A method for tracing information management of polyester blended fabric production process

Abstract

The present invention relates to the field of production traceability technology, specifically a method for traceability information management of polyester blended fabric production process, comprising the following steps: obtaining batch jump nodes, analyzing structure and tension mutation frequency, identifying dense inflection point sections, correlating equipment start and stop behaviors, and constructing a traceability node registration table. In the present invention, by synchronously analyzing the start and stop status of process nodes and tension feedback, a jump index table is constructed, which can accurately identify batch breakpoints and achieve precise positioning of abnormal process sections. Combining the multi-point slope trend in tension feedback with the time synchronization analysis of the structural image, the synergistic relationship between fiber arrangement and torque change can be revealed, and potential abnormal windows can be quickly marked. With the help of the corresponding analysis of tension fluctuations and structural disturbance frequencies, it is helpful to trace the causes of abnormal fluctuations. By hierarchically clustering the dense tension inflection point areas and quantifying the inflection point distribution density, high-risk process sections can be highlighted in the dynamic tension fluctuation sequence.

Description

Technical Field

[0001] The present invention relates to the technical field of production tracing, and in particular to a method for managing production process tracing information of polyester blended fabrics. Background Art

[0002] The field of production traceability involves the collection, recording, and management of relevant information about products throughout their production, processing, and transportation, enabling systematic control over their origins and distribution. Core elements of this technology include raw material identification, production batch correlation, process node tracking, and finished product status integration. Through data collection devices, information encoding methods, and database construction, critical information throughout a product's lifecycle is systematically managed.

[0003] Among them, the traceability information management method for the polyester blended fabric production process refers to the information recording and management method for polyester blended fabrics in specific processes such as spinning, weaving, dyeing and finishing, including the management of raw material ratio information, production process node information, processing equipment status information, time sequence information of each process and operator responsibility information, and synchronizes the above information to the central information system by setting unified coding rules and information entry procedures.

[0004] While existing technologies cover raw material ratios, process nodes, and operation records, most rely on single-dimensional information registration and lack the ability to identify the linkage between material batch changes and equipment behavior. When process anomalies occur, it's impossible to trace the causal relationship between tension fluctuations, structural variations, and equipment startups and shutdowns. This results in post-hoc analysis of the incident based on static data, resulting in coarse traceability and frequent location ambiguity or misjudgment. For example, taking batch interruptions as an example, current technologies only record the point of interruption but do not assess its impact on downstream structural alignment or tension conditions, making it difficult to achieve a fully closed data loop. Furthermore, tension feedback and image sequence data are not integrated into a unified time-series analysis system, failing to capture the dynamic relationship between fiber alignment anomalies and mechanical response. This forces process adjustments to rely more on empirical judgment, impacting traceability accuracy and real-time response. For example, when a tension curve exhibits abnormal fluctuations, existing systems often cannot determine whether it was caused by the sudden startup or shutdown of preceding equipment. The lack of a corresponding mechanism for determining the risk level of sequence skipping ultimately prevents timely identification and intervention of the abnormal process segment. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a polyester blended fabric production process traceability information management method.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a polyester blended fabric production process traceability information management method, comprising the following steps:

[0007] S1: Obtain the continuous process nodes on the polyester blended production line, capture the node indexes of discontinuous node batch sequences, and establish the start-stop tension batch jump index table;

[0008] S2: According to the start-stop tension batch jump index table, the jump section is set as a fixed detection window, the arrangement level spacing slope of multiple consecutive points in the window is calculated, and the window of the mutation position is marked according to the direction of the multi-point slope to obtain the fiber level and torque mutation position set;

[0009] S3: Based on the marked mutation position time period of the fiber layer and the torque mutation position set, the amplitude change frequency in each time segment is counted, and the interactive response area is selected according to the change frequency to generate the tension fluctuation synchronization segment;

[0010] S4: Based on the recorded time segments in the tension fluctuation synchronization segment and the locations of the inflection points, the density of the inflection points is calculated, and the density of all segments is grouped and clustered to generate a tension inflection point dense segment;

[0011] S5: According to the inflection point peak section in the tension inflection point dense section, the start and stop status records of the corresponding equipment are matched, and by screening the risk sections and corresponding process nodes in the start and stop status records, a polyester blended fabric process traceability node registration table is established.

[0012] As a further solution of the present invention, the start-stop tension batch jump index table includes a jump node position index, a corresponding tension feedback index segment, a batch continuity interruption identifier, and a start-stop time interval status identifier; the fiber layer and torque mutation position set includes a mutation detection window position, a layer spacing slope change direction, a torque reversal mark bit, and a mutation node identification number; the tension fluctuation synchronization segment is specifically a trajectory disturbance amplitude distribution segment, a tension switching frequency segment, a frequency overlap mark interval, and a synchronization response identification number; the tension inflection point dense segment includes a tension peak grouping label, a segment density statistic, a dense grouping index number, and a response center time interval; the polyester blended fabric process traceability node registration table includes a risk jump sequence segment position, a jump sequence level score value, a registered process node number, and a traceability level classification label.

[0013] As a further solution of the present invention, the steps for obtaining the start-stop tension batch jump index table are specifically as follows:

[0014] S111: Obtain the start and stop time sequence, material batch sequence, and tension feedback timestamp sequence of continuous process nodes on the polyester blended spinning production line, arrange the start and stop time of each node according to the process flow sequence, calculate the intervals between the start and stop times of adjacent nodes in sequence, and determine whether the adjacent intervals are zero, thereby obtaining a set of nodes with adjacent zero intervals in the start and stop time;

[0015] S112: Divide the tension feedback timestamp sequence according to the time period corresponding to each node in the set of adjacent zero-interval nodes of the start and stop time, extract the tension feedback time segments corresponding to the start and stop time, extract the material batch sequence within the corresponding segments, compare the numbers of the material batch sequences in order, determine whether there is a number jump in the batch sequence, and obtain a list of batch jump nodes in the tension time segments;

[0016] S113: Call the batch jump node list in the tension time segment and the set of zero-interval nodes with adjacent start and stop times to perform cross-judgment, filter the node indexes with zero start and stop times and discontinuous material batches, and generate a start and stop tension batch jump index table.

