Optical fiber storage device and feeding and winding control method thereof

Abstract

The invention relates to the technical field of optical fiber feeding and winding, and discloses an optical fiber storage device and a feeding and winding control method thereof, and the feeding and winding control method comprises the steps: collecting optical fiber stress data in real time based on a sliding window, carrying out the data preprocessing of the stress data, calculating a stress average value, comparing the stress average value with a stress threshold value, and judging whether the optical fiber feeding and winding process is abnormal or not; when it is judged that the optical fiber feeding and winding process is abnormal, processing the to-be-clustered set based on K-means clustering, and adjusting the optical fiber feeding and winding speed according to a clustering result; and after the optical fiber winding is completed, collecting an optical fiber winding completion image, judging whether the winding condition of the optical fiber after the winding is completed is normal or not, and when the winding condition of the optical fiber is judged to be abnormal, determining an early warning type according to the winding condition of the optical fiber, and giving out an early warning. According to the invention, adaptive speed adjustment for different abnormal types is realized, and non-uniform feeding and winding, winding overlapping or optical fiber damage are avoided.

Description

Technical Field

[0001] This invention relates to the field of optical fiber feeding and winding technology, and more specifically, to an optical fiber receiving device and its feeding and winding control method. Background Technology

[0002] Medical laser fibers are commonly used optical transmission elements in minimally invasive surgery, endoscopic surgery, and laser therapy. Their transmission stability and safety directly affect surgical precision and treatment outcomes. During use, medical laser fibers need to be wound or guided by a feeding device according to set tension and speed to ensure smooth fiber transmission during operation and to avoid damage caused by bending, uneven winding, or excessive localized stress. Existing fiber feeding methods typically rely on fixed speed control or simple tension monitoring, which is insufficient to meet the high precision and safety requirements of medical laser fibers in practical use. Due to the thin diameter and high flexibility of medical laser fibers, they are easily affected by surgical procedures, equipment vibration, and external interference during use. Traditional control methods are slow to respond when fiber tension is abnormal or the winding is uneven, easily causing fiber overlap, interruption, or damage, thus affecting laser transmission efficiency and clinical safety.

[0003] Therefore, it is necessary to design an optical fiber receiving device and its winding control method to solve the problems existing in the current technology. Summary of the Invention

[0004] In view of this, the present invention proposes an optical fiber storage device and its winding control method, aiming to solve the problem that abnormal optical fiber storage tension or uneven winding state can easily cause optical fiber winding overlap, interruption or damage.

[0005] In one aspect, the present invention proposes a method for controlling the feeding and winding of an optical fiber receiving device, comprising: The force data of the force sensor during the optical fiber feeding and winding process is collected in real time based on the sliding window. The force data is preprocessed and the average force value of the preprocessed force data is calculated. The average force value is compared with the force threshold to determine whether there is any abnormality in the optical fiber feeding and winding process. When an anomaly is detected in the fiber winding process, data feature values ​​are extracted from the preprocessed stress data. These feature values ​​include the mean, variance, peak value, and slope. A data feature vector is generated based on these feature values. This data feature vector is then integrated with historical anomaly vectors from a historical adjustment set to form a clustering set. The historical adjustment set includes several historical anomaly vectors and several historical adjustment coefficients, with each historical anomaly vector corresponding to a historical adjustment coefficient. The clustering set is then processed using K-means clustering, and the fiber winding speed is adjusted based on the clustering results. After the optical fiber winding is completed, an image of the completed optical fiber winding is acquired. Based on the image, it is determined whether the optical fiber winding condition is normal. If the optical fiber winding condition is determined to be abnormal, the warning type is determined based on the optical fiber winding condition, and a warning is issued.

[0006] Furthermore, based on the real-time acquisition of force data from the force sensor during the fiber winding process using a sliding window, the preprocessing of the force data includes: The force sensor continuously collects the force data of the optical fiber at a fixed sampling frequency, and stores the force data in the data buffer in chronological order to form a time series of force data. Based on the window size and sliding step of the sliding window, continuous force data within the current sliding window range is extracted from the data buffer in real time. The continuous force data within the sliding window is preprocessed, including filtering and normalization.

[0007] Furthermore, when comparing the average force with the force threshold to determine whether there is an abnormality in the fiber winding process, the following steps are included: The force threshold includes a first force threshold and a second force threshold; the first force threshold is less than the second force threshold. When the average force is less than or equal to the first force threshold, it is determined that there is an abnormality in the fiber winding process; When the average force is greater than the first force threshold and less than the second force threshold, it is determined that there is no abnormality in the optical fiber winding process. When the average force is greater than or equal to the second force threshold, it is determined that there is an abnormality in the fiber winding process.

[0008] Furthermore, when processing the set to be clustered based on K-means clustering and adjusting the fiber winding speed according to the clustering results, the process includes: Normalize each data point in the set to be clustered; S1: Initialize K centroids in the set to be clustered, and assign the remaining feature sets to the nearest centroids to form K clusters; S2: Recalculate the centroid of each cluster; S3: Repeat S1 and S2 until the centroid no longer changes; When the cluster containing the feature set does not contain the historical feature set, the historical adjustment coefficient corresponding to the maximum similarity between the feature set and the historical feature set is selected as the initial adjustment coefficient. When the cluster containing the feature set contains a historical feature set, the average value of the historical adjustment coefficients corresponding to all the historical feature sets is selected as the adjustment coefficient, and the fiber winding speed is adjusted according to the adjustment coefficient.

