An online analysis system for bundled short cut fiber performance detection data

Abstract

The present application relates to the technical field of material detection, and discloses an online analysis system for performance detection data of bundled short-cut fibers, which comprises modules such as fixed-position imaging, profile measurement, index conversion, station correspondence, linkage determination, working condition tracing, critical early warning and review correction, etc. The system extracts end profile, bundle edge, hair boundary, cut profile and sizing gray scale distribution through fiber bundle optical images, converts length dispersion, hair amount, bundle hair looseness, cut neatness and sizing uniformity, and corresponds to cutting speed, traction tension, sizing amount, drying temperature, conveying speed and tool state according to batches and stations, determines linkage change of indexes, determines abnormal working condition sources, gives early warning and outputs correction direction when indexes are not over the limit, adjusts the reference after review, and improves the accuracy of quality analysis and the timeliness of early warning.

Description

Technical Field

[0001] The present invention relates to the technical field of material detection, in particular to an online analysis system for performance detection data of clustered chopped fibers. Background Art

[0002] Clustered chopped fibers usually need to go through processes such as precursor introduction, sizing treatment, traction conveying, drying, cutting, spreading detection and packaging collection during production. Since clustered chopped fibers are not single-fiber products, their product quality not only depends on whether the length of a single fiber meets the requirements, but also is affected by many factors such as the integrity of the fiber bundle, the outward spreading degree of fuzz, the looseness degree of the bundled fiber, the neatness degree of the cut, and the sizing uniformity. There is a strong correlation between the above-mentioned performance indicators. For example, fluctuations in cutting speed may simultaneously cause an increase in length dispersion and cut deflection, abnormal traction tension may simultaneously cause loosening of bundled fibers and an increase in the amount of fuzz, and unstable sizing supply may lead to uneven gray distribution of sizing and further affect the edge stability of the bundle body.

[0003] Existing detection methods for clustered chopped fibers are usually based on manual sampling inspection, offline microscopic observation, individual image measurement or output results from independent detection equipment. Such methods can obtain data on length, appearance, moisture content or local defects, but most systems only judge whether a single indicator is qualified, and lack online judgment on synchronous deviation, continuous deviation and interconnected changes among multiple performance indicators. As a result, when quality abnormalities occur, operators still need to analyze the causes of abnormalities based on experience.

[0004] In view of the above and existing related technologies, there are the following defects: the existing data analysis methods usually focus on the display of detection results and qualification judgment, and only give an alarm when a certain performance indicator has exceeded the limited range, and cannot prompt process correction in advance when the performance indicator has not exceeded the limit but continuous deviation has occurred. Due to the continuity of clustered chopped fiber production, if the abnormal trend cannot be identified in the early stage of quality deterioration, subsequent products of the same batch or adjacent batches are prone to continuous problems such as increased fuzz, loose bundled fibers, cut deflection and uneven sizing, resulting in the expansion of quality fluctuation range. Summary of the Invention

[0005] The technical problem to be solved by the present invention is: aiming at the disadvantage in the prior art that most existing systems only give whether a single indicator is qualified, lack linkage judgment among multiple performance indicators, and cannot perform process early warning and correction in advance when the performance indicators have not exceeded the limit but continuous deviation has occurred. Therefore, we propose an online analysis system for performance detection data of clustered chopped fibers.

[0006] To achieve the above objective, the following technical solution is adopted in the present application: an online analysis system for performance detection data of clustered chopped fibers, comprising:

[0007] An attitude-fixing imaging module is used to enable the bundled chopped fibers to pass through the detection field of view in a set orientation and to acquire optical images of the fiber bundle.

[0008] The contour calculation module is used to extract the chopped fiber end contour, bundle edge, filament expansion boundary, cut contour, and sizing grayscale distribution from fiber bundle optical images.

[0009] The index conversion module is used to convert the extracted results into indices such as length dispersion, filament quantity, bundle looseness, cut neatness, and sizing uniformity.

[0010] The workstation correspondence module is used to map performance indicators to cutting speed, traction tension, sizing supply, drying temperature, conveying speed and tool status according to batch flow and workstation sequence.

[0011] The linkage determination module is used to determine the synchronous offset, continuous offset, and mutual interference changes among multiple performance indicators;

[0012] The working condition tracing module is used to determine the source of the main abnormal working conditions based on the linkage judgment results and the corresponding results of the workstation;

[0013] The critical warning module is used to generate warning information when performance indicators have not exceeded the limit range but have shown a continuous deviation.

[0014] The review and correction module is used to output the direction of process correction and adjust the subsequent judgment criteria based on the changes in the corrected performance indicators.

[0015] Preferably, the attitude-fixed imaging module includes a spreading and feeding unit, a fixed-distance conveying unit, a multi-directional illumination unit, and an image acquisition unit:

[0016] The spreading and feeding unit is used to disperse and spread the bundled chopped fibers into the detection channel;

[0017] A fixed-distance transport unit is used to control the imaging distance between the fiber bundle and the detection lens;

[0018] Multi-directional lighting unit, used to illuminate the fiber ends, bundle edges, filament expansion areas and cut areas respectively;

[0019] The image acquisition unit is used to acquire fiber optic images under corresponding illumination conditions.