[0017] As a further solution of the present invention, the steps for obtaining the fiber level and torque mutation position set are specifically as follows:

[0018] S211: extracting the fiber axial arrangement image sequence, adjacent layer spacing value sequence, and torque direction symbol sequence corresponding to the drafting section and twisting section process sections according to the time period position recorded in the start-stop tension batch jump index table, performing time synchronization processing, extracting the matching structure data segment, and generating a structure sequence synchronization data set;

[0019] S212: Based on the structure sequence synchronization data set, the structure sequence in the jump section is set as a fixed detection window of equal width, using the formula:

[0020] ;

[0021] Calculating the detection window Normalized slope rate of change , using the rate of change as an indicator to construct a standard slope change trend value sequence within the window;

[0022] in, Indicates the The change in the segment level spacing, Indicates the The time variation corresponding to the segment, Indicates the benchmark amplitude of the inter-segment level spacing, The base amplitude representing the time difference of the entire segment, which represents the number of spacing point pairs in the window;

[0023] S213: Call each window segment in the standard slope change trend value sequence, determine whether the window segment change trend is discontinuous fluctuation, and detect whether there is a reverse flip mark in the corresponding torque direction symbol sequence, filter out positions where both direction inconsistency and symbol flip exist, mark the filtered window positions, and obtain a set of fiber level and torque mutation positions.

[0024] As a further solution of the present invention, the steps for obtaining the tension fluctuation synchronization section are specifically as follows:

[0025] S311: Based on the marked mutation time periods in the fiber layer and torque mutation position set, extract the fiber arrangement trajectory path sequence and the tension fluctuation value sequence within the corresponding time period, calculate the displacement amplitude difference per unit time in the axial direction of the trajectory path, and count the amplitude change frequency within each time segment. At the same time, perform sign recognition on the tension fluctuation sequence, count the positive and negative value switching frequency within the same time period, and obtain a structure and tension change frequency data comparison table;

[0026] S312: Call the trajectory change frequency and tension switching frequency in the structure and tension change frequency data comparison table, divide the time axis into segments and set fixed intervals, judge for each time slice whether the two types of frequencies exceed their respective corresponding average reference values ​​at the same time, filter out overlapping segments where the fluctuation behavior appears at the same time, summarize them in chronological order, and establish tension fluctuation synchronization segments.

[0027] As a further solution of the present invention, the steps for obtaining the tension inflection point dense section are specifically as follows:

[0028] S411: Based on the time segments in the tension fluctuation synchronization segment, extract the corresponding tension change sequence, locate all local extreme points in the sequence, divide the tension sequence into continuous segments according to a fixed time slice width, count the number of inflection points in each time segment, and generate a tension inflection point number distribution table;

[0029] S412: Call the inflection point quantity value, time span, tension normalized fluctuation amplitude and extreme value change rate of each segment in the tension inflection point quantity distribution table, perform time dimension normalization and unit amplitude unification processing, using the formula:

[0030] ;

[0031] Calculate the current The density of inflection points in each time period , get the tension inflection point density sequence;

[0032] in, Indicates the number of inflection points in the current segment, Indicates the time length corresponding to the current segment, Indicates the standard deviation of the unit time fluctuation of the tension sequence in the current segment, Indicates the extreme tension difference of the current section. Indicates the previous time period The number of inflection points, Indicates the previous time period The extreme difference of tension;

[0033] S413: Based on the density in the tension inflection point density sequence, all segments are hierarchically clustered by density, the group to which the density peak belongs is identified, and it is determined on the time axis whether they constitute a continuous segment fragment. If the density values ​​in the continuous segment are all higher than the set density judgment threshold, the corresponding segment position is recorded to obtain the tension inflection point dense segment.

[0034] As a further solution of the present invention, the steps for obtaining the polyester blended fabric process traceability node registration table are specifically as follows:

[0035] S511: Match the equipment start and stop status records according to the inflection point peak section in the tension inflection point dense section, extract the start and stop records of the equipment corresponding to the section at the time points before and after the section, calculate the time interval between the previous start and stop and the next start and stop, and count the number of equipment nodes skipped in the current process section. Determine whether the number of skipped nodes exceeds the skipping number threshold and whether the start and stop time interval is within the preset start and stop abnormal section range. Filter the process sections that meet the requirements as abnormal process sections and generate a risky skipping process section list.

[0036] S512: Call the number of skip nodes and start-stop time interval data of each process segment in the risk skip process segment list, using the formula:

[0037] ;

[0038] Calculation process section Skip-rank rating , integration obtains the skip-order grade score sequence;

[0039] in, Indicates the number of skip nodes in the target process segment, Indicates the maximum number of skip nodes in all process segments. Indicates the start and stop time interval of the target process section, Indicates the average value of the start and stop time intervals of all process sections;

[0040] S513: According to the score of each process section in the skipping grade scoring sequence, a skipping risk scoring threshold is set, all scoring values ​​are screened, and process section nodes with scores higher than the scoring threshold are retained. The process number, scoring value and corresponding time and location information of the node are combined and registered, and the structural path of the blended fabric traceability system is connected to establish a polyester blended fabric process traceability node registration table.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are:

[0042] In the present invention, by synchronously analyzing the start and stop status of process nodes and tension feedback, a jump index table is constructed, which can accurately identify batch interruption points and achieve precise positioning of abnormal process sections. Combining the time synchronization analysis of multi-point slope trends in tension feedback and structural images, the synergistic relationship between fiber arrangement and torque change can be revealed, and potential abnormal windows can be quickly marked. With the help of the corresponding analysis of tension fluctuations and structural disturbance frequencies, the screening of behavioral overlap periods is achieved, which helps to trace the causes of abnormal fluctuations. Hierarchical clustering of dense tension inflection point areas and quantification of inflection point distribution density can highlight high-risk process sections in dynamic tension fluctuation sequences. Through the linkage analysis of start and stop records and jump sequence nodes, combined with the dual indicators of jump sequence grade score and start and stop interval, high-risk nodes can be extracted and registration and traceability can be completed. This processing logic integrates multi-dimensional cross-comparison between timeline synchronization, structural dynamic analysis and equipment start-up and shutdown behaviors, enabling the correlation between microscopic changes and macroscopic control behaviors in the textile process, thereby achieving accurate tracing of abnormal phenomena and marking of responsible nodes, significantly improving the granularity and response efficiency of the traceability system, reducing abnormality identification delays, and increasing information integration density and decision-making response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a schematic diagram of the main steps of the present invention;

[0044] Figure 2 This is a flow chart of step S1 of the present invention;

[0045] Figure 3 This is a flow chart of step S2 of the present invention;

[0046] Figure 4 This is a flow chart of step S3 of the present invention;

[0047] Figure 5 This is a flow chart of step S4 of the present invention;

[0048] Figure 6 This is a flow chart of step S5 of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0050] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0051] See also Figure 1 The present invention provides a technical solution: a polyester blended fabric production process traceability information management method, comprising the following steps:

[0052] S1: Obtain the continuous process nodes on the polyester blended production line, capture the node indexes of discontinuous node batch sequences, and establish the start-stop tension batch jump index table;

[0053] S2: According to the start-stop tension batch jump index table, the jump section is set as a fixed detection window, the arrangement level spacing slope of multiple consecutive points in the window is calculated, and the window of the mutation position is marked according to the direction of the multi-point slope to obtain the fiber level and torque mutation position set;

[0054] S3: Based on the marked mutation position time period of the fiber layer and torque mutation position set, the amplitude change frequency in each time segment is counted, and the interactive response area is selected according to the change frequency to generate the tension fluctuation synchronization segment;

[0055] S4: Based on the recorded time segments in the tension fluctuation synchronization segment and the location of the inflection point, the density of the inflection points is calculated, and the density of all segments is grouped and clustered to generate a tension inflection point dense segment;

[0056] S5: Based on the inflection point peak section in the tension inflection point dense section, match the start and stop status records of the corresponding equipment, and establish a polyester blended fabric process traceability node registration table by screening the risk sections and corresponding process nodes in the start and stop status records;

[0057] The start-stop tension batch jump index table includes the jump node position index, the corresponding tension feedback index segment, the batch continuity interruption mark, and the start-stop time interval status mark. The fiber layer and torque mutation position set includes the mutation detection window position, the layer spacing slope change direction, the torque reversal mark position, and the mutation node identification number. The tension fluctuation synchronization segment is specifically the trajectory disturbance amplitude distribution segment, the tension switching frequency segment, the frequency overlap mark interval, and the synchronization response identification number. The tension inflection point dense segment includes the tension peak grouping label, the segment density statistic, the dense grouping index number, and the response center time interval. The polyester blended fabric process traceability node registration table includes the risk jump sequence segment position, jump sequence level score value, registered process node number, and traceability level classification label.

[0058] See also Figure 2 The specific steps for obtaining the start-stop tension batch jump index table are as follows:

[0059] S111: Obtain the start and stop time sequence, material batch sequence, and tension feedback timestamp sequence of continuous process nodes on the polyester blended spinning production line, arrange the start and stop time of each node according to the process flow sequence, calculate the intervals between the start and stop times of adjacent nodes in sequence, and determine whether the adjacent intervals are zero, thereby obtaining a set of nodes with adjacent zero intervals in the start and stop time;

[0060] To obtain the start-stop time series, material batch series, and tension feedback timestamp series of continuous process nodes on a polyester blended spinning production line, it is necessary to record the operating status of each process equipment in the drawing, drafting, and twisting sections based on the equipment data acquisition system. Assuming that the drawing, drafting, and twisting sections each have three nodes, the collected start-stop times are as follows: node A1 starts at 08:00 and ends at 08:05, node A2 starts at 08:05 and ends at 08:10, and node A3 starts at 08:10 and ends at 08:15. Then, by comparing the start-stop time differences of adjacent nodes: the start-stop time of A2 minus the start-stop time of A1 is 0 minutes, which indicates that the start-stop times overlap. Here, the start and stop time intervals of adjacent nodes need to be calculated in order and expressed in minutes. During the execution process, the difference between the start and stop times of each pair of adjacent nodes is calculated. If the difference is 0, it is marked as the start and stop times are adjacent to zero. After collecting the time difference results between multiple nodes, the start and stop interval status between each node needs to be marked in a Boolean manner. For example, A1-A2 is TRUE and A2-A3 is FALSE, which means that A1 and A2 have a zero interval, while A2 and A3 have a non-zero interval. Finally, all TRUE state nodes are marked and the node number is output to obtain the set of nodes with adjacent zero intervals in start and stop time.