[0009] Furthermore, when acquiring an image of the completed fiber winding process and determining whether the fiber winding status is normal based on the image, the process includes: Several images of the winding region at different angles are acquired, and the images of the winding region are preprocessed, including grayscale conversion, Gaussian blurring, and contrast enhancement. The preprocessed images of the winding region are then used to extract the fiber contour features and fiber centerline using edge detection and skeletonization methods, respectively. Geometric feature analysis is then performed based on the fiber contour features and fiber centerline.

[0010] Furthermore, when extracting the fiber contour features and fiber centerline from the preprocessed winding region image using edge detection and skeletonization methods respectively, the process includes: The edge detection method described above is used to extract edge information from grayscale images. The grayscale change rate of each pixel in the edge information is calculated in both horizontal and vertical directions to obtain gradient images in both directions. The gradient magnitude and gradient direction of each pixel are calculated based on the gradient images. By comparing the gradient magnitudes of adjacent pixels in the gradient direction, the pixels with the local maximum gradient magnitude are retained, and the blurred pixels around the edge lines are removed to generate a binary edge image. The pixels with the local maximum gradient magnitude correspond to the fiber optic contour features. Skeletonization is performed on the binary edge image to compress the fiber width to a single-pixel width for the fiber centerline; and the skeletonized fiber centerline is then connected by breaks and smoothed.

[0011] Furthermore, when performing geometric feature analysis based on the fiber profile characteristics and fiber centerline, it includes: The geometric feature analysis includes spacing uniformity analysis, discontinuity detection, winding error detection, and uneven tightness detection; The spacing uniformity analysis involves projecting the centerline of each fiber loop along the winding axis to generate the position coordinates of each loop, calculating the spacing between adjacent loops, and statistically analyzing the average value and standard deviation. When the spacing deviation exceeds the spacing deviation threshold, it is determined to be a non-uniform spacing. The discontinuity detection involves performing connected component analysis on the centerline, measuring the length of each continuous line segment, and determining an optical fiber discontinuity when the length of a continuous line segment is lower than a length threshold. The error detection process involves extracting the positions of contour intersection points and determining whether the number of intersection points exceeds the intersection point threshold. When the number of intersection segments exceeds the intersection point threshold, it is determined to be an error. The unevenness detection is performed by detecting protrusions or depressions on the winding surface through local width and brightness.

[0012] When the width or height changes exceed the abnormal change threshold, it is judged as uneven tightness.

[0013] Furthermore, when acquiring an image of the completed fiber winding process and determining whether the fiber winding status is normal based on the completed image, the process also includes: When at least one abnormal behavior is found in the result of the geometric feature analysis, the fiber winding condition after the winding is completed is determined to be abnormal.

[0014] Furthermore, when the fiber optic cable winding condition is determined to be abnormal, the warning type is determined based on the fiber optic cable winding condition, including: The warning types include mechanical defect warnings and quality deviation warnings; When fiber optic cable breaks or twisting occurs, the warning type is determined to be a mechanical defect warning. When uneven spacing or uneven tension occurs, the warning type is determined to be a quality deviation warning.

[0015] On the other hand, this application also provides an optical fiber receiving device for applying the above-mentioned optical fiber receiving device winding control method, including: Optical fiber, winding structure, force sensor, image acquisition device and control unit; The winding structure includes a rotating buckle, a rotating shaft, and an auxiliary winding opening. The rotating shaft is coaxially connected to the rotating buckle. The conveyor belt is sleeved on the rotating shaft and the output shaft of the motor to transmit the power of the motor to the rotating shaft. One end of the optical fiber is wound around the rotating buckle, and the other end of the optical fiber passes through the auxiliary feeding port; The force sensor is installed on the auxiliary feeding port and is used to receive the force data of the auxiliary feeding port during the optical fiber feeding and winding process. Multiple image acquisition devices are provided to acquire images of the winding area after the optical fiber winding is completed from different angles. The control unit is used to control the fiber winding speed based on the collected data during the fiber winding process, and to determine the warning type and issue a warning based on the collected data after the fiber winding is completed.

[0016] It is also provided with an outer buckle, and the rotating buckle is provided with a connecting hole that cooperates with the outer buckle. The outer buckle is placed in the connecting hole and cooperates with the rotating buckle to form an inner and outer double buckle structure.

[0017] Compared with existing technologies, the advantages of this invention are as follows: It employs a sliding window method to collect stress data in real time during the fiber winding process, and performs data preprocessing and average value calculation. This allows for accurate judgment of the fiber winding tension state in a short time, enabling immediate identification of anomalies and improving the response speed of the winding process. After detecting an anomaly, it extracts multi-dimensional feature values ​​such as average, variance, peak value, and slope, combines historical anomaly vectors and historical adjustment coefficients to construct a clustering set, and analyzes it based on the K-means clustering algorithm. This achieves adaptive speed adjustment for different anomaly types, avoiding uneven winding, overlapping winding, or fiber damage. After winding is completed, multiple image acquisition devices acquire images of the winding area from different angles, and automatically identify the winding status. This comprehensively judges the quality indicators such as the neatness and tightness consistency of the finished product, resulting in more accurate detection results. Based on the identification results of the winding status, it automatically determines the warning type and issues corresponding graded warning prompts, enabling operators to quickly take targeted measures. The internal and external dual-card structure provides dual restraint, not only fixing the fiber to the storage device but also restricting the fiber from spreading outwards, preventing fiber detachment. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of the fiber optic cable feeding and winding control method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the optical fiber storage device provided in an embodiment of the present invention without the external clip connected; Figure 3 This is a schematic diagram of the structure of the optical fiber storage device provided in an embodiment of the present invention, with the external clip already connected.