[0020] Preferably, the contour calculation module includes a field-of-view calibration unit, a background correction unit, and a boundary extraction unit:

[0021] The field calibration unit is used to determine the conversion relationship between pixel size and actual size;

[0022] Background correction unit, used to correct the background grayscale of the detection channel;

[0023] The boundary extraction unit is used to extract the fiber end contour, bundle edge, filament expansion boundary, cut contour, and sizing grayscale distribution.

[0024] Preferably, the index conversion module uses the ratio of the standard deviation of the distance between the ends of each fiber within the same detection interval to the target cutting length as the length dispersion, the ratio of the area of ​​the filament expansion boundary extending beyond the edge of the bundle to the area of ​​the bundle as the filament quantity, the ratio of the width fluctuation value of the bundle edge to the average width of the bundle as the bundle looseness, the skew angle of the cut contour relative to the set cutting direction as the cut neatness, and the ratio of the gray difference value of the sizing area to the average gray value of the area as the sizing uniformity index.

[0025] Preferably, the workstation corresponding module includes a batch binding unit, a process acquisition unit, and a phase registration unit:

[0026] Batch binding unit is used to bind performance indicators to raw yarn batch, chopped yarn specification, production time and equipment number;

[0027] The process acquisition unit is used to collect process parameters for the corresponding batch.

[0028] The phase registration unit is used to compensate for performance indicators and process parameters based on the fiber transport time from the cutting station to the inspection field of view.

[0029] Preferably, the linkage judgment module uses the average performance index of five consecutive qualified test intervals as the corresponding index benchmark value, compares the performance index of the current test interval with the corresponding index benchmark value, and obtains the offset direction and offset amplitude; when at least two of the length dispersion, filament amount, bundle looseness, cut neatness and sizing uniformity maintain the same offset in three consecutive test intervals, and the offset amplitude increases in each interval, it is determined that multiple performance indicators have linkage changes.

[0030] Preferably, after a linkage change occurs, the working condition tracing module retrieves the cutting speed, traction tension, sizing supply, drying temperature, conveying speed, and tool status corresponding to the time before the start of the linkage change, calculates the change amplitude, duration, and proximity to the start of the linkage change for each working condition parameter, and identifies the working condition parameter with the largest product of change amplitude, duration, and proximity as the main source of abnormal working conditions.

[0031] Preferably, when the performance indicators do not exceed the limit range, if the linkage judgment module determines that there is a linkage change, and the main abnormal operating condition source determined by the operating condition tracing module maintains the same direction of change in three consecutive detection intervals, the critical warning module generates critical warning information. The critical warning information includes abnormal batch, abnormal performance indicators, main abnormal operating condition source, and warning level.

[0032] Preferably, the verification and correction module generates correction directions based on the main sources of abnormal operating conditions; when the main source of abnormal operating conditions is cutting speed, it outputs the cutting speed correction direction; when the main source of abnormal operating conditions is traction tension, it outputs the traction tension correction direction; when the main source of abnormal operating conditions is sizing supply, it outputs the sizing supply correction direction; when the main source of abnormal operating conditions is drying temperature, it outputs the drying temperature correction direction; when the main source of abnormal operating conditions is conveying speed, it outputs the conveying speed correction direction; and when the main source of abnormal operating conditions is tool status, it outputs the tool maintenance direction.

[0033] Preferably, after the process correction module continuously collects the performance indicators of three detection intervals, if the deviation of the corresponding performance indicator decreases in each interval and the linkage judgment module no longer determines that there is a linkage change, then the corrected operating condition parameter interval and performance indicator interval are written into the subsequent judgment benchmark; if the deviation of the corresponding performance indicator does not decrease or there is still a linkage change, then the original judgment benchmark is maintained and the process correction direction is output again.

[0034] The technical effects and advantages of this invention are as follows:

[0035] In this invention, optical imaging and contour analysis of bundled short-cut fibers are performed using an attitude-fixed imaging module and a contour calculation module. This allows length dispersion, filament quantity, bundle looseness, cut neatness, and sizing uniformity to be uniformly obtained from image data under the same detection channel, avoiding data fragmentation caused by independent outputs from different detection devices. A workstation correspondence module maps performance indicators to cutting speed, traction tension, sizing supply, drying temperature, conveying speed, and tool status according to batch flow and workstation sequence, enabling quality fluctuations to be traced back to the corresponding production stage. A linkage judgment module determines whether multiple performance indicators exhibit synchronous shifts, continuous shifts, or interdependent changes, avoiding reliance on single indicators alone. The system can identify the main sources of abnormal operating conditions from multiple simultaneously changing parameters by calculating the magnitude, duration, and proximity of process parameter changes to the start time of linked changes through the operating condition traceability module; generate warning information when indicators have not exceeded the limit but have shown continuous deviation through the critical warning module, enabling the system to prompt process adjustments before non-conforming products are formed; and verify the results of the corrected performance indicator changes through the review and correction module, and adjust the subsequent judgment criteria accordingly, so that the system can adapt to the production fluctuations of different batches and specifications of fibers, and improve the real-time performance, accuracy, and traceability of bundled chopped fiber quality analysis. Attached Figure Description

