respectively. Then, combined with the material coordinate starting point in the corresponding industrial internet identifier, the cumulative feeding position and the cumulative formation position are converted into the input material coordinate interval and the output material coordinate interval.

[0062] For example, if an input fabric roll is being fed from a material coordinate of 200 meters, and every 2000 encoder pulses from the feed roller correspond to 1 meter of material movement, and the current recording interval is completed after 20000 encoder pulses, then the input material feed amount can be determined to be 10 meters, corresponding to an input material coordinate interval of 200 to 210 meters. If the processing equipment simultaneously produces an 8-meter output fabric roll, and the current material coordinate of the output fabric roll starts from 50 meters, then the corresponding output material coordinate interval is 50 to 58 meters. The material transformation construction module 103 saves the "input 200 to 210 meters corresponds to output 50 to 58 meters" as a coordinate interval mapping relationship. When it is subsequently necessary to determine the upstream position corresponding to approximately 55 meters of the output fabric roll, the corresponding input material position can be further determined within the aforementioned interval based on this transformation record.

[0063] For the slitting process, one input material coordinate range can correspond to multiple output material coordinate ranges. For example, when a wide fabric roll is cut into two narrow fabric rolls along its width, the same input length range can simultaneously correspond to the corresponding length ranges of the two output fabric rolls. The module establishes transformation records for the two output Industrial Internet identifiers, enabling one input Industrial Internet identifier to connect to multiple output Industrial Internet identifiers. For the cutting process, a coordinate range mapping relationship can be established between the corresponding areas in the fabric roll and the resulting cut pieces based on the layout file, cutter movement position, or cutting control data provided by the cutting equipment.

[0064] For processes such as blending or weaving where multiple input materials combine to form the output material, it is also necessary to determine the contribution ratio of each input material. The input material contribution ratio represents the relative contribution of each input material to the formation of the output material within the current output material coordinate range. For blending processes that can be directly measured by mass, this can be calculated based on the actual feed amount of each input material within the current processing range. For example, if two fibers are fed in at 30 kg and 20 kg respectively during a certain processing time period, and together form the corresponding output material range, then the contribution ratio of the first input material is 60%, and the contribution ratio of the second input material is 40%.

[0065] When multiple yarns work together to form fabric during the weaving process, the contribution ratio can be calculated based on the actual feed rate and the predetermined mass per unit length. For example, if the first input yarn is fed 600 meters within the current output material range, with a mass per unit length of 0.02 grams (12 grams), and the second input yarn is fed 400 meters, with a mass per unit length of 0.03 grams (also 12 grams), then both input materials contribute 50%. If the processing equipment can directly output the cumulative mass of each feed channel, then the actual cumulative mass should be used to determine the contribution ratio of the input materials.

[0066] The input material contribution ratio is determined for each output material coordinate interval, rather than just for the entire production batch. When the feeding speed, mixing ratio, or weaving structure changes, the material transformation construction module 103 uses the device encoder pulse position at the time of the change as the new interval boundary, ends the current transformation record, and establishes a new transformation record, so that the coordinate interval mapping relationship and input material contribution ratio are maintained within each transformation record.

[0067] In practice, transformation records can be divided according to the number of encoder pulses on a fixed device, the length of a fixed output material, or the moment when the process status changes. For example, a transformation record can be established for every 1 meter of output fabric. When the equipment detects a change in the feeding ratio, even if the current 1-meter interval has not yet ended, the current transformation record can be terminated early at the changed position, and the next transformation record can be re-established. This can improve the accuracy of local position mapping.

[0068] After the material transformation construction module 103 completes its processing, it saves all transformation records generated during continuous processing in the order of processing time and material coordinates. This ensures that each output material coordinate interval can find its corresponding input material coordinate interval, and further determines the contribution ratio of each input material when multiple input materials exist. Through this method, the processes of merging multiple inputs in blending, separating single inputs in slitting and cutting, and forming multiple materials together in weaving can all be represented using unified transformation records. This provides a clear data foundation for the subsequent generation of material transformation diagrams and for performing reverse defect localization along the processing chain.

[0069] Furthermore, the material transformation construction module is specifically used for: reading the input and output industrial internet identifiers, determining the material mass maintained between the input and output metering positions as the in-machine retention amount; determining the feeding amount and coordinate range of each input material based on the encoder pulses of the input device, converting the feeding amount into mass to form a material segment to be consumed, and adding it to the material queue to be consumed according to the feeding order; when the cumulative mass of the material queue to be consumed reaches the in-machine retention amount, deducting the corresponding mass sequentially from the head of the material queue to be consumed according to the output material formation amount corresponding to each encoder pulse of the output device; when a deduction only consumes a portion of the material segment to be consumed, deducting the mass according to the proportion of the deducted mass to the remaining material segment. The input material coordinate interval is divided according to the proportion of the mass of the material segment to be consumed. The unconsumed part and the corresponding input material coordinate interval are retained for the next deduction. A coordinate interval mapping relationship is established for each input material coordinate interval that is actually deducted in the same output material coordinate interval, and the input material contribution ratio is determined according to the proportion of the deducted mass of each input material to the total deducted mass. When a deduction spans two material segments to be consumed, the fractional pulse position is determined within the current device encoder pulse according to the proportion of the remaining mass of the previous material segment to the output material formation amount corresponding to the current device encoder pulse. The fractional pulse position is used as the boundary of adjacent transformation records, and a transformation record including the coordinate interval mapping relationship and the input material contribution ratio is generated.

[0070] In this embodiment, the material transformation module is used to solve the problem of misalignment between input and output positions caused by the fact that the input material does not immediately form output material after entering the equipment during continuous textile processing. The module first reads the input and output industrial internet identifiers corresponding to the current processing process and determines the input and output metering positions. The input metering position is the position where the encoder of the input device actually measures the amount of material fed, and the output metering position is the position where the encoder of the output device actually measures the amount of output material formed. The mass of material located between the two and continuously existing inside the equipment under normal operating conditions is defined as the in-machine retention. The in-machine retention can be obtained through calibration before the equipment is put into production. For example, after the equipment is in a stable operating state and no new material is added, the mass of output material that continues to form after the input metering position is measured. This mass can be used as the in-machine retention under the current process state; it can also be obtained by accumulating the mass of material held in each transmission section inside the equipment. For equipment where the amount of material held inside the machine changes significantly after the process parameters are changed, the corresponding in-machine retention can be saved for different operating states.