[0019] The components include: 1. Optical fiber; 2. Rotary buckle; 3. Rotating shaft; 4. Auxiliary feeding and winding port; 5. Conveyor belt; 6. Motor; 7. Force sensor; 8. Image acquisition device; 9. Control unit; 10. External buckle; 11. Connecting hole. Detailed Implementation

[0020] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey its scope to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] In some embodiments of this application, see Figure 1 As shown, a method for controlling the feeding and winding of optical fiber receiving devices is proposed, including: S100: Based on the sliding window, the force data of the force sensor during the optical fiber feeding and winding process is collected in real time. The force data is preprocessed and the average force value of the preprocessed force data is calculated. The average force value is compared with the force threshold to determine whether there is any abnormality in the optical fiber feeding and winding process. S200: When an anomaly is detected in the fiber feeding and winding process, data feature values ​​are extracted from the preprocessed force data. These feature values ​​include the mean, variance, peak value, and slope. A data feature vector is generated based on these feature values. This data feature vector is then integrated with historical anomaly vectors from the historical adjustment set to form a set to be clustered. The historical adjustment set includes several historical anomaly vectors and several historical adjustment coefficients, with each historical anomaly vector corresponding to a historical adjustment coefficient. The set to be clustered is then processed using K-means clustering, and the fiber feeding and winding speed is adjusted based on the clustering results. S300: After the optical fiber winding is completed, an image of the completed optical fiber winding is acquired. Based on the image, it is determined whether the optical fiber winding condition is normal. If the optical fiber winding condition is determined to be abnormal, the warning type is determined based on the optical fiber winding condition, and a warning is issued.

[0022] Specifically, during the fiber winding process, force sensors installed at the auxiliary winding port collect real-time force data of the fiber. A sliding window method is used for data acquisition, processing continuously acquired force data in segments, with statistical analysis performed simultaneously on the data within each window, ensuring rapid acquisition of the latest force information. The acquired force data is filtered (e.g., low-pass filtered) to remove high-frequency noise; data normalization ensures comparability between data from different sensors or different fiber batches. The average value of the pre-processed force data is calculated; this average value is compared with a preset force threshold; if the average value exceeds or falls below the threshold, an anomaly is determined in the fiber winding process. Multidimensional feature values ​​are extracted from the pre-processed force data, including average, variance, peak value, and slope; the average value reflects the overall tension level, variance reflects the tension fluctuation amplitude, peak value reflects instantaneous anomalies, and slope reflects the rate of tension change. These feature values ​​are combined to form a data feature vector; the current feature vector is combined with historical anomaly vectors and their corresponding historical adjustment coefficients to construct a clustering set for determining the anomaly type and severity. The K-means clustering algorithm is used to process the clustered sets. Based on the clustering results, the anomaly type and severity are mapped to the fiber winding speed adjustment strategy. The motor speed is automatically adjusted according to the analysis results to achieve dynamic speed regulation. After the fiber winding is completed, high-definition images are acquired from different angles by multiple image acquisition devices placed around the winding area. The images cover the entire winding area, ensuring no blind spots. Edge detection and fiber centerline extraction are performed on the acquired images to analyze the fiber arrangement uniformity and interlayer spacing; it is determined whether there are problems such as overlapping, uneven tightness, or abnormal stacking during winding. The warning level is determined according to the type and degree of winding anomaly; corresponding warning prompts are issued for operators or automatic measures to be taken. For example, in a medical laser fiber winding operation, after the fiber winding is completed, the image acquisition device takes images of the fiber winding area from multiple angles. The control unit processes the acquired images, including grayscale conversion, edge detection, and centerline extraction. Analysis results showed that a section of the optical fiber had a broken centerline, indicating a fiber optic interruption. Simultaneously, the number of intersections in some areas exceeded a set threshold, indicating mis-winding. According to the control unit's preset rules, fiber interruptions and mis-winding are considered mechanical defects, thus the fiber winding condition was deemed abnormal. The control unit categorized this anomaly as a "mechanical defect warning" and issued a system alert to notify the operator for inspection and handling. This example demonstrates that this method can promptly detect winding anomalies after fiber winding is completed, distinguish different types of anomalies, and issue targeted warnings, thereby avoiding fiber damage and risks to surgical safety.

[0023] Understandably, the fiber optic winding control method achieves rapid identification and dynamic speed adjustment of tension anomalies during the winding process by real-time acquisition of stress data, data feature extraction, and cluster analysis. Simultaneously, it automatically detects the fiber winding quality through multi-angle image acquisition and generates graded early warnings. This method can improve the uniformity and stability of fiber winding and reduce the risk of fiber damage.

[0024] In some embodiments of this application, when preprocessing the force data of the force sensor during the optical fiber winding process by real-time acquisition of the force data based on a sliding window, the process includes: The force sensor continuously collects force data on the optical fiber at a fixed sampling frequency and stores the force data in time sequence into a data buffer to form a time series of force data. Based on the window size and sliding step of the sliding window, continuous force data within the current sliding window range is extracted from the data buffer in real time. The continuous force data within the sliding window is preprocessed, including filtering and normalization.

[0025] Specifically, the force sensor continuously collects force data of the optical fiber during the winding process at a fixed sampling frequency (e.g., several times per second, which can be set to 100~1000Hz depending on the fiber winding speed). The collected data is stored in a data buffer in chronological order to form a time series, ensuring data continuity and reflecting the trend of optical fiber tension change over time. The data buffer can adopt a ring buffer structure, allowing the latest data to overwrite the oldest data, ensuring data continuity during sliding window processing. Based on the preset sliding window size (e.g., 100 sampling points), continuous force data within the current sliding window range is extracted from the data buffer in real time. Each time the window is updated, the latest data segment is automatically acquired for analysis to ensure real-time performance and continuity. The window size and step size can be adjusted according to the fiber winding speed and control accuracy to balance processing efficiency and monitoring sensitivity. Low-pass filtering, moving average filtering, or weighted filtering algorithms are used to remove high-frequency noise and occasional spikes in the data acquisition process, thereby improving the stability and reliability of the data. The filtered data is then normalized according to a preset range, for example, by mapping the force values ​​to the range of 0 to 1, to eliminate data differences between different batches or different sensors, providing a unified data foundation for subsequent feature extraction and cluster analysis. After data preprocessing, stable and standardized force data within the sliding window can be generated, providing a reliable basis for subsequent anomaly detection and dynamic speed control.