[0036] The disclosure of this invention is illustrated with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. In the drawings, the same reference numerals are used to refer to the same parts:

[0037] Figure 1 This is an overall architecture diagram of an online analysis system for performance testing data of bundled chopped fibers according to the present invention;

[0038] Figure 2 This is a flowchart illustrating the operational logic of an online analysis system for testing the performance of bundled chopped fibers, as described in this invention.

[0039] Figure 3 This is a schematic diagram of the attitude-fixed imaging module structure of an online analysis system for performance testing data of bundled short-cut fibers according to the present invention;

[0040] Figure 4 This is a flowchart of the contour calculation and index conversion of an online analysis system for performance testing data of bundled short-cut fibers according to the present invention;

[0041] Figure 5 The workstation corresponding to the phase registration logic diagram of the online analysis system for performance testing data of bundled short-cut fibers of the present invention;

[0042] Figure 6 This is a flowchart illustrating the linkage judgment process of an online analysis system for performance testing data of bundled chopped fibers according to the present invention.

[0043] Figure 7 This is a working condition traceability logic diagram of an online analysis system for performance testing data of bundled chopped fibers according to the present invention;

[0044] Figure 8 This is a flowchart illustrating the critical warning and correction process of an online analysis system for performance testing data of bundled short-cut fibers according to the present invention. Detailed Implementation

[0045] It is readily understood that, based on the technical solution of this invention, those skilled in the art can propose various interchangeable structural methods and implementations without altering the essential spirit of the invention. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solution of this invention.

[0046] Reference Figure 1 As shown, this invention provides a technical solution: an online analysis system for performance testing data of bundled chopped fibers, comprising a data acquisition layer, a core analysis layer, and an output feedback layer. The data acquisition layer includes optical detection data sources and process and batch data sources; the optical detection data source includes a spreading and feeding unit, a multi-directional illumination unit, and an image acquisition unit, used to form optical images of the fiber bundle; the process and batch data source includes a batch binding unit, a process acquisition unit, and a phase registration unit, used to provide batch information, process parameters, and time registration information corresponding to the fiber bundle production process.

[0047] The core analysis layer includes an online analysis platform, which sequentially includes an attitude-fixed imaging module, a contour calculation module, an index conversion module, a workstation correspondence module, a linkage judgment module, a working condition traceability module, a critical warning module, and a review and correction module. These modules are connected sequentially according to the data flow, forming a closed loop from fiber optic image acquisition to performance index output, working condition traceability, critical warning, and review and correction.

[0048] The output feedback layer includes performance index output, early warning result output, correction guidance unit, and benchmark adjustment unit. Performance index output is used to output length dispersion, filament quantity, bundle looseness, cut neatness, and sizing uniformity. Early warning result output is used to output abnormal batches, abnormal performance indicators, main sources of abnormal operating conditions, and early warning levels. The correction guidance unit outputs cutting speed, traction tension, sizing supply, drying temperature, conveyor speed, and tool maintenance direction. The benchmark adjustment unit is used to adjust subsequent judgment benchmarks after verifying the effectiveness of the correction.

[0049] Reference Figure 2 The attitude-fixed imaging module includes a spreading and feeding unit, a fixed-distance conveying unit, a multi-directional illumination unit, and an image acquisition unit. The spreading and feeding unit feeds the bundled chopped fibers to be inspected into the inspection channel, and uses guide plates, limiting sidewalls, or spreading rollers to spread the fiber bundle into a single layer or thin layer on the conveyor belt. The fixed-distance conveying unit guides the fiber bundle through the inspection field of view along the conveying direction and controls the imaging distance between the plane of the fiber bundle and the lens of the image acquisition unit, ensuring that the image ratio remains consistent across different inspection intervals.

[0050] Multi-directional illumination units are positioned above and below the detection field of view. The upper illumination highlights the fiber end contours, bundle edges, and cut contours, while the lower illumination enhances the contrast of the filament expansion boundaries and fiber overlap areas. An image acquisition unit is positioned above the detection field of view to acquire optical images of the fiber bundle as it passes through. To reduce the impact of fiber bundle stacking on contour recognition, the spreading and feeding unit restricts the fiber bundles entering the detection channel to a preset width, and adjacent fiber bundle segments must not form a stacked state that obstructs the main area of ​​the detection field of view. The image acquisition unit acquires images within the same detection interval under the corresponding upper illumination, lower illumination, and combined upper and lower illumination conditions, ensuring that the fiber ends, filament expansion boundaries, cut contours, and sizing grayscale distribution have distinguishable image sources. Each detection interval corresponds to a set of fiber bundle optical images, which includes at least image data reflecting the fiber ends, bundle edges, filament expansion areas, cut areas, and sizing grayscale distribution.