[0071] The module continuously determines the feed amount of each input material based on the encoder pulses of the input device, and uses the material coordinate mapping relationship under the corresponding Industrial Internet identifier to determine the input material coordinate range corresponding to that feed amount. For loose fibers that can be directly measured by mass, the actual feed mass is used directly; for materials such as yarn or fabric that are measured by length, the mass is converted based on the feed length and the mass per unit length. The mass per unit length can be obtained by converting the measured value of the current material or the linear density and areal density in the production process record. The input Industrial Internet identifier, the input material coordinate range, and the material mass corresponding to one continuous feed together form the material segment to be consumed.

[0072] The consumable material queue is a collection of consumable material segments arranged in the order they enter the input metering position. The head of the queue represents the earliest material segment to enter the equipment and participate in subsequent outputs, while the tail of the queue represents the most recently entered material segment. Each consumable material segment records at least the input industrial internet identifier, the input material coordinate range, the initial mass, and the current remaining mass. New material segments are added to the tail of the queue according to the feeding order, thus maintaining the sequential transmission relationship of input materials within the equipment.

[0073] Before the cumulative mass of the material queue reaches the in-machine retention limit, newly entering materials are considered to be used to establish or replenish the material occupancy within the equipment, and a coordinate interval mapping relationship is not immediately established with the current output material. Once the cumulative mass of the material queue reaches the in-machine retention limit, it indicates that the material that first entered the input metering position has the conditions to reach the output metering position and participate in the formation of output material. At this point, queue deduction begins based on the encoder pulses of the output device. Each time an encoder pulse is generated at the output end, the mass to be deducted from the material queue is first determined based on the pre-calibrated output material formation amount corresponding to a single output device encoder pulse, and then deductions are made sequentially starting from the head of the queue.

[0074] For example, if the in-machine retention is 500 grams, and the cumulative mass of the material to be consumed in the queue reaches 500 grams, then each encoder pulse at the output end corresponds to 10 grams of output material. Therefore, for each encoder pulse detected at the output end, 10 grams are deducted from the head of the material to be consumed queue. If the remaining mass of the material segment at the head of the queue is greater than 10 grams, only 10 grams are deducted from that segment; if the remaining mass is exactly 10 grams, that segment is completely consumed after this deduction and removed from the queue; if the remaining mass is less than 10 grams, the entire remaining mass of that segment is deducted first, and then the remaining amount is deducted from the next segment to be consumed.

[0075] When a deduction consumes only a portion of a material segment to be consumed, the input material coordinate interval corresponding to that segment is synchronously divided based on the proportion of the actual deducted mass to the current remaining mass of that segment. For yarns or fabrics that are continuous along the feeding direction and whose mass per unit length remains constant within the current segment, the material coordinate interval can be divided according to the same proportion. For example, if the remaining mass of a material segment to be consumed is 20 grams, and the corresponding input material coordinate interval is 100 meters to 104 meters, and 5 grams are deducted this time, accounting for one-quarter of the remaining mass, then 100 meters to 101 meters is determined as the actual input material coordinate interval consumed this time, and 101 meters to 104 meters and the corresponding 15 grams of remaining mass are retained at the beginning of the material queue to be consumed for deduction in the next output. In this way, the input material coordinates will not be repeatedly calculated or lost because a material segment spans multiple output pulses.

[0076] After each deduction of the output material quantity, a coordinate interval mapping relationship is established between the actual input material coordinate interval deducted and the current output material coordinate interval. The current output material coordinate interval is determined by the output Industrial Internet identifier and the material coordinates corresponding to the encoder pulse of the output device. If an output material coordinate interval is formed by the deduction of multiple input material segments, each input Industrial Internet identifier and its actual deducted input material coordinate interval are recorded separately, and the contribution ratio of each input material is determined based on the proportion of the actual deducted mass of each input material to the total deducted mass.

[0077] For example, if an encoder pulse from an output device corresponds to 10 grams of output material, and material A at the head of the current queue has only 3 grams remaining, and then material B provides the remaining 7 grams, then both material A and material B are mapped to the coordinate range of the current output material. The input material contribution ratio of material A is 30%, and the input material contribution ratio of material B is 70%. This contribution ratio is determined based on the actual mass deducted from the queue of materials to be consumed, rather than solely based on the production formula or the theoretical ratio of the entire batch of materials. Therefore, it can reflect the actual source composition of the local output material.

[0078] When a deduction spans two segments of material to be consumed, a fractional pulse position is determined within the current output device encoder pulse. The fractional pulse position refers to the proportion within the output range corresponding to a complete output device encoder pulse where the previous segment of material to be consumed is exactly completely consumed. For example, if the current output device encoder pulse corresponds to 10 grams of output material, and the previous segment of material to be consumed has 3 grams remaining, then the previous segment of material to be consumed is consumed at 30% of the output range corresponding to that pulse. Therefore, the position 0.3 pulses after the current integer pulse is determined as the fractional pulse position.

[0079] The fractional pulse positions are further converted into specific boundary positions within the current output material coordinate range. Assuming one output device encoder pulse corresponds to a 0.5 mm output material length, then 0.3 pulses correspond to 0.15 mm. This 0.15 mm position serves as the transformation boundary between the output areas corresponding to the previous input material segment and the next input material segment. Therefore, even if the switching between two input material segments occurs within a single complete device encoder pulse, an independent local coordinate range mapping relationship can be formed, without forcibly assigning the entire pulse's corresponding output material to one of the input material segments.

[0080] The module ultimately uses adjacent integer or fractional pulse positions as the boundaries of the transformation record, writing the corresponding input Industrial Internet identifier, input material coordinate range, output Industrial Internet identifier, output material coordinate range, actual deducted mass, and input material contribution ratio into the same transformation record. By continuously executing the above queue deduction and boundary determination process, a continuous transformation record arranged in the order of the output material coordinates can be obtained. This allows the subsequent identifier resolution and verification module to accurately recover the correspondence between different input materials after they have been retained, transferred, and output within the continuous processing equipment and the local area of ​​the output material.

[0081] The identifier resolution and verification module 104 is used to connect the transformation records corresponding to continuous processing steps according to the resolution relationship between industrial Internet identifiers, form a material transformation diagram, and verify the material transformation diagram based on the continuity of material coordinates, the conservation relationship of the contribution ratio of input materials, and the overlap relationship of processing time.