[0026] Understandably, the force sensor continuously collects fiber optic force data at a fixed sampling frequency and stores it in a time-series data buffer. A sliding window is used to extract continuous force data within the current window range from the buffer in real time, enabling real-time tracking of fiber tension changes. This method ensures rapid response to tension anomalies, allowing for timely detection of abnormalities and reducing the risk of fiber damage. The force data within the sliding window is filtered and denoised to eliminate high-frequency noise and measurement spikes. Normalization converts the force data into a unified standard range, eliminating the influence of differences between different batches of fibers or sensors on the analysis results. This improves the accuracy of subsequent feature extraction and anomaly analysis, providing a reliable foundation for dynamic control. The pre-processed sliding window data can be used to calculate characteristic values ​​such as the average force, variance, peak value, and trend. This supports subsequent anomaly detection, cluster analysis, and dynamic adjustment of the winding speed, enabling intelligent control of the winding process. It improves adaptability while ensuring winding uniformity, ensuring stable force and neat arrangement of the fiber during the winding process.

[0027] In some embodiments of this application, when comparing the average force with a force threshold to determine whether there is an abnormality in the fiber winding process, the method includes: The force threshold includes a first force threshold and a second force threshold; the first force threshold is less than the second force threshold. When the average force is less than or equal to the first force threshold, it is determined that there is an abnormality in the fiber winding process; When the average force is greater than the first force threshold and less than the second force threshold, it is determined that there is no abnormality in the fiber winding process. When the average force is greater than or equal to the second force threshold, it is determined that there is an abnormality in the fiber winding process.

[0028] Specifically, the force thresholds include a first force threshold and a second force threshold, where the first force threshold is less than the second force threshold. The first force threshold is used to detect situations where the tension is too low during the fiber winding process, and the second force threshold is used to detect situations where the tension is too high during the fiber winding process. The thresholds can be preset or dynamically adjusted according to parameters such as fiber diameter, material properties, winding speed, and storage tray specifications to adapt to different fiber batches and production conditions. When the average force within the sliding window is less than or equal to the first force threshold, it is determined that there is an abnormality in the fiber winding process, indicating that the fiber may be loose, not tightly wound, or have insufficient tension. When the average force is greater than the first force threshold but less than the second force threshold, it is determined that the fiber winding process is normal, the tension is within a reasonable range, and the winding is uniform and stable. When the average force is greater than or equal to the second force threshold, it is determined that there is an abnormality in the fiber winding process, indicating that the fiber may be too tight, or there is a risk of excessive force or damage. The average force is calculated based on the sliding window data and compared with a preset threshold in real time. Automatic judgment can be achieved through a logic judgment module or software algorithm without manual intervention. When an anomaly is detected, subsequent feature analysis, clustering processing or dynamic speed adjustment can be triggered to quickly respond to fiber tension anomalies.

[0029] Understandably, by setting a first and a second force threshold and comparing the average force calculated by the sliding window, accurate judgment of anomalies in the fiber winding process is achieved. The dual-threshold judgment method can simultaneously identify two types of anomalies: excessively low and excessively high fiber tension, ensuring that the winding process remains within a safe and reasonable tension range. This method has a simple and clear judgment logic, is easy to implement in real-time control, and can quickly trigger anomaly handling or dynamic speed adjustment measures, thereby preventing damage caused by fiber slack, uneven winding, or excessive force, and improving the uniformity and stability of fiber storage.

[0030] In some embodiments of this application, when processing the set to be clustered based on K-means clustering and adjusting the fiber winding speed according to the clustering results, the following steps are included: Normalize each data point in the cluster set; S1: Initialize K centroids in the set to be clustered, and assign the remaining feature sets to the nearest centroids to form K clusters; S2: Recalculate the centroid of each cluster; S3: Repeat S1 and S2 until the centroid no longer changes; When the cluster containing the feature set does not contain the historical feature set, the historical adjustment coefficient corresponding to the maximum similarity between the feature set and the historical feature set is selected as the initial adjustment coefficient. When the cluster containing the feature set contains a historical feature set, the average value of the historical adjustment coefficients corresponding to all the historical feature sets is selected as the adjustment coefficient, and the fiber winding speed is adjusted according to the adjustment coefficient.

[0031] Specifically, the feature vector set generated during the fiber winding process is first normalized, mapping the data of each feature dimension to a uniform numerical range (e.g., 0-1) to eliminate the influence of different feature dimensions and value ranges on the clustering results, thus ensuring the accuracy and stability of the clustering analysis. Subsequently, the normalized feature set is processed using the K-means clustering algorithm: K centroids are initialized in the set to be clustered, and their initial positions can be determined randomly, heuristically, or based on historical data; then, the remaining feature sets are assigned to the nearest centroids according to their distance from each centroid, forming K initial clusters. The clustering process is continuously optimized through iteration: in each iteration, the centroid of each cluster is recalculated, i.e., the average of all feature vectors within the cluster is calculated as the new centroid position; then, all feature sets are reassigned to the nearest centroid cluster. This iterative process is repeated until the centroid positions stabilize and no longer change or the change is less than a preset threshold, at which point clustering is complete. When determining the adjustment coefficient for fiber winding speed based on clustering results, historical feature sets are used as a reference: when the cluster containing the feature set does not contain historical feature sets, the similarity between the current feature set and historical feature sets is calculated, and the historical adjustment coefficient corresponding to the maximum similarity is selected as the initial adjustment coefficient; when the cluster contains historical feature sets, the historical adjustment coefficients corresponding to all historical feature sets within the cluster are taken, and their average value is calculated as the current adjustment coefficient. Subsequently, the motor output is adjusted in real time according to this adjustment coefficient to achieve dynamic adjustment of the fiber winding speed, enabling the fiber tension to quickly return to the normal range when abnormalities occur.