[0061] S112: Divide the tension feedback timestamp sequence according to the time period corresponding to each node in the set of adjacent zero-interval nodes of the start and stop time, extract the tension feedback time segments corresponding to the start and stop time, extract the material batch sequence within the corresponding segments, compare the numbers of the material batch sequences in order, determine whether there is a number jump in the batch sequence, and obtain a list of batch jump nodes in the tension time segments;

[0062] The tension feedback timestamp sequence is divided into multiple subsegments based on the time range of each pair of nodes in the set of nodes with adjacent zero start and stop times. The matching principle is that the tension data time range must cover the start and end times of the corresponding start and stop periods. For example, if the start and stop times of nodes A1-A2 are 08:00-08:10, the data corresponding to the 08:00 to 08:10 period in the tension sequence is extracted. Tension feedback data can be collected at a frequency of once per second, resulting in a total of 10 minutes x 60 seconds = 600 sampling points within this period. The batch numbers of the materials within this period are also extracted. For example, if the batch numbers are B1 → B3 → B2, the sequence is numbered. If the sequence number rolls back (for example, B3 is followed by B2) or jumps across segments (for example, B1 is followed by B3), the numbering is determined to be discontinuous. The specific judgment criteria involves converting the batch numbers into a sequence of equally spaced natural numbers and performing a difference calculation. For example, B1→B3→B2 is mapped to 1→3→2. The differences between adjacent items are 2 and -1. Since the differences are not 1, the numbering is considered discontinuous, and the segment is marked as a batch jump. This process is repeated for all time periods with zero start and stop times, ultimately compiling a list of batch jump nodes in the tension time segment.

[0063] S113: Calling the batch jump node list in the tension time segment and the set of nodes with adjacent zero intervals of start and stop time to perform cross-judgment, screening the node indexes with both zero start and stop time and discontinuous material batches, and generating a start and stop tension batch jump index table;

[0064] A cross-screening process is performed on the list of batch transition nodes within the tension time segment and the set of nodes with adjacent start / stop times of zero. Nodes meeting two conditions are extracted: a zero start / stop interval and discontinuous batch sequence numbers. For example, if the start / stop interval between nodes A2 and A3 is zero, and the batch number they process is B3→B1, the natural sequence is 3→1, with a difference of -2, indicating a valid transition. This node is then marked as a valid abnormal node. The node numbers that meet these two conditions are then collected, chronologically numbered, and indexed. The output should include fields such as the start and end time periods, material batch numbers, number difference, and whether each transition node is a transition flag. The resulting index table contains entries such as: node A2, time 08:05-08:10, batch sequence B3→B1, number difference of -2, and transition flag of 1. After screening all nodes, a set of abnormal node index data is generated, forming the start / stop tension batch transition index table.

[0065] See also Figure 3 , the specific steps for obtaining the fiber level and torque mutation position set are:

[0066] S211: Extract the fiber axial arrangement image sequence, adjacent layer spacing value sequence, and torque direction symbol sequence corresponding to the drafting section and twisting section process sections according to the time period position recorded in the start-stop tension batch jump index table, perform time synchronization processing, extract the matching structure data segment, and generate a structure sequence synchronization data set;

[0067] Extract the fiber axial arrangement image sequence, layer spacing value sequence and torque direction symbol sequence in the drafting section and twisting section process section. The image sequence is collected in real time by the photoelectric imaging device in the process site at a high frame rate (for example, 20 frames per second) of the production section. The size of each frame image is set to Pixels, record the axial arrangement of fibers, obtain the fiber arrangement trajectory contour by image edge extraction, use the gradient edge grayscale difference maximum method to determine the boundary points between adjacent layers in the image, extract the pixel distance between adjacent layers in each frame of the image, and convert it into millimeters to form a layer spacing sequence. , if the coordinates of the four boundary points extracted are , then the spacing value sequence is Pixels, corresponding to a physical spacing of The torque direction is derived from the fiber rotation direction in the image, that is, the curvature direction extraction algorithm is used on the trajectory curve to obtain the positive and negative polarity of the torque rotation in each frame. If the positive and negative signs alternate in consecutive frames, such as frame is positive (+1), frame If it is negative (-1), it is judged that a direction reversal has occurred. Finally, the image data of this time period is converted into: a fiber structure trajectory image sequence, a layer spacing numerical sequence, and a torque direction symbol sequence, and integrated into one under the same time index to obtain a structural parameter extraction result set.

[0068] S212: Based on the structure sequence synchronization data set, the structure sequence in the jump segment is set as a fixed detection window of equal width, using the formula:

[0069] ;

[0070] Calculating the detection window Normalized slope rate of change , using the rate of change as an indicator to construct a standard slope change trend value sequence within the window;

[0071] in, Indicates the The change in the segment level spacing, Indicates the The time variation corresponding to the segment, Indicates the benchmark amplitude of the inter-segment level spacing, Indicates the reference amplitude of the time difference of the entire segment, Indicates the number of spacing point pairs in the window.

[0072] Set the jump segment as a fixed detection window, for example, each window length is 0.5 seconds, corresponding to 10 frames of image frames, from which the three consecutive point spacing sequences and corresponding timestamp data are extracted. The original spacing unit is millimeter and the time unit is second. To eliminate the influence of unit differences, all participating items need to be normalized. Assume that the three consecutive level spacing changes in a window are mm, corresponding to the time change of seconds, the base level spacing of the acquisition section is Millimeters, time base is seconds, then the normalized slope rate of change The calculation is done using the following formula:

[0073] ;

[0074] The data is calculated as follows:

[0075] ;

[0076] It represents the intensity of the change in fiber-level spacing per unit time within a certain detection window. Its physical meaning is the rate of change of the amplitude of the fiber structure disturbance within the window. A larger value indicates a more drastic change in the fiber arrangement within the window and a more unstable structure. This indicator comprehensively considers the relative proportional relationship between the spacing change and the sampling time resolution, and uniformly measures the degree of disturbance between different windows under a unit normalized scale. This value can be used as a direct basis for judging the significance of structural disturbances, and provides a quantitative basis for the direction and amplitude of structural fluctuations for subsequent abnormal node identification and mutation segment screening, playing a key role in trend identification.

[0077] Repeat the above calculation process for all detection windows to form a sequence of normalized slope change values ​​corresponding to each window. This sequence serves as an indicator of perturbation within the segment, used to determine the intensity of dynamic perturbation behavior in each window and, in turn, generate a sequence of standard slope change trend values. The formula is beneficial because, by introducing a normalized denominator, it reduces data of different scales to a unified scale, improving the stability and accuracy of cross-segment perturbation comparisons.