[0082] The identifier resolution and verification module 104 connects the transformation records formed by different processing steps according to the resolution relationship between industrial internet identifiers, and performs consistency verification on the connection results to form a material transformation diagram that can truly reflect the processing evolution relationship between raw materials, yarns, fabric rolls, and cut pieces. The resolution relationship between industrial internet identifiers refers to the ability to obtain other industrial internet identifiers that have a direct processing input relationship or a direct processing output relationship with a material based on a certain industrial internet identifier, through industrial internet identifier resolution service or local identifier resolution database. For example, when yarn A and yarn B are woven into fabric roll C, yarn A and yarn B establish direct processing associations with fabric roll C respectively; when fabric roll C is further cut into fabric rolls D and E, fabric roll C establishes direct processing associations with fabric rolls D and E respectively. After reading the above association relationship, the identifier resolution and verification module 104 connects the corresponding transformation records generated by the material transformation construction module 103 according to the processing sequence.

[0083] A material transformation diagram is a directed relational structure formed by Industrial Internet identifiers and their corresponding material coordinate ranges as nodes, and transformation records representing the processing and conversion relationships between nodes. The direction in the diagram points from the input material to the output material according to the material processing direction. For example, the material coordinate range corresponding to yarn is connected to the material coordinate range corresponding to fabric rolls through weaving, and the material coordinate range corresponding to fabric rolls continues to connect to the material coordinate range corresponding to the next fabric roll or cut piece after slitting or cutting. One input material can connect to multiple output materials, and multiple input materials can also jointly connect to one output material. Therefore, the material transformation diagram can simultaneously represent the branching relationships formed by slitting and the converging relationships formed by blending and weaving.

[0084] When establishing the material transformation diagram, the identifier resolution and verification module 104 first queries adjacent processing steps based on the input and output industrial internet identifiers in each transformation record. When the output industrial internet identifier of the previous processing step matches the input industrial internet identifier of the next processing step, a direct connection is established between the two transformation records; when the same industrial internet identifier corresponds to multiple consecutive transformation records, they are then arranged according to the material coordinate range and processing time sequence. Thus, the industrial internet identifier provides the basis for identity connections across processing steps, and the material coordinate range further provides the basis for local location connections within the same material.

[0085] Material coordinate continuity verification is used to determine whether the coordinate intervals of adjacent transformation records on the same material can be continuously connected. For the same roll of fabric or yarn used continuously in the processing direction, the end coordinate of the previous transformation record should, in principle, be consistent with the start coordinate of the next transformation record, or the difference between the two should be within the preset measurement error range. The preset measurement error range can be determined based on the measurement accuracy of the meter wheel, the equipment encoder, or the feeding mechanism. For example, if a roll of fabric passes through two processing intervals continuously, and the first transformation record corresponds to material coordinates from 100 meters to 110 meters, and the next one corresponds to 110 meters to 120 meters, then the two are continuous; if the next one starts from 112 meters, then there are 2 meters in the middle that are not covered by the transformation record, and the module marks this position as a coordinate breakpoint; if the next one starts from 109 meters, then there is 1 meter of overlapping coverage, which is also marked as abnormal. For the measurement error allowed by the equipment, such as allowing the difference between the preceding and following coordinates to not exceed 0.02 meters, 110 meters and 110.01 meters can be considered continuous.

[0086] For the slitting and cutting processes, material coordinate continuity does not require continuous coordinate values ​​between different output materials. Instead, it verifies whether the input material coordinate range in each transformation record can form a complete and consistent mapping with the corresponding output material coordinate range. For example, when a 200-meter to 210-meter area of ​​the same fabric roll is cut into multiple pieces, the coordinate relationship is considered valid as long as the input material area corresponding to each piece is within the 200-meter to 210-meter range and there are no records exceeding the valid coordinate range of the current input material. This avoids the erroneous direct comparison of coordinate values ​​between different material objects.

[0087] The input material contribution ratio conservation check is mainly used in processing processes such as blending and weaving where multiple input materials jointly form the output material. For the same output material coordinate range, the contribution ratio of all input materials participating in forming that range should be able to fully reflect the source of the output material. In practice, the sum of the contribution ratios of all input materials can be compared with 100%, and an allowable deviation can be set according to the feeding metering accuracy. For example, if an output fabric range is formed by three types of input yarns, with corresponding input material contribution ratios of 40%, 35%, and 25%, the sum is 100%, which passes the check. If the recorded values ​​are 40%, 35%, and 15%, the sum is only 90%, and the allowable deviation is 2%, then the module determines that there is a missing contribution ratio or abnormal feeding data. For processing processes where production losses are recorded separately, the clearly defined loss ratio can be deducted first, and then the consistency between the remaining input material contribution ratio and the actual output quantity can be checked.

[0088] Processing time overlap verification is used to determine whether input materials, output materials, and corresponding process evidence segments participating in the same transformation process have a reasonable shared processing interval in time. Processing time overlap does not require that the start and end times of each record be exactly the same, but rather that the corresponding data actually occur within the same continuous processing process. For example, if an output fabric interval is formed from 10:00:20 to 10:00:30, and the feeding record for input yarn A is from 10:00:19 to 10:00:31, and the feeding record for input yarn B is from 10:00:20 to 10:00:30, then both input records cover this output formation time, and a valid connection can be established. If the feeding record for an input material occurs at 9:50 and does not overlap with the output formation time, then even if the Industrial Internet identifier relationship matches, this record will not be directly connected to the current output material coordinate interval.

[0089] In the specific implementation process, for example, yarn A and yarn B together form fabric roll C at the weaving station, and fabric roll C is then cut into fabric roll D. The identifier resolution and verification module 104 first determines the processing connections from A and B to C and from C to D based on the industrial internet identifier resolution relationship. Then it checks whether the 0-meter to 20-meter material coordinate range formed by C in the weaving process is continuous with the 0-meter to 20-meter material coordinate range used in the subsequent cutting process. Then it checks whether the sum of the input material contribution ratios of A and B to this range satisfies the conservation relationship, and checks whether the feeding time of A and B, the formation time of C, and the collection time of the corresponding process evidence fragments have reasonable overlap. When all conditions are met, the relevant transformation records are retained in the material transformation diagram. When coordinate breakpoints, abnormal contribution ratios, or processing time mismatches occur, the corresponding connections are marked as abnormal, and unconfirmed abnormal connections are prevented from being directly used for subsequent defect reverse location.

[0090] Through the above processing, the transformation records that were originally stored separately in the continuous processing steps can be transformed into a material transformation diagram with clear direction, material position relationship and processing time relationship. At the same time, it can eliminate unreasonable connections caused by identification reading errors, equipment metering deviation, missing material data or processing record mismatch, and provide a verified data basis for subsequently determining the upstream material coordinate range and processing time window corresponding to the defect by reverse along the material transformation diagram.