[0032] Understandably, by normalizing the feature vector set of the fiber winding process and classifying the feature set based on the K-means clustering algorithm, dynamic adjustment of the fiber winding speed is achieved. Using historical feature set references and average adjustment coefficients, a rapid response is possible when fiber tension is abnormal, automatically adjusting the winding speed to a suitable range, ensuring uniform fiber tension and stable winding. This effectively prevents damage caused by fiber slack, uneven winding, or excessive force.

[0033] In some embodiments of this application, when acquiring an image of the completed optical fiber winding and determining whether the optical fiber winding status is normal based on the completed image, the process includes: Several images of the winding region from different angles were acquired, and these images were preprocessed, including grayscale conversion, Gaussian blurring, and contrast enhancement. The preprocessed winding region images were then used to extract the fiber contour features and fiber centerline using edge detection and skeletonization methods, respectively. Geometric feature analysis was performed based on the fiber contour features and fiber centerline.

[0034] Specifically, multiple image acquisition devices are set up to capture images of the wound area after the optical fiber winding is completed from different angles to obtain complete winding information. Each acquisition device can be an industrial camera, with resolution and shooting angle adjusted according to the fiber diameter and storage tray size to ensure clear images covering the entire winding area. Multi-angle images can compensate for potential occlusion or local distortion problems from a single angle. The acquired images are converted to grayscale to reduce data dimensionality and highlight fiber edge information. Gaussian blurring is applied to the grayscale images to remove high-frequency noise and stray textures, improving the robustness of edge detection. Contrast enhancement is used to highlight the grayscale difference between the fiber and the background, facilitating subsequent edge extraction and contour analysis. Edge detection methods (such as the Canny operator) are used to extract fiber contour features from the preprocessed images to obtain the fiber's shape boundary information. Skeletonization processing is performed on the edge images to extract the fiber centerline, which reflects the fiber path and winding direction. Contour features and centerline data can be used to further calculate fiber arrangement density, winding uniformity, and overlapping areas. Based on the extracted fiber outline and centerline, geometric feature analysis is performed, including spacing uniformity analysis, discontinuity detection, winding error detection, and uneven tightness detection. By comparing with preset standard values ​​or historical normal samples, it is determined whether the fiber winding is uniform and tight, and whether there are abnormal winding or overlapping areas. The analysis results can be used to trigger early warnings and dynamically adjust the subsequent fiber winding speed.

[0035] Understandably, by acquiring multi-angle images of the fiber after winding, and preprocessing these images such as grayscale conversion, Gaussian blurring, and contrast enhancement, combined with edge detection and skeletonization methods to extract fiber contour features and centerline, accurate analysis of the fiber winding geometry can be achieved. This method can comprehensively and accurately determine whether the fiber winding is uniform and tight, and whether there are overlapping or abnormal areas. The use of multi-angle acquisition and integrated image processing techniques improves the reliability and accuracy of winding state detection and reduces the risk of fiber damage.

[0036] In some embodiments of this application, geometric feature analysis based on fiber profile characteristics and fiber centerline includes: Geometric feature analysis includes spacing uniformity analysis, discontinuity detection, winding error detection, and uneven tension detection; Spacing uniformity analysis involves projecting the centerline of each fiber loop along the winding axis to generate the position coordinates of each loop, calculating the spacing between adjacent loops, and statistically analyzing the average and standard deviation. When the spacing deviation exceeds the spacing deviation threshold, it is determined to be non-uniform spacing. Discontinuity detection involves performing connected component analysis on the centerline and measuring the length of each continuous segment. If the length of a continuous segment is lower than a length threshold, it is determined to be an optical fiber discontinuity. Error detection involves extracting the positions of contour intersection points and determining whether the number of intersection points exceeds the intersection point threshold. When the number of intersection segments exceeds the intersection point threshold, it is determined to be an error. Uneven tension detection involves checking for protrusions or depressions on the winding surface by measuring local width and brightness.

[0037] When the width or height changes exceed the abnormal change threshold, it is judged as uneven tightness.

[0038] Specifically, the spacing uniformity analysis involves projecting the centerline of each fiber loop along the winding axis to generate the position coordinates of each loop; calculating the spacing between adjacent loops and statistically analyzing the average spacing and standard deviation of the entire winding area; when the spacing deviation exceeds a preset spacing deviation threshold, an anomaly of uneven spacing in the fiber winding is determined; this analysis method can effectively detect situations where the fiber loops are arranged too densely or too sparsely, ensuring the uniformity of fiber winding. Discontinuity detection involves performing connected component analysis on the centerline to identify each continuous fiber segment; measuring the length of each continuous segment, and determining that the fiber has a discontinuity when the length of a continuous segment is lower than a preset length threshold; this step can promptly detect fiber breakage or detachment, preventing subsequent abnormal fiber winding and damage to the finished product. Wrapping error detection involves extracting the positions of the fiber outline intersections and counting the number of intersection points; when the number of intersection points exceeds a preset intersection point threshold, a wrapping error is determined, i.e., the fiber has crossed or overlapped during the winding process; this detection can detect abnormal fiber winding direction or local overlap problems, improving the standardization and reliability of fiber winding. Uneven tension detection uses local width measurement and brightness analysis to detect protrusions or depressions on the fiber winding surface. When the local width or height change exceeds the abnormal change threshold, it is determined that there is an uneven tension problem in the fiber winding. This method can identify uneven tension distribution caused by excessive or insufficient force on the fiber during winding, thereby improving the fiber alignment quality.