[0078] S213: Calling each window segment in the standard slope change trend value sequence, determining whether the window segment change trend presents discontinuous fluctuations, and detecting whether there is a reverse flip sign in the corresponding torque direction symbol sequence, screening positions where both direction inconsistency and sign flipping occur, marking the screened window positions, and obtaining a set of fiber level and torque mutation positions;

[0079] Call each window segment in the standard slope change trend value sequence to determine whether the slope change direction is continuous and consistent in the three-point sequence. Specifically, if the three-segment normalized slope change rate sequence is , indicating that the direction changes from rising to falling and then rising again, which is inconsistent. Then match the symbol change of the window at the corresponding time point in the torque direction symbol sequence. If the symbol of this segment is , indicating that a reversal has occurred. Windows that meet these two conditions are called mutation windows and are marked. For example, in the detection section, if the direction of window 5 is inconsistent and torque reversal also occurs, then window number 5 is added to the mark set, and the final output is the window index that meets the double judgment, forming a set of fiber level and torque mutation positions. This result is used to identify discontinuous nodes in the fiber structure and match them with the tension fluctuation logic chain. This result shows that the synchronous disturbance window has been screened out from the two aspects of slope change and torque reversal, which can be directly used for the subsequent identification of structural misalignment behavior.

[0080] See also Figure 4 The specific steps for obtaining the tension fluctuation synchronization section are as follows:

[0081] S311: Based on the marked mutation time periods in the fiber layer and torque mutation location set, the fiber arrangement trajectory path sequence and the tension fluctuation value sequence within the corresponding time period are extracted. The displacement amplitude difference per unit time in the axial direction of the trajectory path is calculated, and the frequency of amplitude changes within each time segment is counted. At the same time, the tension fluctuation sequence is sign-recognized, and the frequency of positive and negative value switching within the same time period is counted to obtain a data comparison table of structure and tension change frequency.

[0082] First, the fiber arrangement trajectory path sequence and tension fluctuation value sequence corresponding to the time period are extracted. The fiber arrangement trajectory path can be identified by the pixel displacement of continuous edge points in the structural image in the axial direction and arranged in chronological order. The tension fluctuation value sequence is collected by the tension sensors of the drawing section and the twisting section at the corresponding timestamps. The unit time displacement of the structural trajectory in the axial direction can be expressed as the change in the pixel center of gravity of the two previous and next image frames divided by the frame time interval. For example, if the center positions of the two previous and next frames are 145 pixels and 151 pixels respectively, and the frame time interval is 0.04 seconds, the displacement rate is (151-145) / 0.04=150px / s. The identification of the positive and negative switching frequency of tension fluctuations requires segmenting the continuous values ​​in the tension value sequence into positive and negative values ​​in a sliding window manner. For the transformation segment, if a tension sequence is [18, 20, -15, -17, 14, -12], there will be three positive and negative reversals at 18→-15, -17→14, and 14→-12, respectively. Therefore, the tension fluctuation frequency in this time period is 3 times. After completing the statistics of the trajectory displacement frequency and the tension positive and negative switching frequency, the two are aligned with the unified time base. By judging whether the trajectory disturbance frequency and the tension switching frequency increase at the same time in the same segment, it is determined whether they constitute a synchronous response. In the example, if the trajectory displacement frequency of a segment exceeds 5 times / second and the tension fluctuation frequency exceeds 3 times / second, and both appear in the time interval of 2 to 4 seconds, then the segment is considered a synchronous response segment. All segments that meet the above conditions will be collectively output as the final result tension fluctuation synchronization segment.

[0083] S312: Calling the trajectory change frequency and tension switching frequency in the structure and tension change frequency data comparison table, the time axis is segmented and fixed intervals are set. For each time slice, it is determined whether the two types of frequencies simultaneously exceed their respective corresponding average reference values. The overlapping segments of the fluctuation behavior that appear at the same time position are selected, and the segments are summarized in chronological order to establish the tension fluctuation synchronization segment;

[0084] The call structure and tension change frequency data comparison table are used to divide the time axis into equal intervals, each segment is 5 seconds. The trajectory displacement change frequency and tension sign switching frequency are extracted for each time window. The trajectory change frequency baseline value is set to 12 times / window, and the tension switching frequency baseline value is 2 times / window. These values ​​are obtained by averaging normal fluctuation behavior under 20 sample windows. Taking a single sample as an example, the trajectory change frequency in window t1 is 16 times and the tension switching frequency is 3 times, both exceeding the baseline values. Therefore, window t1 is determined to be a synchronous fluctuation segment. Subsequently, the segments that meet this condition simultaneously in all time periods are continuously marked. If two or more consecutive segments are synchronous segments, the segment is marked as an overlapping response segment. Finally, an index table structure is established with time index as the unit. For example, time windows t1-t3 and t7-t8 are overlapping response segments, indicating that there is synchronous amplification behavior between trajectory disturbance and tension change in these segments, resulting in a tension fluctuation synchronous segment.

