A matrix ultraviolet curing method for improving flatness of optical fiber ribbons

Abstract

The application discloses a kind of matrix ultraviolet curing methods for improving the flatness of optical fiber ribbon, belong to optical communication component manufacturing technical field, it includes arrangement multiple optical fibers to form geometric reference template and coat preset curing material;Using low-power ultraviolet light to the geometric reference template is short-time exposure, generates optical fiber ribbon;Through real-time monitoring, the real-time stress distribution parameter and real-time flatness parameter of optical fiber ribbon are obtained, and monitoring data is generated;The monitoring data is carried out feature extraction, generates stress-flatness feature vector and through preset dynamic collaborative control algorithm, control preset programmable matrix ultraviolet light source to optical fiber ribbon is carried out matrix ultraviolet curing, generates the optical fiber ribbon after curing.Using real-time monitoring optical fiber ribbon internal stress and surface morphology, and based on monitoring result through dynamic collaborative control algorithm driving programmable matrix ultraviolet light source carries out pixelization irradiation closed-loop control mode, can actively manage stress and deformation in curing process, to improve the flatness of final optical fiber ribbon.

Description

Technical Field

[0001] This invention relates to the field of optical communication component manufacturing technology, and in particular to a matrix ultraviolet curing method for improving the flatness of optical fiber ribbons. Background Technology

[0002] Currently, in the field of fiber optic ribbon manufacturing, the flatness of the fiber optic ribbon directly affects its subsequent laying efficiency and signal transmission performance. Since fiber optic ribbons are typically formed by bonding multiple optical fibers with resin and then curing them under ultraviolet light, the resin shrinks during the curing process, generating internal stress. Furthermore, differences in the arrangement of the multiple optical fibers and uneven resin coating thickness can easily lead to an imbalance in the internal stress distribution of the fiber optic ribbon, resulting in problems such as surface unevenness and localized warping. This stress concentration is particularly pronounced in the production of high-density fiber optic ribbons, such as those with 12 cores or more, and can severely impact the bending performance and lifespan of the fiber optic ribbon, becoming a key issue restricting the improvement of fiber optic ribbon product quality.

[0003] To improve the flatness of fiber optic ribbons, traditional technologies often employ a single ultraviolet light source for overall curing. Curing parameters such as irradiation intensity and time are typically preset based on prior experiments. Some solutions use optical inspection equipment to check the surface flatness of the fiber optic ribbon after curing to determine product quality. A few improved solutions incorporate simple surface morphology monitoring during curing, but this only obtains surface contour information and cannot detect internal stress distribution. Furthermore, the monitoring results are only used for post-processing quality assessment and cannot provide real-time intervention in the curing process, thus failing to fundamentally solve the flatness problem caused by stress during curing.

[0004] However, traditional curing methods have significant drawbacks: Firstly, because real-time data on the internal stress of the fiber ribbon cannot be obtained, the preset curing parameters are difficult to adapt to stress variations in different batches and regions, easily leading to over- or under-curing in some areas and exacerbating stress concentration. Secondly, the overall irradiation mode of traditional light sources cannot achieve differentiated energy output, making it impossible to make targeted curing adjustments for areas with high stress. Furthermore, the lack of coordination with mechanical stress control makes it difficult to effectively release internal stress. In addition, traditional technologies lack a linkage mechanism between quality feedback and parameter optimization. Even if flatness issues are detected, parameters must be manually readjusted, which not only reduces production efficiency but also makes it difficult to ensure that the adjusted parameters are suitable for subsequent production, failing to meet the mass production requirements of high-precision fiber ribbons. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a matrix UV curing method for improving the flatness of optical fiber ribbons. This method employs a closed-loop control approach that monitors the internal stress and surface morphology of the optical fiber ribbon in real time and drives a programmable matrix UV light source for pixelated irradiation based on the monitoring results using a dynamic collaborative control algorithm. This approach can proactively manage stress and deformation during the curing process, thereby improving the final flatness of the optical fiber ribbon.

[0006] The above objectives can be achieved through the following approach:

[0007] A matrix UV curing method for improving the flatness of optical fiber ribbons includes: arranging multiple optical fibers to form a geometric reference template with a preset tension deviation; coating the geometric reference template with a preset curing material to obtain a geometric reference template coated with matrix material; short-term exposure of the matrix-coated geometric reference template with preset low-power UV light to transform the matrix-coated geometric reference template into a geometrically locked structure, generating an optical fiber ribbon; obtaining real-time stress distribution parameters and real-time flatness parameters of the optical fiber ribbon by real-time monitoring of the in-situ stress of the internal stress and surface morphology of the optical fiber ribbon, generating monitoring data; extracting features from the monitoring data to generate a stress-flatness feature vector characterizing the current overall state of the optical fiber ribbon; generating pixelated irradiation control commands for driving the light source based on the stress-flatness feature vector through a preset dynamic collaborative control algorithm; and controlling a preset programmable matrix UV light source composed of multiple independently addressable light-emitting units to perform matrix UV curing on the optical fiber ribbon according to the pixelated irradiation control commands, generating a cured optical fiber ribbon.