[0039] It is understandable that by analyzing the geometric features of the fiber profile and centerline, a comprehensive inspection of the fiber winding quality can be achieved. Specifically, this includes spacing uniformity analysis, discontinuity detection, incorrect winding detection, and uneven tension detection. Spacing uniformity analysis can determine whether the fiber coils are evenly spaced, promptly detecting spacing anomalies; discontinuity detection can identify fiber breakage or detachment, preventing wire breakage during winding; incorrect winding detection can detect fiber crossings, overlaps, or winding errors, improving the standardization of fiber handling; uneven tension detection can identify abnormal fiber tension distribution, preventing damage caused by excessive tightness or looseness. Through the above comprehensive analysis methods, this invention can accurately determine the winding state of the fiber after winding, improving the uniformity and stability of fiber handling and reducing the risk of fiber damage.

[0040] In some embodiments of this application, when acquiring an image of the completed optical fiber winding and determining whether the optical fiber winding status is normal based on the completed image, the method further includes: If the results of geometric feature analysis show at least one abnormal behavior, the fiber winding condition after the winding is completed is determined to be abnormal.

[0041] Specifically, a comprehensive analysis of the geometric characteristics of fiber winding is performed, including spacing uniformity analysis, discontinuity detection, winding error detection, and uneven tension detection. If at least one abnormal behavior is observed in the analysis results—namely, the fiber coil spacing deviation exceeding a preset spacing deviation threshold; the presence of discontinuous segments with lengths below a length threshold; the number of fiber outline intersections exceeding a cross-line point threshold (indicating winding error); or the fiber's local width or height variation exceeding an abnormal variation threshold (indicating uneven tension)—then the fiber winding condition is deemed abnormal. Once at least one abnormal indicator is detected, the current fiber winding state is immediately marked as abnormal, and an early warning signal can be triggered or abnormal data can be recorded for subsequent processing. In some embodiments of this application, when it is determined that the fiber optic winding condition is abnormal, the method for determining the warning type based on the fiber optic winding condition includes: Early warning types include mechanical defect warnings and quality deviation warnings; When fiber optic cable breaks or twisting occurs, the warning type is determined to be a mechanical defect warning. When uneven spacing or uneven tension occurs, the warning type is determined to be a quality deviation warning.

[0042] Specifically, the warning types are divided into two categories: mechanical defect warnings and quality deviation warnings. Mechanical defect warnings mainly target anomalies such as breakage and mis-winding caused by equipment or mechanical reasons during the fiber winding process. Quality deviation warnings mainly target anomalies affecting the winding quality, such as uneven spacing or uneven tension during the fiber winding process. When geometric feature analysis or centerline detection results show that the fiber has discontinuities or the number of contour intersections exceeds a threshold, i.e., mis-winding occurs, the anomaly is determined to be a mechanical defect. When the analysis results show that the fiber spacing deviation exceeds a set threshold or the local width / height change exceeds the abnormal change threshold, i.e., uneven spacing or uneven tension occurs, the anomaly is determined to be a quality deviation. Based on the determination results, the corresponding warning type is associated with the current fiber winding status.

[0043] In another preferred embodiment based on the above embodiments, see [reference] Figure 2-3 As shown, this embodiment provides an optical fiber receiving device for applying the above-described optical fiber receiving device winding control method, including: 1. Fiber optic cable, winding structure, 7. Force sensor, 8. Image acquisition device, and 9. Control unit; The winding structure includes a rotating buckle 2, a rotating shaft 3, and an auxiliary winding opening 4. The rotating shaft 3 is coaxially connected to the rotating buckle 2. The conveyor belt 5 is sleeved on the rotating shaft 3 and the output shaft of the motor 6, and is used to transmit the power of the motor 6 to the rotating shaft 3. One end of optical fiber 1 is wrapped around the rotating buckle 2, and the other end of optical fiber 1 passes through the auxiliary feeding port 4; Force sensor 7 is installed on auxiliary feeding port 4 to receive force data of auxiliary feeding port 4 during the feeding and winding of optical fiber 1; Multiple image acquisition devices 8 are provided to acquire images of the winding area of ​​the optical fiber 1 after winding is completed from different angles; Control unit 9 is used to control the winding speed of optical fiber 1 based on the collected data during the winding process, and to determine the warning type and issue a warning based on the collected data after the winding of optical fiber 1 is completed. It is also provided with an outer buckle 10, and a connecting hole 11 that cooperates with the outer buckle 10 is provided on the rotating buckle 2. The outer buckle 10 is placed in the connecting hole 11 and cooperates with the rotating buckle 2 to form an inner and outer double buckle structure.

[0044] It is also equipped with an outer casing, which is used to enclose the optical fiber 1, the winding structure, the force sensor 7, the image acquisition device 8, the control unit 9, and the outer buckle 10 inside.