[0085] See also Figure 5 The specific steps for obtaining the dense tension inflection point section are as follows:

[0086] S411: Based on the time segments in the tension fluctuation synchronization segment, extract the corresponding tension change sequence, locate all local extreme points in the sequence, divide the tension sequence into continuous segments based on a fixed time slice width, count the number of inflection points in each time segment, and generate a tension inflection point number distribution table;

[0087] Based on the time segment recorded in the tension fluctuation synchronization segment, the corresponding tension change signal needs to be synchronously expanded according to the time axis in actual application. Combined with the original data sequence output by the tension sensor, the tension signal is continuously sampled with a millisecond resolution. For example, by monitoring the tension for 30 consecutive seconds at a frequency of 100Hz, 3000 tension sampling points can be obtained. The fluctuation direction between each point in the sequence and the adjacent points is judged by first-order difference calculation, and then the position of the extreme point is determined. When the tension value changes from increasing to decreasing or from decreasing to increasing and meets the set extreme value threshold change amplitude (such as the change amplitude exceeds 1.5N), it can be marked as an inflection point. Suppose a certain The segment tension sampling sequence is: {52,54,56,55,53,51,53,56,59,61,60,58,55}, among which the 4th bit (56→55) and the 12th bit (61→60) can be used as peaks, and the inflection points are then determined to be 4 and 12. After the same rule detection is performed on the entire 30-second sequence, the detected inflection points are assigned to each 10-second time segment. Assuming that there are 6 inflection points in the first segment, 9 in the second segment, 5 in the third segment, and 12 in the fourth segment, this operation process can be embedded in production monitoring to reflect the distribution of tension disturbance points in real time and generate a tension inflection point quantity distribution table.

[0088] S412: Call the inflection point quantity value, time span, tension normalized fluctuation amplitude and extreme value change rate of each segment in the tension inflection point quantity distribution table, perform time dimension normalization and unit amplitude unification processing, using the formula:

[0089] ;

[0090] Calculate the current The density of inflection points in each time period , get the tension inflection point density sequence;

[0091] in, Indicates the number of inflection points in the current segment, Indicates the time length corresponding to the current segment, Indicates the standard deviation of the unit time fluctuation of the tension sequence in the current segment, Indicates the extreme tension difference of the current section. Indicates the previous time period The number of inflection points, Indicates the previous time period The tension extreme value difference, all parameters are unified in unit time to ensure the overall dimensional consistency, density value As a quantitative indicator of the degree of segment disturbance aggregation.

[0092] Call the inflection point values ​​and tension characteristic fluctuation values ​​in the tension inflection point quantity distribution table, and normalize the indicators in all sections. , tension extreme difference , tension variance , time span Substitute into the formula to calculate the inflection point density. The calculation steps are as follows: Taking the second section as an example, the number of inflection points , time length , tension extreme difference , the previous segment , standard deviation of tension fluctuation , the number of inflection points in the previous segment , according to the formula

[0093] ;

[0094] Get tension density , the result shows that there is intensive disturbance behavior in this section per unit time. If this value is compared with the density judgment benchmark value of 2.8, it can be confirmed that it meets the judgment conditions. The parameters are explained as follows: is the number of inflection points in the current section, obtained from the inflection point statistics process, The time span of the current segment is set to 10 seconds. is the square value of the tension standard deviation within the segment, which is calculated by taking the square of the sample standard deviation. is the difference between the maximum and minimum values ​​in the tension sequence, is the difference between the ranges of two adjacent segments, is the difference in the number of inflection points in the two segments, all of which are normalized by time to express the intensity of disturbance fluctuations. The benefit of this formula is that it can three-dimensionally characterize short-term disturbances by integrating the tension fluctuation amplitude and the adjacent change rate, and generate a tension inflection point density sequence.

[0095] S413: Based on the density in the tension inflection point density sequence, all segments are hierarchically clustered by density, the group to which the density peak belongs is identified, and a determination is made on the time axis whether they constitute a continuous segment fragment. If the density values ​​in the continuous segment are all higher than the set density determination threshold, the corresponding segment position is recorded to obtain the tension inflection point dense segment;

[0096] According to the values ​​obtained in the tension inflection point density sequence, the The values ​​are clustered and identified. First, they are arranged in ascending order according to the density values ​​and then divided into three groups: low density group, medium density group, and high density group. Among them, the low density group is a segment with a density value between 0.0 and 1.5. This range reflects the situation that the number of disturbance inflection points per unit time is relatively sparse or the tension fluctuation changes slowly, which often corresponds to the period when the tension operation of the production process is stable or the interference is weak. The medium density group corresponds to the segment with a density value greater than or equal to 1.5 and less than or equal to 2.8. This interval is the transition area from the stable disturbance state to the significant disturbance state. The high density group is a segment with a density value greater than 2.8. The tension disturbance in this group has the characteristics of frequent inflection point changes and violent tension fluctuations. For example, if the second segment calculated in the above example is 3.115, it is classified into the high density group. Then, all the time-continuous segments in the high density group are combined to determine whether they are continuous on the time axis and whether the density meets the set density threshold greater than 2.8. Segments 2 and 3 are set. , then the two constitute a set of continuous segments, and the time start and end range is recorded as 20~40 seconds. Finally, the segment position is extracted as the dense disturbance response area, marked as the core focus area, and the tension inflection point dense segment is obtained.

[0097] See also Figure 6 The specific steps for obtaining the polyester blended fabric process traceability node registration form are as follows:

[0098] S511: Match the equipment start / stop status records based on the inflection point peak segments in the tension inflection point dense segment, extract the start / stop records of the equipment corresponding to the segment at the time points before and after the segment, calculate the time interval between the previous start / stop and the next start / stop, and count the number of equipment nodes skipped in the current process segment. Determine whether the number of skipped nodes exceeds the skipping threshold and whether the start / stop time interval is within the preset start / stop abnormal segment range. Filter the process segments that meet the requirements as abnormal process segments and generate a risky skipping process segment list.