[0091] In this embodiment, the identifier resolution and verification module is also used to perform connection and closed-loop verification on the transformation records already established between consecutive processing steps, in order to determine whether the material coordinate correspondence, input material contribution ratio, and processing time relationship between different processing steps can be consistently transmitted. The module first reads the transformation records formed in the aforementioned manner and arranges them according to the order of the processing steps. When the output Industrial Internet identifier in one transformation record is consistent with the input Industrial Internet identifier in the transformation record of the next processing step, the processing time corresponding to the two transformation records is further compared. As long as the time range in which the previous processing step forms the current intermediate material and the effective time range in which the next processing step consumes the current intermediate material have a common coverage, the two transformation records are considered to have the conditions for establishing a connection, and a candidate connection is formed. A candidate connection represents a temporary association between two adjacent transformation records that have not yet undergone closed-loop verification.

[0092] When multiple transformation records exist for the same intermediate material, the identifier resolution and verification module collects all the endpoints of the input and output material coordinate intervals involved in these transformation records and arranges them in ascending order of material coordinates. The material region between two adjacent endpoints is defined as an atomic coordinate unit. An atomic coordinate unit is the smallest coordinate interval that is no longer further subdivided according to the boundaries of existing transformation records during the current material transformation diagram verification process. Its function is to ensure that no transformation record boundary crosses within the same atomic coordinate unit, thereby enabling the interval to be propagated using a uniquely determined processing connection relationship.

[0093] For example, if the same fabric roll has material coordinate ranges of 0m to 8m, 8m to 15m, 0m to 10m, and 10m to 15m in adjacent processing steps, the resulting endpoints are 0m, 8m, 10m, and 15m. These adjacent endpoints form three atomic coordinate units: 0m to 8m, 8m to 10m, and 10m to 15m. Subsequent verification uses these atomic coordinate units as the basic objects, thus avoiding errors caused by direct comparison when different processing steps use different range boundaries.

[0094] Boundary verification markers are generated for both endpoints of each atomic coordinate unit. Each boundary verification marker consists of the current Industrial Internet identifier, the endpoint material coordinates, and the endpoint side identifier. The endpoint side identifier distinguishes the start and end points of the atomic coordinate unit along the feeding direction. For example, if an Industrial Internet identifier corresponds to an atomic coordinate unit of 8 meters to 10 meters, then two boundary verification markers are generated to represent the 8-meter start point and the 10-meter end point. The boundary verification markers are only used to track the endpoint propagation results during the verification process and do not change the original material coordinates.

[0095] Simultaneously, each atomic coordinate unit is assigned a unit contribution flag of 1. The unit contribution flag indicates that the current atomic coordinate unit is considered as a complete output contribution unit, used to verify whether this complete contribution can be restored to its complete state after being split upstream and returned to the current atomic coordinate unit. Here, the value 1 represents the total contribution of the current atomic coordinate unit, and does not represent specific mass or length.

[0096] Boundary check markers are then propagated along the candidate connections in the opposite direction to the processing direction. When a boundary check marker passes through a transformation record, it is converted to the endpoint position of the corresponding input material coordinate interval based on the coordinate interval mapping relationship in the transformation record. If an output material coordinate interval is formed by multiple input materials, the unit contribution marker is simultaneously split according to the contribution ratio of each input material. For example, if the unit contribution marker of an atomic coordinate unit is 1, and the input material contribution ratios for the two types of input materials are 60% and 40% respectively, then after propagation upstream, upstream contribution amounts of 0.6 and 0.4 are formed respectively, and upstream boundary markers corresponding to the two types of input materials are obtained respectively. This propagation continues through multiple consecutive processing steps in the same manner, ultimately obtaining the set of upstream boundary markers and the upstream contribution amount corresponding to the current atomic coordinate unit.

[0097] After reverse propagation is completed, the above results are not directly used as valid verification results. Instead, forward propagation is performed along the original candidate connection in the normal processing direction. Forward propagation uses the same transformation records and coordinate interval mapping relationships as reverse propagation, remapping upstream boundary markers to downstream levels, and re-summarizing upstream contributions belonging to the same output material coordinate interval according to the transformation records. When propagation returns to the original atomic coordinate unit, the material coordinates of the return endpoint and the return contribution are obtained. At the same time, the processing time range actually covered by each transformation record is retained during each level of propagation. Propagation to the next processing step is only allowed if there is a common processing time overlap interval in the current level.

[0098] For example, a certain atomic coordinate unit corresponds to a fabric roll of 120 to 121 meters, with a unit contribution marker of 1. After reverse propagation, the upstream contribution of yarn A is 0.55 and the upstream contribution of yarn B is 0.45. After re-propagating them forward along the original candidate connection, if the reconstructed fabric roll has a starting end of 120.002 meters and an ending end of 121.003 meters, the returned contribution is 0.998, and the device encoder's single pulse corresponds to 0.005 meters, and the metering resolution of the input material contribution ratio is 0.005, then both the endpoint difference and the contribution difference are within the allowable range, and the propagation path passes the verification.

[0099] If the difference between the material coordinates of the returned endpoint and the material coordinates of the original endpoint obtained after forward propagation is greater than the material displacement corresponding to a single device encoder pulse, it indicates that the corresponding candidate connection has caused coordinate distortion that cannot be explained by the device positioning resolution. If the difference between the returned contribution and 1 is greater than the measurement resolution of the input material contribution ratio, it indicates that there is a loss or duplication of input material contribution in the corresponding processing path. If a common processing time overlap interval cannot be obtained at a certain level of forward propagation, it indicates that although the corresponding transformation records can be connected by the Industrial Internet identifier, the actual processing time cannot constitute the same material transfer process.

[0100] When any of the above situations occur, the identifier resolution and verification module checks the propagation results of each level of candidate connection according to the processing order from upstream to downstream. The candidate connection that causes the endpoint coordinates, return contribution, or processing time relationship to exceed the corresponding verification condition for the first time is identified as an abnormal connection. Using the location where the anomaly first occurs for localization can avoid mistaking multiple subsequent deviations caused by the continued propagation of the anomaly as multiple independent errors.