[0045] Specifically, the fiber optic cable receiving device is used to achieve automated fiber feeding, tension equalization, and winding fixation. This device includes an optical fiber 1, a feeding structure, a force sensor 7, an image acquisition device 8, and a control unit 9. The feeding structure includes a rotating latch 2, a rotating shaft 3, and an auxiliary feeding port 4, wherein the rotating shaft 3 is coaxially connected to the rotating latch 2. The output shaft of the motor 6 is connected to the rotating shaft 3 via a conveyor belt 5, transmitting the power of the motor 6 to the rotating shaft 3, thereby driving the rotating latch 2 to feed the optical fiber 1. One end of the optical fiber 1 is wound around the rotating latch 2, and the other end passes through the auxiliary feeding port 4, allowing the optical fiber to smoothly enter the receiving structure. The force sensor 7 is installed on the auxiliary feeding port 4 to collect real-time force data of the optical fiber during the feeding process, providing the control unit 9 with a basis for monitoring the fiber tension. Multiple image acquisition devices 8 are provided to acquire images of the winding area after the optical fiber feeding is completed from different angles, enabling analysis and detection of the optical fiber winding quality. The control unit 9 controls the fiber winding speed in real time based on the data collected by the force sensor 7, ensuring uniform and stable tension during the winding process. After the fiber winding is completed, the control unit 9 analyzes the image data to determine the winding state and issues warnings for mechanical defects or quality deviations based on abnormal conditions. To further improve the fixation reliability after the fiber winding is completed, the device is also equipped with an outer buckle 10. The rotating buckle 2 has a connection hole 11 that mates with the outer buckle 10. The outer buckle 10 is placed in the connection hole 11 and mates with the rotating buckle 2 to form an inner and outer double-clamp structure. After the fiber winding is completed, the outer buckle 10 is engaged, which can effectively fix the fiber wound on the rotating buckle 2 and prevent the fiber from loosening or falling off due to spring force or other factors. The entire outer shell is used to enclose all components such as the fiber 1, winding structure, force sensor 7, image acquisition device 8, control unit 9, and outer buckle 10, thereby providing physical protection and environmental isolation for the above components. The outer shell can be made of metal materials (such as aluminum alloy, stainless steel) or high-strength engineering plastics to ensure sufficient mechanical strength and impact resistance. To facilitate equipment installation, maintenance, and fiber optic replacement, the outer casing is equipped with an openable cover, and transparent windows or light-transmitting covers are provided in certain locations, allowing operators to observe the internal fiber winding status without opening the casing. A fiber inlet is located at the position corresponding to the auxiliary winding port 4, and a transparent protective cover or light-transmitting hole is provided at the corresponding position of the image acquisition device 8 to ensure image quality and prevent dust, liquid, and foreign objects from entering. This casing not only prevents damage to the internal structure from external dust, liquid, and mechanical impact, but also avoids fiber damage or winding abnormalities caused by accidental contact by operators.

[0046] In summary, by using a sliding window method to collect stress data in real time during the fiber winding process, and performing data preprocessing and average value calculation, the tension state of the fiber winding can be accurately determined in a short time, enabling immediate identification of anomalies and improving the response speed of the winding process. After an anomaly is detected, multi-dimensional feature values ​​such as average, variance, peak value, and slope are extracted, and a clustering set is constructed by combining historical anomaly vectors and historical adjustment coefficients. Analysis is then performed based on the K-means clustering algorithm, achieving adaptive speed adjustment for different anomaly types and avoiding uneven winding, overlapping winding, or fiber damage. After winding is completed, multiple image acquisition devices are used to acquire images of the winding area from different angles, and the winding status is automatically identified. This allows for a comprehensive assessment of quality indicators such as the neatness and tightness of the finished product, resulting in more accurate detection results. Based on the identification results of the winding status, the warning type is automatically determined, and corresponding graded warning prompts are issued, enabling operators to quickly take targeted measures. The internal and external dual-card structure provides dual restraint, not only fixing the fiber to the storage device but also preventing the fiber from spreading outward and falling off.

[0047] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0048] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0049] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0050] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0051] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for controlling the feeding and winding of an optical fiber receiving device, characterized in that, include: The force data of the force sensor during the optical fiber feeding and winding process is collected in real time based on the sliding window. The force data is preprocessed and the average force value of the preprocessed force data is calculated. The average force value is compared with the force threshold to determine whether there is any abnormality in the optical fiber feeding and winding process. When an anomaly is detected in the fiber winding process, data feature values ​​are extracted from the preprocessed stress data. These feature values ​​include the mean, variance, peak value, and slope. A data feature vector is generated based on these feature values. This data feature vector is then integrated with historical anomaly vectors from a historical adjustment set to form a clustering set. The historical adjustment set includes several historical anomaly vectors and several historical adjustment coefficients, with each historical anomaly vector corresponding to a historical adjustment coefficient. The clustering set is then processed using K-means clustering, and the fiber winding speed is adjusted based on the clustering results. After the optical fiber winding is completed, an image of the completed optical fiber winding is acquired. Based on the image, it is determined whether the optical fiber winding condition is normal. If the optical fiber winding condition is determined to be abnormal, the warning type is determined based on the optical fiber winding condition, and a warning is issued.

2. The fiber optic cable feeding and winding control method according to claim 1, characterized in that, Based on the real-time acquisition of force data from force sensors during the fiber winding process using a sliding window, the preprocessing of the force data includes: The force sensor continuously collects the force data of the optical fiber at a fixed sampling frequency, and stores the force data in the data buffer in chronological order to form a time series of force data. Based on the window size and sliding step of the sliding window, continuous force data within the current sliding window range is extracted from the data buffer in real time. The continuous force data within the sliding window is preprocessed, including filtering and normalization.

3. The fiber optic cable feeding and winding control method according to claim 2, characterized in that, When comparing the average force with the force threshold to determine whether there is an abnormality in the fiber winding process, the following steps are included: The force threshold includes a first force threshold and a second force threshold; the first force threshold is less than the second force threshold. When the average force is less than or equal to the first force threshold, it is determined that there is an abnormality in the fiber winding process; When the average force is greater than the first force threshold and less than the second force threshold, it is determined that there is no abnormality in the optical fiber winding process. When the average force is greater than or equal to the second force threshold, it is determined that there is an abnormality in the fiber winding process.