[0051] To ensure a unified basis for subsequent calculations, this embodiment defines the conveying direction as the X-direction, the direction perpendicular to the conveying direction and located within the fiber bundle spreading plane as the Y-direction, and the direction perpendicular to the fiber bundle spreading plane as the Z-direction. Each pixel in the image acquired by the image acquisition unit can be mapped to its actual position within the detection field of view. Specifically, the X-direction is used to calculate fiber length and cutting direction deviation, the Y-direction is used to calculate bundle width and filament expansion range, and the Z-direction is used to define the imaging distance and illumination incident direction.

[0052] Reference Figure 3 The contour calculation module includes a field-of-view calibration unit, a background correction unit, and a boundary extraction unit. The field-of-view calibration unit is used to determine the conversion relationship between pixel size and actual size. Let the actual length corresponding to a unit pixel in the detected image be... If the pixel distance between any two points in an image is d, then the actual distance D between the two points is: .

[0053] The background correction unit is used to correct the background grayscale of the detection channel. Let the original fiber bundle optical image be I(x,y), the background image of the detection channel without fibers be B(x,y), and the corrected image be I'(x,y). Then, it can be processed using the following formula:

[0054] I'(x,y)=I(x,y)-B(x,y);

[0055] When the detection channel is affected by changes in illumination intensity, the background correction unit can also normalize I'(x,y) based on the average gray value of the edge region of the detection field of view, so that the gray values ​​of the images obtained at different detection times can be compared.

[0056] The boundary extraction unit extracts the fiber end contour, bundle edge, filament outward boundary, cut contour, and sizing grayscale distribution from the corrected image. The fiber end contour is used to determine the positions of the two ends of the chopped fiber; the bundle edge is used to determine the outer boundary of the main body of the fiber bundle; the filament outward boundary is used to determine the monofilament or filament region radiating outward from the bundle edge; the cut contour is used to determine the shape of the fiber cut end; and the sizing grayscale distribution is used to reflect the distribution state of the sizing layer on the fiber bundle surface. During boundary extraction, the system first extracts the fiber region from the background-corrected image, and then divides the fiber region into a bundle region and a filament outward expansion region according to the area of ​​the connected region, the aspect ratio of the region, and the edge continuity; among them, the continuous and dense region extending along the conveying direction is determined as the bundle region, and the region located outside the bundle edge and distributed in a fine linear discrete pattern is determined as the filament outward expansion region.

[0057] In this embodiment, the boundary extraction unit first performs grayscale thresholding on the corrected image to obtain the fiber region, and then performs edge tracking on the fiber region to form a set of contour points. For the bundle edge and the filament outward boundary, the boundary extraction unit uses the continuous dense area within the fiber region as the bundle region and the discrete fine line-like area outside the bundle region as the filament outward expansion region. For the cut contour, the boundary extraction unit extracts the end boundary line intersecting with the set cutting direction within the fiber end region. For the sizing grayscale distribution, the boundary extraction unit statistically analyzes the grayscale value changes within the bundle region to obtain the grayscale set of the sizing region.

[0058] The index conversion module converts the image features obtained by the contour measurement module into performance indicators. Each detection interval is denoted as K, the number of chopped fibers extracted within detection interval K is n, the target cutting length is L0, and the measured length of the i-th chopped fiber is L. i The average length is:

[0059] .

[0060] The standard deviation of length is:

[0061] .

[0062] Length dispersion D L Calculate using the following formula:

[0063] D L =σ L / L0.

[0064] Length dispersion D L The larger the value, the more unstable the distribution of chopped fiber lengths within the same detection range. When the end of a fiber is obscured or its outline is incomplete, the fiber is not included in the length dispersion calculation and is marked as an invalid length sample in the detection record to avoid misjudgment of length dispersion caused by abnormal images.

[0065] The amount of fuzz is calculated based on the area of ​​the fuzz's outer boundary extending beyond the edge of the bundle. Let A be the area of ​​the bundle region within the detection interval K. b The area of ​​the outer expansion region of the hair is A. h The amount of hair H is then calculated using the following formula:

[0066] H=A h / A b .

[0067] The larger the amount of fuzz H, the more fuzzes are emitted from the edge of the fiber bundle. The area Ab of the bundle body region is calculated from the pixel area within the closed range of the bundle body edge, and the area Ah of the fuzz outward expansion region is calculated from the pixel area located outside the bundle body edge and identified as the fuzz region.