[0101] After identifying an abnormal connection, it is disconnected from the current material transformation diagram, and the material coordinate positions corresponding to both ends of the abnormal connection are set as verification breakpoints. A verification breakpoint indicates that the transformation relationship before and after the current position cannot be directly used as a continuous valid path; subsequent defect reverse localization will not continue to cross this abnormal connection when reaching the verification breakpoint. The remaining candidate connections, verified through coordinate, contribution, and processing time closed-loop checks, continue to be connected according to the Industrial Internet identifier, material coordinate order, and processing time relationship, thus forming a verified material transformation diagram. This provides a data foundation with clearly defined material coordinate boundaries and valid processing connection relationships for defect reverse localization.

[0102] The defect reverse location module 105 is used to obtain the finished product industrial internet identifier and defect coordinate range, reverse the coordinate range mapping along the verified material transformation diagram, determine the upstream material coordinate range and processing time window corresponding to the defect coordinate range, and extract the process evidence fragments corresponding to the processing time window as the source tracing result.

[0103] The defect reverse location module 105 is used to trace the source when a quality defect occurs in the finished product. It uses the finished product industrial internet identifier and the defect coordinate range as the entry point, and performs coordinate range mapping step by step in the opposite direction of the processing direction along the verified material transformation diagram. It determines the upstream material coordinate range and processing time window related to the formation of the defect from the material position corresponding to the finished product, and extracts the corresponding process evidence fragments from the processing time window, thereby forming a traceability result that can reflect the defect location, material source and processing status.

[0104] The finished product industrial internet identifier is used to uniquely identify the finished product object that needs defect tracing, and its generation and parsing methods are consistent with those of the aforementioned industrial internet identifier. The defect coordinate range refers to the range of the detected quality defect location within the material coordinate system of the finished product. For finished products directly formed from fabric rolls, the material coordinate system along the length direction can be used; for finished products formed from cut pieces, the aforementioned two-dimensional material coordinate system can be used. When the quality inspection equipment only outputs the defect center location, the defect coordinate range can be determined according to the defect length, width, or defect outline given by the inspection equipment. For example, if a fabric defect is detected within the length range of 485.2 meters to 485.8 meters of a fabric roll, then 485.2 meters to 485.8 meters is taken as the current defect coordinate range.

[0105] Reverse coordinate interval mapping refers to using the correspondence between the output material coordinate intervals and input material coordinate intervals recorded in the material transformation diagram to deduce the input material positions that participated in forming the output material position from the output material position. Reverse processing is not simply querying the entire upstream batch level by level according to the Industrial Internet identifier; instead, it retains the local material coordinate intervals that actually map to the current defect coordinate interval at each processing step, thus allowing the traceability range to propagate upstream along the processing chain.

[0106] In practical implementation, the defect reverse location module 105 first uses the finished product industrial internet identifier to determine the most recent one or more transformation records connected to the finished product in the material transformation diagram, and determines which output material coordinate interval the defect coordinate interval falls into. If the defect coordinate interval is entirely within one output material coordinate interval, then the transformation record is directly used for reverse mapping; if the defect coordinate interval spans multiple output material coordinate intervals, then the defect coordinate interval is divided into multiple sub-intervals according to the interval boundaries of each transformation record, and reverse mapping is performed upstream for each sub-interval, and then the upstream material coordinate interval corresponding to each sub-interval is retained.

[0107] When both the input and output material coordinate ranges in the transformation record are continuous length ranges, the upstream material coordinate range can be determined based on the relative position within each range. For example, a transformation record indicates that an input fabric roll of 200 to 210 meters corresponds to an output fabric roll of 50 to 58 meters. If the defect coordinate range is located between 54 and 55.6 meters of the output fabric roll, then the starting point of the defect in the output range is located at the halfway point of the entire output range, and the ending point is located at the 70% point of the entire output range. Therefore, after mapping upstream, the corresponding input fabric roll is approximately 205 to 207 meters. In actual implementation, the module can further refine the mapping based on the encoder pulses stored in the transformation record, prioritizing the position corresponding to the encoder pulses to reduce linear scaling errors caused by processing expansion and contraction.

[0108] For branching relationships formed by slitting or cutting, the defect reverse location module 105 selects the transformation record directly corresponding to the current finished product's industrial internet identifier, without entering other output branches unrelated to the current finished product. For example, if an input fabric roll simultaneously forms two slitting fabric rolls, when a defect is located in one of the slitting fabric rolls, it only reverse-locates the actual corresponding area in the input fabric roll along the transformation record connected to the corresponding slitting fabric roll, without adding the material coordinates of the other slitting fabric roll to the traceability result.

[0109] For blended or woven fabrics, the same output material coordinate range may correspond to multiple input material coordinate ranges simultaneously. In this case, the defect reverse localization module 105 continues to reverse locate along each input branch, retaining the corresponding input material contribution ratio. For example, if a fabric defect coordinate range corresponds to two types of input yarns, with input material contribution ratios of 60% and 40% respectively, then the upstream material coordinate ranges corresponding to both yarns are included in the traceability results, and the input material contribution ratios of 60% and 40% are recorded respectively. This avoids missing other materials that actually participated in forming the defect area by only returning one upstream material during multi-input processing.

[0110] The processing time window refers to the actual processing time range during which the current upstream material coordinate interval is transformed into the downstream material coordinate interval through the corresponding processing station. The defect reverse location module 105 determines the processing time window based on the correspondence between the equipment encoder pulse range in the transformation record and the equipment encoder pulse and acquisition time stored in the process evidence acquisition module 102. For example, if the equipment encoder pulse range corresponding to the upstream material coordinate interval of 205 meters to 207 meters is 810000 to 814000, and the process evidence acquisition record indicates that this pulse range occurs from 10:15:20 to 10:15:32, then 10:15:20 to 10:15:32 is determined as the current processing time window.

[0111] For defects requiring tracing across multiple processing steps, the corresponding upstream material coordinate range and processing time window are determined for each processing step. For example, a defect in a finished cut piece is first reverse-located to a material coordinate range within a fabric roll. Then, based on the weaving transformation record corresponding to the fabric roll, the coordinate range of the yarn material involved in forming that fabric roll range is further located, and the cutting processing time window and weaving processing time window are obtained respectively. The reverse process continues until a preset upstream material type is reached or no earlier valid transformation record exists in the material transformation diagram.

[0112] After determining the processing time window, the defect reverse location module 105 queries process evidence segments according to the processing station and processing time window. If the processing time window falls entirely within a single process evidence segment, that process evidence segment is extracted; if the processing time window spans multiple process evidence segments, all process evidence segments that overlap with the processing time window and have a corresponding relationship in the material coordinate range are extracted. The extracted content includes the actual process sensor data and quality inspection data recorded during the corresponding processing period, thereby enabling the determination of whether there are tension abnormalities, temperature abnormalities, feeding abnormalities, or recorded quality abnormalities when the defect area passes through each processing step.