[0099] First, the time interval covered by the inflection point dense segment is retrieved in sequence according to the timestamp, and the log data of the operating status of all equipment in the time period is split according to the equipment number to obtain the start and stop time points of each equipment before and after the segment. For example, the segment occurs between 13:00 and 13:45 on June 13, 2024. Equipment A stops running at 12:55 and is reactivated at 13:47. The start and stop time interval is 52 minutes. This process corresponds to the start and stop time calculation link. Subsequently, the equipment nodes that do not appear in the start and stop records within the segment time range are identified, and the number of nodes skipped from the logical order is counted. For example, if the process segment should pass through nodes A→B→C→D→E in sequence, but only nodes A and E appear in the start and stop records, the number of skipped nodes is 3. , respectively B, C, and D. If the number of skipped nodes exceeds the threshold and the start-stop time interval is within the abnormal section (for example, the preset abnormal start-stop interval is set to be greater than 30 minutes), the current process section is judged to constitute a risky sequence skipping section. The threshold for the number of skipped nodes can be set to 2 based on the sample statistical median value. If the current number of skipped nodes is 3, the condition is met and it is judged to be a sequence skipping section. The setting process of the start-stop abnormal section is obtained by analyzing the average time of all start-stop samples. Assuming that the average value of all start-stop interval samples is 25 minutes and the standard deviation is 6 minutes, the abnormal start-stop section is defined as the average value plus 2 times the standard deviation, that is, greater than 37 minutes is abnormal. The 52 minutes here is significantly higher than this interval, so it meets the abnormal section condition. Finally, the section is judged to be a sequence skipping section and enters the sequence skipping process section list.

[0100] S512: Call the number of jump nodes and start-stop time interval data of each process segment in the risk jump process segment list, using the formula:

[0101] ;

[0102] Calculation process section Skip-rank rating , integration obtains the skip-order grade score sequence;

[0103] in, Indicates the number of jump nodes in this process segment. Indicates the maximum number of skip nodes in all process segments. Indicates the start and stop time interval of the process section. It represents the average value of the start and stop time intervals of all process sections.

[0104] First, the number of skip nodes for each jump sequence segment obtained by calling the previous substep and start-stop interval time ,Will By the maximum number of hopping nodes in all hopping segments For normalization, assuming that the maximum number of jumps in all process segments is 4, then if the current process segment skips 3 nodes, the normalized number is , and then set the start and stop time interval of each segment Average start-stop time interval with all segments For comparison, assuming minutes, the start and stop interval of the current process section is 52 minutes, then the deviation ratio is , put it into the formula: ; The parameters are explained as follows: For the scoring results, is the number of hop nodes (currently 3), , , The calculation result is a score of 1.39275, which is higher than the preset sequence jump risk score threshold (for example, the threshold is set to 1.2). This score constitutes a high-level risk score and enters the high-risk node set. The benefit of the formula is that it effectively integrates the two risk factors into the same dimension by standardizing the number of sequence jumps and introducing the start-stop offset ratio to form a product score, which is conducive to identifying abnormal sequence jump process sections in the scoring system.

[0105] S513: Based on the score of each process segment in the skipping grade scoring sequence, a skipping risk scoring threshold is set, all scoring values ​​are screened, and process segment nodes with scores higher than the scoring threshold are retained. The process number, scoring value, and corresponding time and location information of the node are combined and registered, and the structural path of the blended fabric traceability system is connected to establish a polyester blended fabric process traceability node registration table;

[0106] Based on the score value, all process sections are sorted from high to low according to the score, and the process sections are classified according to the set jump risk score threshold. If the score value of a process section exceeds the threshold of 1.2, it is judged as a jump risk node. The jump grade score sequence is used as input, and the score value and process section number are read one by one. The node information with a score higher than 1.2 is extracted. For example, the process section numbered A102 and scored 1.39 enters the registration range. At the same time, the time period when the jump occurs is extracted, such as 13:00 to 13:45 on June 13, 2024. The registration record structure is constructed, including four items: process section number, time interval, jump grade score, number of jump nodes, and start-stop interval. After completion, all information is filled in the blending process traceability registration table, which exists as a structural positioning information item in the subsequent process backtracking.

[0107] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A polyester blended fabric production process traceability information management method, characterized in that: The following steps are involved: S1: Obtain the continuous process nodes on the polyester blended production line, capture the node indexes of discontinuous node batch sequences, and establish the start-stop tension batch jump index table; The steps for obtaining the start-stop tension batch jump index table are specifically as follows: S111: Obtain the start and stop time sequence, material batch sequence, and tension feedback timestamp sequence of continuous process nodes on the polyester blended spinning production line, arrange the start and stop time of each node according to the process flow sequence, calculate the intervals between the start and stop times of adjacent nodes in sequence, and determine whether the adjacent intervals are zero, thereby obtaining a set of nodes with adjacent zero intervals in the start and stop time; S112: Divide the tension feedback timestamp sequence according to the time period corresponding to each node in the set of adjacent zero-interval nodes of the start and stop time, extract the tension feedback time segments corresponding to the start and stop time, extract the material batch sequence within the corresponding segments, compare the numbers of the material batch sequences in order, determine whether there is a number jump in the batch sequence, and obtain a list of batch jump nodes in the tension time segments; S113: Calling the batch jump node list in the tension time segment and the set of nodes with adjacent zero intervals of start and stop time to perform cross-judgment, screening the node indexes with both zero start and stop time and discontinuous material batches, and generating a start and stop tension batch jump index table; S2: According to the start-stop tension batch jump index table, the jump section is set as a fixed detection window, the arrangement level spacing slope of multiple consecutive points in the window is calculated, and the window of the mutation position is marked according to the direction of the multi-point slope to obtain the fiber level and torque mutation position set; S3: Based on the marked mutation position time period of the fiber layer and the torque mutation position set, the amplitude change frequency in each time segment is counted, and the interactive response area is selected according to the change frequency to generate the tension fluctuation synchronization segment; S4: Based on the recorded time segments in the tension fluctuation synchronization segment and the locations of the inflection points, the density of the inflection points is calculated, and the density of all segments is grouped and clustered to generate a tension inflection point dense segment; S5: According to the inflection point peak section in the tension inflection point dense section, the start and stop status records of the corresponding equipment are matched, and by screening the risk sections and corresponding process nodes in the start and stop status records, a polyester blended fabric process traceability node registration table is established.