[0068] The looseness of the filament bundle is calculated based on the fluctuation value of the bundle edge width. m measurement positions are selected along the conveying direction, and the bundle width at the j-th measurement position is W. j The average width of the beam is W avg The standard deviation of the beam width is σ W The looseness S of the bundled fibers is calculated using the following formula:

[0069] S=σ W / W avg .

[0070] The larger the fiber bundle looseness S, the more obvious the fluctuation in fiber bundle width and the worse the bundle stability.

[0071] Cut neatness is represented by the skew angle of the cut contour relative to the set cutting direction. Let θ be the skew angle between the i-th cut contour and the set cutting direction. i If the mean angle of the incision within the detection interval K is C, then:

[0072] .

[0073] The smaller the cut neatness index C, the closer the cut is to the set cutting direction and the neater the cut is; the larger the C, the more obvious the cut deviation.

[0074] The uniformity of sizing is calculated based on the sizing grayscale distribution within the bundle region. Let the maximum grayscale value in the sizing region be G. max The minimum grayscale value is G. min The average grayscale value is G. avg The uniformity of sizing, U, is calculated using the following formula:

[0075] U=(G max -G min ) / G avg .

[0076] The smaller the sizing uniformity index U, the smaller the difference in sizing grayness and the more uniform the sizing distribution; the larger the U, the more uneven the sizing distribution.

[0077] Therefore, the index conversion module forms a performance index set Q within each detection interval. K :

[0078] Q K ={D L ,H,S,C,U};

[0079] Where D L H represents length dispersion, S represents filament quantity, C represents filament looseness, and U represents sizing uniformity.

[0080] Reference Figure 4The corresponding module at each workstation includes a batch binding unit, a process acquisition unit, and a phase registration unit. The batch binding unit sets the performance index Q for each detection interval. K It is linked to the raw yarn batch, chopped specifications, production period, equipment number, and test number to form a batch test record.

[0081] The process acquisition unit collects the set of process parameters P corresponding to this batch. K :

[0082] P K ={V,T,A,D,R,B};

[0083] Where V is the cutting speed, T is the traction tension, A is the sizing supply, D is the drying temperature, R is the conveying speed, and B is the tool condition. The tool condition can be determined by the cumulative number of cuts, the tool wear level, or the tool vibration state.

[0084] Because there is a transport time between the fiber bundle and the cutting, traction, sizing, drying, and conveying stations, and then to the inspection field of view, the phase registration unit performs time compensation on the performance index sequence and process parameter sequence according to the transport time of the fiber bundle from the front-end station to the inspection field of view. Let the process parameter acquisition time be t. ip The performance index was detected at time t. iq If the phase compensation between the two is Δi, then:

[0085] Δi=t iq -t ip ;

[0086] The phase registration unit shifts the performance index sequence along the time axis according to Δi, so that the performance index Q within the detection interval K is... K This corresponds to the actual process parameters P experienced by the fiber bundle at the front-end station. K After registration is completed, the corresponding module at each workstation outputs the corresponding operating condition data and performance indicators, ensuring that subsequent linkage judgments and operating condition traceability are based on the same batch and the same workstation phase.

[0087] Reference Figure 5 The linkage judgment module is used to determine whether there are synchronous offsets, continuous offsets, or interdependent changes among multiple performance indicators. The linkage judgment module first establishes corresponding indicator benchmark values. For any indicator X among length dispersion, filament quantity, bundle looseness, cut neatness, and sizing uniformity, the indicator values ​​of five consecutive qualified test intervals are selected as benchmark samples, and the corresponding indicator benchmark value XB is calculated.

[0088] X B =(X1+X2+X3+X4+X5) / 5;

[0089] The index value of the current detection interval K is X. K The offset is ΔX K :

[0090] ΔX K =X K -X B ;

[0091] Offset direction signX K Determined according to the following rules:

[0092] When ΔX K When ΔX > 0, the offset direction is positive; when ΔX K When ΔX < 0, the offset direction is the reverse offset; when ΔX K When = 0, the offset direction is no offset.

[0093] Offset amplitude P XK Calculate using the following formula:

[0094] P XK =|ΔX K | / X B ;

[0095] For each detection interval, the linkage judgment module calculates the offset direction and offset amplitude of length dispersion, filament quantity, bundle looseness, cut neatness, and sizing uniformity. When at least two of the above five indicators have the same offset direction in the same detection interval, and the at least two indicators maintain the same offset direction in three consecutive detection intervals, while the corresponding offset amplitude increases in each interval, the linkage judgment module determines that multiple performance indicators have a linkage change.

[0096] Specifically, let the three consecutive detection intervals be K1, K2, and K3, and let the offset magnitude of the same index X in the three detection intervals be P respectively. XK1 P XK2 and P XK3 When P is satisfied XK1 <P XK2 <P XK3 When the deviation of the indicator increases in intervals, it is determined that the deviation is increasing. If at least two indicators meet the conditions of unidirectional deviation and increasing in intervals, the output linkage change is valid; otherwise, the output linkage change is invalid. This judgment method can identify the state where a single indicator has not exceeded the limit but multiple indicators have deteriorated synchronously, thus avoiding judging the fiber bundle quality based solely on the single qualified threshold.