4. The fiber optic cable feeding and winding control method according to claim 3, characterized in that, When processing the set to be clustered based on K-means clustering and adjusting the fiber winding speed according to the clustering results, the following steps are included: Normalize each data point in the set to be clustered; S1: Initialize K centroids in the set to be clustered, and assign the remaining feature sets to the nearest centroids to form K clusters; S2: Recalculate the centroid of each cluster; S3: Repeat S1 and S2 until the centroid no longer changes; When the cluster containing the feature set does not contain the historical feature set, the historical adjustment coefficient corresponding to the maximum similarity between the feature set and the historical feature set is selected as the initial adjustment coefficient. When the cluster containing the feature set contains a historical feature set, the average value of the historical adjustment coefficients corresponding to all the historical feature sets is selected as the adjustment coefficient, and the fiber winding speed is adjusted according to the adjustment coefficient.

5. The fiber optic cable feeding and winding control method according to claim 4, characterized in that, When acquiring an image of the completed fiber winding process, and determining whether the fiber winding status is normal based on the image, the process includes: Several images of the winding region at different angles are acquired, and the images of the winding region are preprocessed, including grayscale conversion, Gaussian blurring, and contrast enhancement. The preprocessed images of the winding region are then used to extract the fiber contour features and fiber centerline using edge detection and skeletonization methods, respectively. Geometric feature analysis is then performed based on the fiber contour features and fiber centerline.

6. The fiber optic cable feeding and winding control method according to claim 5, characterized in that, When extracting fiber contour features and fiber centerline from the preprocessed wrapped region image using edge detection and skeletonization methods, respectively, the following steps are taken: The edge detection method described above is used to extract edge information from grayscale images. The grayscale change rate of each pixel in the edge information is calculated in both horizontal and vertical directions to obtain gradient images in both directions. The gradient magnitude and gradient direction of each pixel are calculated based on the gradient images. By comparing the gradient magnitudes of adjacent pixels in the gradient direction, the pixels with the local maximum gradient magnitude are retained, and the blurred pixels around the edge lines are removed to generate a binary edge image. The pixels with the local maximum gradient magnitude correspond to the fiber optic contour features. Skeletonization is performed on the binary edge image to compress the fiber width to a single-pixel width for the fiber centerline; and the skeletonized fiber centerline is then connected by breaks and smoothed.

7. The fiber optic cable feeding and winding control method according to claim 6, characterized in that, When performing geometric feature analysis based on the fiber profile characteristics and fiber centerline, the following are included: The geometric feature analysis includes spacing uniformity analysis, discontinuity detection, winding error detection, and uneven tightness detection; The spacing uniformity analysis involves projecting the centerline of each fiber loop along the winding axis to generate the position coordinates of each loop, calculating the spacing between adjacent loops, and statistically analyzing the average value and standard deviation. When the spacing deviation exceeds the spacing deviation threshold, it is determined to be a non-uniform spacing. The discontinuity detection involves performing connected component analysis on the centerline, measuring the length of each continuous line segment, and determining an optical fiber discontinuity when the length of a continuous line segment is lower than a length threshold. The error detection process involves extracting the positions of contour intersection points and determining whether the number of intersection points exceeds the intersection point threshold. When the number of intersection segments exceeds the intersection point threshold, it is determined to be an error. The uneven tightness detection is performed by detecting protrusions or depressions on the winding surface through local width and brightness. When the width or height changes exceed the abnormal change threshold, it is judged as uneven tightness.

8. The fiber optic cable feeding and winding control method according to claim 7, characterized in that, When acquiring an image of the completed optical fiber winding, and determining whether the optical fiber winding status is normal based on the completed image, the method further includes: When at least one abnormal behavior is found in the result of the geometric feature analysis, the fiber winding condition after the winding is completed is determined to be abnormal.

9. The fiber optic cable feeding and winding control method according to claim 7, characterized in that, When an abnormal fiber optic winding condition is determined, the warning type is specified based on the fiber optic winding condition, including: The warning types include mechanical defect warnings and quality deviation warnings; When fiber optic cable breaks or twisting occurs, the warning type is determined to be a mechanical defect warning. When uneven spacing or uneven tension occurs, the warning type is determined to be a quality deviation warning.

10. A fiber optic cable receiving device, used in applying the fiber optic cable receiving device winding control method as described in any one of claims 1-9, characterized in that, include: Optical fiber, winding structure, force sensor, image acquisition device and control unit; The winding structure includes a rotating buckle, a rotating shaft, and an auxiliary winding opening. The rotating shaft is coaxially connected to the rotating buckle. The conveyor belt is sleeved on the rotating shaft and the output shaft of the motor to transmit the power of the motor to the rotating shaft. One end of the optical fiber is wound around the rotating buckle, and the other end of the optical fiber passes through the auxiliary feeding port; The force sensor is installed on the auxiliary feeding port and is used to receive the force data of the auxiliary feeding port during the optical fiber feeding and winding process. Multiple image acquisition devices are provided to acquire images of the winding area after the optical fiber winding is completed from different angles. The control unit is used to control the fiber winding speed based on the collected data during the fiber winding process, and to determine the warning type and issue a warning based on the collected data after the fiber winding is completed. It is also provided with an outer buckle, and the rotating buckle is provided with a connecting hole that cooperates with the outer buckle. The outer buckle is placed in the connecting hole and cooperates with the rotating buckle to form an inner and outer double buckle structure.

Images