2. The polyester blended fabric production process traceability information management method according to claim 1, characterized in that: The start-stop tension batch jump index table includes the jump node position index, the corresponding tension feedback index segment, the batch continuity interruption mark, and the start-stop time interval status mark. The fiber layer and torque mutation position set includes the mutation detection window position, the layer spacing slope change direction, the torque reversal mark bit, and the mutation node identification number. The tension fluctuation synchronization segment is specifically the trajectory disturbance amplitude distribution segment, the tension switching frequency segment, the frequency overlap mark interval, and the synchronization response identification number. The tension inflection point dense segment includes the tension peak grouping label, the segment density statistic, the dense grouping index number, and the response center time interval. The polyester blended fabric process traceability node registration table includes the risk jump sequence segment position, the jump sequence level score value, the registered process node number, and the traceability level classification label.

3. The polyester blended fabric production process traceability information management method according to claim 1, characterized in that: The steps for obtaining the fiber level and torque mutation position set are specifically as follows: S211: extracting the fiber axial arrangement image sequence, adjacent layer spacing value sequence, and torque direction symbol sequence corresponding to the drafting section and twisting section process sections according to the time period position recorded in the start-stop tension batch jump index table, performing time synchronization processing, extracting the matching structure data segment, and generating a structure sequence synchronization data set; S212: Based on the structure sequence synchronization data set, the structure sequence in the jump section is set as a fixed detection window of equal width, using the formula: ； Calculating the detection window Normalized slope rate of change , using the rate of change as an indicator to construct a standard slope change trend value sequence within the window; in, Indicates the The change in the segment level spacing, Indicates the The time variation corresponding to the segment, Indicates the benchmark amplitude of the inter-segment level spacing, Indicates the reference amplitude of the time difference of the entire segment, Indicates the number of spacing point pairs in the window; S213: Call each window segment in the standard slope change trend value sequence, determine whether the window segment change trend is discontinuous fluctuation, and detect whether there is a reverse flip mark in the corresponding torque direction symbol sequence, filter out positions where both direction inconsistency and symbol flip exist, mark the filtered window positions, and obtain a set of fiber level and torque mutation positions.

4. The polyester blended fabric production process traceability information management method according to claim 3, characterized in that: The steps for obtaining the tension fluctuation synchronization section are specifically as follows: S311: Based on the marked mutation time periods in the fiber layer and torque mutation position set, extract the fiber arrangement trajectory path sequence and the tension fluctuation value sequence within the corresponding time period, calculate the displacement amplitude difference per unit time in the axial direction of the trajectory path, and count the amplitude change frequency within each time segment. At the same time, perform sign recognition on the tension fluctuation sequence, count the positive and negative value switching frequency within the same time period, and obtain a structure and tension change frequency data comparison table; S312: Call the trajectory change frequency and tension switching frequency in the structure and tension change frequency data comparison table, divide the time axis into segments and set fixed intervals, judge for each time slice whether the two types of frequencies exceed their respective corresponding average reference values ​​at the same time, filter out overlapping segments where the fluctuation behavior appears at the same time, summarize them in chronological order, and establish tension fluctuation synchronization segments.

5. The polyester blended fabric production process traceability information management method according to claim 4, characterized in that: The steps for obtaining the tension inflection point dense section are specifically as follows: S411: Based on the time segments in the tension fluctuation synchronization segment, extract the corresponding tension change sequence, locate all local extreme points in the sequence, divide the tension sequence into continuous segments according to a fixed time slice width, count the number of inflection points in each time segment, and generate a tension inflection point number distribution table; S412: Call the inflection point quantity value, time span, tension normalized fluctuation amplitude and extreme value change rate of each segment in the tension inflection point quantity distribution table, perform time dimension normalization and unit amplitude unification processing, using the formula: ； Calculate the current The density of inflection points in each time period , get the tension inflection point density sequence; in, Indicates the number of inflection points in the current segment, Indicates the time length corresponding to the current segment, Indicates the standard deviation of the unit time fluctuation of the tension sequence in the current segment, Indicates the extreme tension difference of the current section. Indicates the previous time period The number of inflection points, Indicates the previous time period The extreme difference of tension; S413: Based on the density in the tension inflection point density sequence, all segments are hierarchically clustered by density, the group to which the density peak belongs is identified, and it is determined on the time axis whether they constitute a continuous segment fragment. If the density values ​​in the continuous segment are all higher than the set density judgment threshold, the corresponding segment position is recorded to obtain the tension inflection point dense segment.

6. The polyester blended fabric production process traceability information management method according to claim 5, characterized in that: The specific steps for obtaining the polyester blended fabric process traceability node registration form are as follows: S511: Match the equipment start and stop status records according to the inflection point peak section in the tension inflection point dense section, extract the start and stop records of the equipment corresponding to the section at the time points before and after the section, calculate the time interval between the previous start and stop and the next start and stop, and count the number of equipment nodes skipped in the current process section. Determine whether the number of skipped nodes exceeds the skipping number threshold and whether the start and stop time interval is within the preset start and stop abnormal section range. Filter the process sections that meet the requirements as abnormal process sections and generate a risky skipping process section list. S512: Call the number of skip nodes and start-stop time interval data of each process segment in the risk skip process segment list, using the formula: ； Calculation process section Skip-rank rating , integration obtains the skip-order grade score sequence; in, Indicates the number of skip nodes in the target process segment, Indicates the maximum number of skip nodes in all process segments. Indicates the start and stop time interval of the target process section, Indicates the average value of the start and stop time intervals of all process sections; S513: According to the score of each process section in the skipping grade scoring sequence, a skipping risk scoring threshold is set, all scoring values ​​are screened, and process section nodes with scores higher than the scoring threshold are retained. The process number, scoring value and corresponding time and location information of the node are combined and registered, and the structural path of the blended fabric traceability system is connected to establish a polyester blended fabric process traceability node registration table.

Images