[0097] Reference Figure 6The working condition tracing module is used to determine the source of the main abnormal working condition after a linkage change occurs. When the linkage judgment module outputs that a linkage change has been established, the working condition tracing module retrieves the corresponding process parameters before the start of the linkage change. The process parameters include cutting speed, traction tension, sizing supply, drying temperature, conveying speed, and tool status.

[0098] For any process parameter Y, the operating condition traceability module calculates the variation range M of that process parameter. Y Duration T Y And the proximity N to the start of the linkage change Y Let Y B Let Y0 be the process parameter baseline value, and Y0 be the process parameter value at the sampling time before the linkage change begins. Then the change amplitude M Y Calculate using the following formula:

[0099] M Y =|Y0-Y B | / Y B ;

[0100] Duration T Y The number of consecutive sampling intervals in which the process parameter changes in the same direction is determined. Let r be the number of consecutive sampling intervals in which the process parameter changes in the same direction. The operating condition traceability module will use T... Y Let it be denoted as r.

[0101] Proximity N Y The time interval between the start time of the process parameter change and the start time of the linkage change is determined. Let the start time of the process parameter change be t. Y The start time of the linkage change is t. L The detection interval duration is T. c ,but:

[0102] ;

[0103] When the change in process parameters is closer to the start of the linkage change, N Y The larger N is; the further away the change in process parameters is from the start of the linkage change, the greater N becomes. Y The smaller.

[0104] The operating condition traceability module calculates the comprehensive comparison value F for each process parameter. Y :

[0105] F Y =M Y ×T Y ×N Y ;

[0106] The cutting speed, traction tension, sizing supply, drying temperature, conveying speed, and tool condition are respectively corresponding to F YCompare and put F Y The largest process parameter was identified as the primary source of abnormal operating conditions. If the F-value of two process parameters... Y If they are the same, then prioritize selecting the one that is closest to the start time of the linked change, N. Y Larger process parameters; if N Y If they are still the same, then select the range of change M. Y Larger process parameters.

[0107] In this way, the operating condition tracing module does not simply output all fluctuating process parameters, but determines the main sources of abnormal operating conditions based on the magnitude, duration and timing of the changes, thus transforming the identification of abnormal causes from manual experience judgment into executable online calculation logic.

[0108] Reference Figure 7 The critical warning module generates a warning message when performance indicators do not exceed the specified limits. For each performance indicator, the system pre-sets a corresponding limit range. Within the current detection interval, if the length dispersion, filament quantity, bundle looseness, cut neatness, and sizing uniformity do not exceed their respective limits, but the linkage judgment module has determined that there is a linkage change, and the main abnormal working condition source determined by the working condition tracing module maintains the same direction of change in three consecutive detection intervals, then the critical warning module generates a critical warning message.

[0109] Critical early warning information includes abnormal batches, abnormal performance indicators, main sources of abnormal operating conditions, and warning levels. Abnormal performance indicators are those involved in the linked changes. The main sources of abnormal operating conditions are the process parameters output by the operating condition tracing module. The warning level is determined based on the maximum deviation of the indicators involved in the linked changes. Let P be the maximum normalized deviation of the indicators involved in the linked changes. max The normalized offset magnitude corresponding to the performance index exceeding the limit boundary is P. limit The warning ratio E is calculated using the following formula:

[0110] E=P max / P limit ;

[0111] When 0 < E < 0.4, the warning level is Level 1; when 0.4 ≤ E < 0.7, the warning level is Level 2; and when 0.7 ≤ E < 1, the warning level is Level 3. Level 1 warnings are used to alert for potential issues, Level 2 warnings are used to suggest adjustments and preparations, and Level 3 warnings are used to prompt immediate correction. Since the critical warning module only operates when performance indicators are within the specified range, it can output warning information before quality defects materialize.

[0112] The review and correction module generates correction directions based on the main sources of abnormal operating conditions. If the main source of abnormal operating conditions is the cutting speed, the review and correction module outputs the cutting speed correction direction based on the offset direction of the cutting speed relative to its process reference value; when the cutting speed is higher than the process reference value, it outputs a direction to decrease the cutting speed, and when the cutting speed is lower than the process reference value, it outputs a direction to increase the cutting speed.

[0113] If the main source of abnormal operating conditions is traction tension, the verification and correction module outputs the traction tension correction direction based on the offset direction of the traction tension relative to its process reference value; when the traction tension is higher than the process reference value, it outputs the direction of reducing traction tension, and when the traction tension is lower than the process reference value, it outputs the direction of increasing traction tension.

[0114] If the main source of abnormal operating conditions is the amount of sizing supplied, the verification and correction module outputs the direction of sizing supply correction based on the offset direction of the sizing supply relative to its process reference value; when the sizing supply is higher than the process reference value, it outputs the direction of reducing the sizing supply; when the sizing supply is lower than the process reference value, it outputs the direction of increasing the sizing supply.