[0113] For example, after reverse mapping, the defect coordinate range corresponding to the industrial internet identifier of a certain finished product is determined to originate from the 205-meter to 207-meter section of the fabric roll. The weaving processing time window for this range is from 10:15:20 to 10:15:32. After querying the corresponding process evidence segment, it is found that the warp tension significantly deviates from the current process setting range between 10:15:24 and 10:15:28. Simultaneously, online quality inspection data records weaving defects at the corresponding material location. Therefore, this process evidence segment, along with the upstream material coordinate range of the 205-meter to 207-meter section of the fabric roll, is jointly written into the traceability results. If the specific yarn material coordinate range can be located further upstream, the corresponding processing time window and process evidence segment are then added.

[0114] The final traceability results include at least the finished product's Industrial Internet identifier, defect coordinate range, progressively determined upstream Industrial Internet identifiers, upstream material coordinate ranges, processing station, processing time window, and corresponding process evidence fragments. When multiple input materials exist, the contribution ratio of each input material is also included. Thus, quality defects can be traced step-by-step from their specific material location in the finished product to the upstream material location and corresponding processing process that actually contributed to that location, rather than simply remaining at the batch or roll level. This provides a traceability basis for analyzing the causes of textile defects, showing a correspondence between location and processing process.

[0115] In this embodiment, the defect reverse location module is also used to trace the material area that actually corresponds to the specific defect location in the finished product upstream of the processing chain along the verified material transformation diagram when a local quality defect occurs in the finished product. The reverse location result is verified by using process evidence fragments and forward feedback results, thereby avoiding misjudging upstream materials that have no direct location correspondence with the defect as the source of the defect based solely on the association relationship between industrial Internet identifiers.

[0116] The defect reverse localization module first reads the finished product's industrial internet identifier and the detected defect coordinate range, then defines this range as the defect seed range. The defect seed range is the initial propagation range of this reverse localization within the finished product's material coordinates, and its boundaries are determined by the defect detection results. For example, if online quality inspection equipment determines that there is a weaving defect between 125.40 meters and 125.70 meters in the finished fabric, then 125.40 meters to 125.70 meters is considered the defect seed range. For two-dimensional cut pieces, the smallest coverage area of ​​the defect contour in the material coordinates can be used as the defect seed range.

[0117] Assign a defect contribution value of 1 to the defect seed interval. The defect contribution value represents the relative amount of the current backpropagation branch's contribution to the original defect seed interval's source contribution. A value of 1 indicates all source contributions that have not yet been split. The defect contribution value does not represent the severity of the defect, nor does it represent the probability of the defect occurring. Instead, it is used to record the proportion of the source contribution borne by each backpropagation path when multiple input materials jointly form a single output material.

[0118] Subsequently, following the verified material transformation diagram in the opposite direction of the processing direction, we search for transformation records that intersect with the defect seed interval. If the defect seed interval only covers a portion of the output material coordinate interval of a certain transformation record, we only perform reverse coordinate interval mapping on the intersecting portion, without propagating the entire output region corresponding to the entire transformation record upstream. Based on the coordinate interval mapping relationship stored in the transformation record, the intersecting output material coordinate intervals are converted into the corresponding upstream material coordinate intervals.

[0119] When the current output material is formed by multiple input materials, the current defect contribution is split according to the contribution ratio of each input material. For example, if a section of fabric corresponding to a defect seed interval is formed by yarn A and yarn B, with input material contribution ratios of 0.6 and 0.4 respectively, then the defect contribution of 1 is split into 0.6 and 0.4, and continues to propagate backward along the upstream material coordinate intervals corresponding to yarn A and yarn B respectively. If there are multiple input relationships in the upstream processing of yarn A, then 0.6 is used as the current defect contribution for that path and is further split according to the input material contribution ratio in the next transformation record. This preserves the contribution paths of each upstream material to the same defect area during multi-level mixing and weaving processes.

[0120] During the reverse propagation process, upon reaching each upstream material coordinate interval, the next candidate connection is checked for the existence of the aforementioned verification breakpoint. A verification breakpoint indicates that the transformation relationship on both sides of that position has not passed the closed-loop verification of material coordinates, input material contribution ratio, or processing time. Therefore, when the reverse propagation reaches a verification breakpoint, the current path is immediately terminated, and the data on the other side of the verification breakpoint is not used to continue inferring the upstream source. Other paths without verification breakpoints continue propagation until a preset upstream material type is reached or no earlier valid transformation record exists.

[0121] For each upstream material coordinate interval obtained from reverse positioning, the device encoder pulses corresponding to the start and end coordinates are determined according to the material coordinate mapping table. The device encoder pulse range is formed by the device encoder pulses corresponding to the two endpoints. Then, the times corresponding to the beginning and end of the device encoder pulse range are obtained using the aforementioned pulse time series, and the time range between the two is determined as the processing time window. If the material coordinate endpoint is located between two integer device encoder pulses, the endpoint time is determined according to the corresponding decimal pulse position between adjacent pulse times.

[0122] For example, if the encoder pulse range corresponding to a certain upstream material coordinate interval is 420000 to 420600, and in the pulse time sequence, 420000 corresponds to 10:12:15.200 seconds, and 420600 corresponds to 10:12:18.500 seconds, then the processing time window for this upstream material coordinate interval is from 10:12:15.200 seconds to 10:12:18.500 seconds. This processing time window is obtained directly from the encoder pulses corresponding to the material position, rather than being estimated based on the overall start and end time of the production batch.

[0123] After determining the processing time window and the equipment encoder pulse range, a dual screening process is performed on the existing process evidence segments. Only when the acquisition time range of the process evidence segment overlaps with the current processing time window, and the equipment encoder pulse range recorded in the process evidence segment also overlaps with the current equipment encoder pulse range, is the process evidence segment extracted and bound to the current upstream material coordinate range. By simultaneously satisfying both the time overlap and pulse range overlap conditions, irrelevant data generated when the same processing equipment processes other materials at similar times can be excluded.

[0124] After obtaining the upstream material coordinate range with process evidence fragments, it is necessary to perform forward re-entry verification on the reverse positioning results. Forward re-entry refers to remapping the upstream material coordinate range to the finished product side by sequentially calling the same coordinate range mapping relationship in the opposite direction according to the actual reverse path just traversed. Using the original reverse path instead of searching for other paths ensures that forward verification and reverse positioning are for the same set of processing connections. The material coordinate range finally mapped to the finished product side is defined as the re-entry coordinate range.