[0115] If the main source of abnormal operating conditions is drying temperature, the verification and correction module outputs the drying temperature correction direction based on the offset direction of the drying temperature relative to its process reference value; when the drying temperature is higher than the process reference value, it outputs the direction of decreasing the drying temperature, and when the drying temperature is lower than the process reference value, it outputs the direction of increasing the drying temperature.

[0116] If the main source of abnormal operating conditions is the conveying speed, the verification and correction module outputs the conveying speed correction direction according to the offset direction of the conveying speed relative to its process reference value; when the conveying speed is higher than the process reference value, it outputs the direction of reducing the conveying speed, and when the conveying speed is lower than the process reference value, it outputs the direction of increasing the conveying speed.

[0117] If the main source of abnormal operating conditions is the tool condition, the verification and correction module outputs the tool maintenance direction, which includes at least one of the following: stopping the machine to check the cutting edge, cleaning the tool attachments, replacing the tool, or correcting the cutting gap.

[0118] After the process correction is completed, the verification and correction module continuously collects performance indicators for three detection intervals and compares the corrected performance indicators with the original offset state. Let P be the offset magnitude of the performance indicators involved in the linkage change within the three consecutive detection intervals after correction. a P b and P c When P is satisfied a >P b >P c Furthermore, if the linkage judgment module no longer determines that there is a linkage change, the review and correction module will determine that the correction is valid and write the corrected operating condition parameter range and performance index range into the subsequent judgment benchmark.

[0119] If the performance index deviation does not decrease in each of the three consecutive detection intervals after correction, or if the linkage judgment module still determines that there is a linkage change, the review and correction module determines that the correction has not met the requirements, maintains the original judgment benchmark, and outputs the process correction direction again. When outputting the process correction direction again, the review and correction module continues to call the calculation results of the operating condition traceability module. If the main abnormal operating condition source has changed, the correction direction is output according to the new main abnormal operating condition source; if the main abnormal operating condition source has not changed, the adjustment continues along the original correction direction.

[0120] Reference Figure 8 During system operation, bundled short-cut fibers first enter the detection channel. The attitude-fixing imaging module guides the fiber bundle through the detection field of view in a set posture and acquires optical images of the fiber bundle. The contour calculation module performs field of view calibration, background correction, and boundary extraction on the optical images of the fiber bundle to obtain the fiber end contour, bundle edge, filament expansion boundary, cut contour, and sizing grayscale distribution. The index conversion module converts the above extraction results into length dispersion, filament quantity, bundle looseness, cut neatness, and sizing uniformity. The station correspondence module maps the above performance indicators to cutting speed, traction tension, sizing supply, drying temperature, conveying speed, and tool status according to batch flow and station sequence. The linkage judgment module determines whether there are linkage changes in multiple performance indicators. The working condition traceability module identifies the main abnormal working condition source after linkage changes occur. The critical warning module generates critical warning information when the performance indicators have not exceeded the limits but have shown continuous deviation. The verification and correction module outputs the process correction direction and verifies the corrected detection results.

[0121] When the verification results show that the offset amplitude has decreased and the linkage change has disappeared, the system updates the subsequent judgment benchmark and enters the next round of testing; when the verification results show that the offset amplitude has not decreased or the linkage change has not disappeared, the system outputs the process correction direction again until the performance indicators return to stability. Through this operation process, this system forms a complete closed loop from optical inspection, index conversion, workstation correspondence, linkage judgment, working condition tracing, critical warning to verification and correction. Therefore, the system does not perform optical inspection or alarm prompts in isolation, but achieves a continuous closed loop of attitude imaging, contour calculation, index conversion, workstation correspondence, linkage judgment, working condition tracing, critical warning, verification and correction, and benchmark update under the same batch of fiber bundles, thereby supporting the early identification of quality degradation and the location of the main abnormal working condition sources before the performance indicators exceed the limits.

[0122] The technical scope of this invention is not limited to the content described above. Those skilled in the art can make various modifications and variations to the above embodiments without departing from the technical concept of this invention, and all such modifications and variations should fall within the protection scope of this invention.

Claims

1. An online analysis system for performance testing data of bundled chopped short fibers, characterized in that, include: An attitude-fixing imaging module is used to enable the bundled chopped fibers to pass through the detection field of view in a set orientation and to acquire optical images of the fiber bundle. The contour calculation module is used to extract the chopped fiber end contour, bundle edge, filament outward expansion boundary, cut contour, and sizing grayscale distribution from fiber bundle optical images. The index conversion module is used to convert the extracted results into indices such as length dispersion, filament quantity, bundle looseness, cut neatness, and sizing uniformity. The workstation correspondence module is used to map performance indicators to cutting speed, traction tension, sizing supply, drying temperature, conveying speed and tool status according to batch flow and workstation sequence. The linkage determination module is used to determine the synchronous offset, continuous offset, and mutual interference changes among multiple performance indicators; The working condition tracing module is used to determine the source of the main abnormal working conditions based on the linkage judgment results and the corresponding results of the workstation; The critical warning module is used to generate warning information when performance indicators have not exceeded the limit range but have shown a continuous deviation. The review and correction module is used to output the direction of process correction and adjust the subsequent judgment criteria based on the changes in the corrected performance indicators. The contour calculation module includes a field-of-view calibration unit, a background correction unit, and a boundary extraction unit. The field of view calibration unit is used to determine the conversion relationship between pixel size and actual size; The background correction unit is used to correct the background grayscale of the detection channel; The boundary extraction unit is used to extract the fiber end contour, bundle edge, filament outer boundary, cut contour, and sizing grayscale distribution. The index conversion module uses the ratio of the standard deviation of the distance between the ends of each fiber in the same detection interval to the target cutting length as the length dispersion, the ratio of the area of ​​the filament expansion boundary beyond the edge of the bundle to the area of ​​the bundle as the filament amount, the ratio of the width fluctuation value of the bundle edge to the average width of the bundle as the bundle looseness, the skew angle of the cut contour relative to the set cutting direction as the cut neatness, and the ratio of the gray difference value of the sizing area to the average gray value of the area as the sizing uniformity index. The linkage judgment module uses the average performance index of five consecutive qualified test intervals as the corresponding index benchmark value, compares the performance index of the current test interval with the corresponding index benchmark value, and obtains the offset direction and offset amplitude. When at least two of the length dispersion, filament amount, bundle looseness, cut neatness and sizing uniformity maintain the same offset in three consecutive test intervals, and the offset amplitude increases in each interval, it is determined that multiple performance indicators have linkage changes. The verification and correction module generates correction directions based on the main sources of abnormal operating conditions. When the main source of abnormal operating conditions is cutting speed, it outputs the cutting speed correction direction; when the main source of abnormal operating conditions is traction tension, it outputs the traction tension correction direction; when the main source of abnormal operating conditions is sizing supply, it outputs the sizing supply correction direction; when the main source of abnormal operating conditions is drying temperature, it outputs the drying temperature correction direction; when the main source of abnormal operating conditions is conveying speed, it outputs the conveying speed correction direction; and when the main source of abnormal operating conditions is tool status, it outputs the tool maintenance direction.

2. The online analysis system for performance testing data of bundled chopped fibers according to claim 1, characterized in that: The attitude-fixed imaging module includes a spreading and feeding unit, a fixed-distance conveying unit, a multi-directional illumination unit, and an image acquisition unit. The spreading and feeding unit is used to disperse and spread the bundled short-cut fibers into the detection channel; The fixed-distance delivery unit is used to control the imaging distance between the fiber bundle and the detection lens; The multi-directional lighting unit is used to illuminate the fiber end, bundle edge, filament expansion area and cut area respectively; The image acquisition unit is used to acquire fiber optic images under the corresponding illumination conditions.

3. The online analysis system for performance testing data of bundled chopped fibers according to claim 1, characterized in that: The workstation corresponding module includes a batch binding unit, a process acquisition unit, and a phase registration unit: The batch binding unit is used to bind performance indicators with raw yarn batch, chopped yarn specification, production period and equipment number; The process acquisition unit is used to acquire the process parameters of the corresponding batch; The phase registration unit is used to compensate for performance indicators and process parameters based on the fiber transport time from the cutting station to the inspection field of view.

4. The online analysis system for performance testing data of bundled chopped fibers according to claim 1, characterized in that: After a linkage change occurs, the working condition tracing module retrieves the cutting speed, traction tension, sizing supply, drying temperature, conveying speed, and tool status corresponding to the time before the linkage change begins. It then calculates the change amplitude, duration, and proximity to the time of the linkage change for each working condition parameter, and identifies the working condition parameter with the largest product of change amplitude, duration, and proximity as the main source of abnormal working conditions.

5. The online analysis system for performance testing data of bundled chopped fibers according to claim 1, characterized in that: When the performance indicators do not exceed the limit range, if the linkage judgment module determines that there is a linkage change, and the main abnormal operating condition source determined by the operating condition tracing module maintains the same direction of change in three consecutive detection intervals, the critical warning module generates critical warning information. The critical warning information includes abnormal batch, abnormal performance indicators, main abnormal operating condition source, and warning level.

6. The online analysis system for performance testing data of bundled chopped fibers according to claim 1, characterized in that: After the process is corrected, the review and correction module continuously collects the performance indicators of three detection intervals. If the deviation of the corresponding performance indicator decreases in each interval and the linkage judgment module no longer determines that there is a linkage change, the corrected working condition parameter interval and performance indicator interval are written into the subsequent judgment benchmark. If the deviation of the corresponding performance index does not decrease or there are still linked changes, the original judgment criteria will be maintained and the process correction direction will be output again.

Images