[0125] During the forward propagation process, the defect contribution carried by each backpropagation branch is re-aggregated according to the corresponding convergence relationship. For example, if the value 1 is split into 0.6 and 0.4 during backpropagation, and both branches re-enter the same path that forms the current finished product area after their respective upstream verifications, then the defect contribution after forward aggregation should be restored to 1. If one of the branches cannot return through a valid connection, the aggregated value may only be 0.6, indicating that the source relationship obtained from the backpropagation is incomplete.

[0126] Finally, both the spatial boundary and the defect contribution are verified simultaneously. The starting point of the feedback coordinate interval is compared with the starting point of the defect seed interval, and the ending point of the feedback coordinate interval is compared with the ending point of the defect seed interval. The difference between the two endpoints must not exceed the material displacement corresponding to a single encoder pulse of the current material. Simultaneously, the difference between the forward-summed defect contribution and 1 must not exceed the metering resolution of the input material contribution ratio. The metering resolution of the input material contribution ratio is calculated from the smallest change that the metering device used to determine the input material feed quality or feed length can distinguish. For example, if the system can record the input material contribution ratio with a minimum ratio unit of 0.01, the metering resolution is 0.01.

[0127] For example, if the defect seed range is 125.400 meters to 125.700 meters, and a single device encoder pulse corresponds to 0.005 meters, then a forward regression yields 125.403 meters to 125.696 meters. The difference at the starting end is 0.003 meters, and the difference at the ending end is 0.004 meters, both not exceeding 0.005 meters. If the defect contribution after forward aggregation is 0.996, and the input material contribution ratio measurement resolution is 0.01, then the contribution difference is also within the allowable range, and the reverse path passes verification. Conversely, if the error at any endpoint exceeds the material displacement corresponding to a single device encoder pulse, or if the defect contribution cannot be recovered to 1 within the allowable error range, the corresponding process evidence fragment will not be used as the final tracing result.

[0128] The source tracing result is formed by combining the verified upstream material coordinate range, corresponding processing time window, and process evidence fragments. Therefore, defect reverse location is not just about finding upstream batches based on identification association, but rather starting from the actual location of the defective material, completing closed-loop verification through coordinate reverse propagation, contribution amount breakdown, dual constraints of equipment encoder pulses and time, and forward retracing of the original path. This ensures that the final retained process evidence has a verifiable positional correspondence with the defective area of ​​the finished product.

[0129] Figure 2This paper compares the effectiveness of the present invention with traditional batch traceability methods in reverse locating defects in finished products. Taking the defect coordinate range of 125.4 m to 125.7 m of finished fabric as an example, Figure (a) shows the traditional batch traceability method, which mainly relies on batch or material identification to query upstream production records. It can only determine that the defect comes from the entire upstream roll of fabric from 0 m to 120 m, and the corresponding process time window to be searched is about 600 s. It cannot further determine the local material location corresponding to the defect. Figure (b) shows the result of traceability using the present invention. The system takes the finished product industrial internet identification and defect coordinate range as input, performs coordinate range mapping in reverse along the verified material transformation diagram, locates the finished product defect range of 125.4 m to 125.7 m to the upstream roll of fabric from 54.2 m to 54.8 m, and determines a processing time window of about 3.3 s based on the encoder pulse of the corresponding equipment. The corresponding process evidence fragment is extracted from this processing time window. As can be seen, the present invention can reduce the traditional traceability of the entire batch of materials to the coordinate range of specific materials, and simultaneously compress the time range of the process evidence that needs to be retrieved, thereby improving the positioning accuracy of textile defect tracing.

[0130] Although this application discloses preferred embodiments as described above, it is not intended to limit this application. Any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of this application. Therefore, the scope of protection of this application should be determined by the scope defined in the claims of this application.

Claims

1. A textile processing traceability system based on identifier resolution, characterized in that, include: The identifier coordinate generation module is used to generate industrial internet identifiers for raw materials, yarns, fabric rolls and cut pieces entering the textile processing chain, and to establish material coordinates associated with the corresponding material location for each industrial internet identifier. The process evidence acquisition module is used to read the input and output industrial internet identifiers at the processing station, simultaneously collect equipment encoder pulses, process sensor data and quality inspection data, and divide the process sensor data and quality inspection data into process evidence segments corresponding to the material coordinate range based on the equipment encoder pulses. The material transformation construction module is used to establish a transformation record between the input material coordinate range and the output material coordinate range during the blending, weaving, slitting or cutting process, based on the input material feed amount, output material formation amount and equipment encoder pulses. The transformation record includes the coordinate range mapping relationship and the contribution ratio of the input material. The identifier resolution and verification module is used to connect the transformation records corresponding to continuous processing steps according to the resolution relationship between industrial internet identifiers, form a material transformation diagram, and verify the material transformation diagram based on the continuity of material coordinates, the conservation relationship of the contribution ratio of input materials, and the overlap relationship of processing time. The defect reverse location module is used to obtain the finished product industrial internet identifier and defect coordinate range, reverse the coordinate range mapping along the verified material transformation diagram, determine the upstream material coordinate range and processing time window corresponding to the defect coordinate range, and extract the process evidence fragments corresponding to the processing time window as the traceability result.

2. The textile processing traceability system based on identifier resolution according to claim 1, characterized in that, The identifier coordinate generation module is specifically used for: Read the feeding direction, starting position, and effective length of the yarn or fabric roll, and generate a coordinate definition record including the coordinate origin, positive coordinate direction, coordinate unit, and effective coordinate range; Based on the coordinate definition record, the material displacement corresponding to a single pulse of the device encoder is used as the smallest dividing unit to divide the effective coordinate range into continuous coordinate units, and a coordinate unit identifier corresponding to each coordinate unit is generated. Read the actual material position and cumulative device encoder pulse at the boundary of the first coordinate unit and the hundredth coordinate unit, and determine the pulse displacement conversion parameter based on the difference between the actual material position and the cumulative difference between the device encoder pulse; The boundary of the subsequent coordinate unit is generated using the pulse displacement conversion parameter. The actual material position is reread every 100 coordinate units. When the deviation between the actual material position and the boundary of the corresponding coordinate unit is greater than the material displacement corresponding to a single pulse, the current actual material position and the corresponding cumulative device encoder pulse are established as a calibration breakpoint. A new coordinate unit is generated based on the subsequent device encoder pulse increment, with the calibration breakpoint as the starting point. Bind the coordinate definition record, coordinate unit identifier, calibration breakpoint and corresponding cumulative device encoder pulse to the industrial internet identifier to generate a material coordinate mapping table; After reading the industrial internet identifier at the processing station, the material coordinate mapping table is obtained, the nearest calibration breakpoint before the current cumulative device encoder pulse is determined, and the current material coordinate range is determined based on the actual material position corresponding to the calibration breakpoint and the pulse increment of the current cumulative device encoder pulse relative to the calibration breakpoint.

3. The textile processing traceability system based on identifier resolution according to claim 1, characterized in that, The process evidence collection module is specifically used for: Read the input and output industrial internet identifiers, obtain the corresponding material coordinate mapping table, and record the pulse sequence number and pulse time of the encoder pulses of the continuous equipment to form a pulse time sequence; Collect process sensor data and quality inspection data with sampling time, determine the equipment encoder pulses adjacent to the sampling time of each data item according to the pulse time sequence, convert each data item into the corresponding pulse position, and form a pulse domain data sequence; Based on the material coordinate mapping table, the pulse domain data sequence is converted into a material coordinate sequence, and segment boundaries are inserted at the locations where quality inspection data is generated, where the input industrial internet identifier changes, and where process sensor data crosses the preset process allowable range. Candidate evidence segments are formed using the encoder pulses between adjacent segment boundaries, and the first pulse, the last pulse, the first material coordinates, and the last material coordinates of the candidate evidence segments are recorded. Closedness verification is performed based on the cumulative material displacement from the first pulse to the last pulse and the coordinate span from the first material coordinate to the last material coordinate. When the difference between the two is greater than the material displacement corresponding to a single device encoder pulse, the candidate evidence segment is split and closedness verification is re-executed when the difference first exceeds the pulse position corresponding to the material displacement. The process sensing data and quality inspection data in the candidate evidence fragments that have passed the closure verification are bound with the input industrial internet identifier, the output industrial internet identifier, the equipment encoder pulse range and the material coordinate range to form process evidence fragments and output them for the material transformation construction module to call.

4. The textile processing traceability system based on identifier resolution according to claim 1, characterized in that, The material transformation construction module is specifically used for: Read the input and output industrial internet identifiers, and determine the material mass maintained between the input metering position and the output metering position as the in-machine retention amount; The feed amount and coordinate range of each input material are determined based on the encoder pulse of the input device. The feed amount is converted into mass to form a segment of material to be consumed, and then added to the queue of material to be consumed according to the feeding order. When the cumulative mass of the material queue to be consumed reaches the amount of material retained in the machine, the corresponding mass is deducted sequentially from the head of the material queue to be consumed based on the amount of material formed by the encoder pulse of each device at the output end. When a deduction consumes only a portion of the material segment to be consumed, the input material coordinate interval is divided according to the proportion of the deducted mass to the mass of the material segment to be consumed. The unconsumed portion and the corresponding input material coordinate interval are retained for the next deduction. Establish a coordinate interval mapping relationship for each input material coordinate interval that is actually deducted from the same output material coordinate interval, and determine the contribution ratio of each input material based on the proportion of the deducted mass to the total deducted mass. When a deduction spans two material segments to be consumed, the fractional pulse position is determined within the current device encoder pulse based on the proportion of the remaining mass of the previous material segment to the output material formation amount corresponding to the current device encoder pulse. The fractional pulse position is used as the boundary of adjacent transformation records, and a transformation record including coordinate interval mapping relationship and input material contribution ratio is generated.

5. The textile processing traceability system based on identifier resolution according to claim 1, characterized in that, The identifier parsing and verification module is specifically used for: Analyze the transformation records of continuous processing steps, and form candidate connections for records where the output Industrial Internet identifier is the same as the input Industrial Internet identifier of the next process and the processing time overlaps. All input material coordinate interval endpoints and output material coordinate interval endpoints that are candidate materials connected to the same intermediate material are collected and divided into atomic coordinate units according to adjacent endpoints; For each atomic coordinate unit, a boundary check mark consisting of an Industrial Internet identifier, endpoint material coordinates, and endpoint side is generated at both endpoints, and a unit contribution mark with a value of 1 is assigned. Propagate the boundary verification markers backward along the candidate connection, and split the unit contribution markers according to the input material contribution ratio of each transformation record to obtain the upstream boundary marker set and upstream contribution amount; Propagate the upstream boundary marker set and upstream contribution along the original candidate connection in the forward direction to reconstruct the return endpoint material coordinates and return contribution. When the difference between the material coordinates at the return endpoint and the material coordinates at the original endpoint is greater than the material displacement corresponding to a single device encoder pulse, the difference between the return contribution and 1 is greater than the metering resolution of the input material contribution ratio, or there is no overlapping processing time interval in the forward propagation, the candidate connection that first appears under the above conditions according to the processing sequence is identified as an abnormal connection, the abnormal connection is disconnected and its two ends are set as verification breakpoints, and the remaining candidate connections form a verified material transformation diagram.

6. The textile processing traceability system based on identifier resolution according to claim 1, characterized in that, The defect reverse location module is specifically used for: Read the finished product's industrial internet identifier and defect coordinate range, use the defect coordinate range as the defect seed range, and assign a defect contribution value of 1. The transformation record that intersects with the defect seed interval is found in reverse along the verified material transformation diagram. The upstream material coordinate interval is obtained according to the coordinate interval mapping relationship. The defect contribution is split according to the input material contribution ratio. The path propagation is terminated when the reverse path reaches the verification breakpoint. The encoder pulse range of the corresponding equipment is determined based on the start and end coordinates of each upstream material coordinate range, and the corresponding processing time window is determined based on the pulse time sequence. Process evidence segments that simultaneously satisfy time overlap and pulse range overlap are extracted using processing time windows and equipment encoder pulse ranges, and then bound to the corresponding upstream material coordinate range; The coordinate range of upstream materials with process evidence segments is mapped forward along the original reverse path to obtain the coordinate range of finished product side feedback, and the defect contribution is summarized in the forward direction. When the difference between the two endpoints of the feedback coordinate interval and the corresponding endpoint of the defect seed interval is not greater than the material displacement corresponding to a single device encoder pulse, and the difference between the summed defect contribution and 1 is not greater than the input material contribution ratio measurement resolution, the corresponding upstream material coordinate interval, processing time window, and process evidence fragment are retained as the tracing result.