[0008] Optionally, the step of organizing multiple optical fibers to form a geometric reference template with a preset tension deviation includes: applying a preset independent tension to the multiple optical fibers to generate multiple tension optical fibers; and using a fiber splitting device to organize the multiple tension optical fibers to obtain a coplanar geometric reference template with uniform spacing.

[0009] Optionally, generating the fiber ribbon includes: controlling the irradiation intensity and exposure time of the low-power ultraviolet light to cause a preliminary gelation reaction on the surface of the ultraviolet-cured matrix material; forming a gel shell on the surface of the geometric reference template through the preliminary gelation reaction to obtain a geometrically locked structure and generate the fiber ribbon.

[0010] Optionally, the step of acquiring monitoring data includes: scanning the fiber optic strip using a preset optical coherence tomography sensor to acquire raw scan data containing depth and polarization information; performing image reconstruction on the raw scan data to generate a stress distribution image representing the internal stress distribution and a smoothness image representing the surface morphology; extracting real-time stress distribution parameters from the stress distribution image and extracting real-time smoothness parameters from the smoothness image, and combining the two to generate monitoring data.

[0011] Optionally, the step of generating the stress-smoothness feature vector includes: normalizing the parameters with different physical dimensions in the monitoring data to generate standardized monitoring data; performing principal component analysis on the standardized monitoring data to extract the main feature components used to characterize the main direction of data change; and combining the main feature components to construct the stress-smoothness feature vector.

[0012] Optionally, the step of generating pixelated irradiation control instructions includes: generating phase-modulated curing path parameters for guiding the curing reaction sequence based on the stress-flatness feature vector; generating a time-varying dynamic scanning irradiation mode based on the phase-modulated curing path parameters; and generating pixelated irradiation control instructions based on the dynamic scanning irradiation mode using the dynamic collaborative control algorithm.

[0013] Optionally, the dynamic collaborative control algorithm includes: based on the dynamic scanning illumination mode, analyzing the illumination area, intensity gradient, and scanning timing parameters that change over time in the mode; establishing a spatial mapping relationship between the dynamic scanning illumination mode and the light-emitting unit array of the programmable matrix ultraviolet light source, mapping the illumination area to the physical coordinates of the light-emitting units, and generating a spatial pixelated mapping table; based on the spatial pixelated mapping table, performing spatiotemporal decomposition of the intensity gradient of the dynamic scanning illumination mode to generate an initial illumination intensity sequence for each light-emitting unit; based on the stress-flatness feature vector, calculating the deviation value between the current stress-flatness feature vector and the preset target feature vector, and collaboratively correcting the initial illumination intensity sequence according to the deviation value; encapsulating the corrected illumination parameters of each light-emitting unit into a pixelated illumination control instruction, wherein the pixelated illumination control instruction includes the address code, intensity value, and time control signal of each light-emitting unit.

[0014] Optionally, the step of controlling the programmable matrix ultraviolet light source for matrix ultraviolet curing includes: parsing the pixelated irradiation control command to obtain a spatialized irradiation intensity distribution and a time-sequential irradiation time sequence; sending an independent driving signal to each light-emitting unit of the programmable matrix ultraviolet light source according to the irradiation intensity distribution and the irradiation time sequence; and adjusting the driving signal in real time according to the continuously updated pixelated irradiation control command during the irradiation process to dynamically maintain the flatness of the fiber ribbon.

[0015] Optionally, the method further includes: performing flatness detection on the optical fiber strip after curing to generate a detection result characterizing the final product quality; comparing the detection result with a preset quality target value to generate an error signal; and adjusting the internal parameters of the dynamic collaborative control algorithm based on the error signal to generate optimized control algorithm parameters.

[0016] Based on the same inventive concept, the present invention also provides a matrix UV curing system for improving the flatness of optical fiber ribbons, the system comprising:

[0017] The geometric reference template construction module is used to organize multiple optical fibers to form a geometric reference template with a preset tension deviation;

[0018] A matrix material coating module is used to coat a preset curing material onto the geometric reference template to obtain a geometric reference template coated with matrix material.

[0019] The geometric locking structure generation module is used to expose the geometric reference template coated with matrix material to preset low-power ultraviolet light for a short time, transforming the geometric reference template coated with matrix material into a geometric locking structure and generating an optical fiber ribbon;

[0020] The in-situ stress and flatness monitoring module is used to obtain the real-time stress distribution parameters and real-time flatness parameters of the optical fiber by real-time monitoring of the in-situ stress of the internal stress and surface morphology of the optical fiber, and to generate monitoring data.

[0021] The stress-flatness feature extraction and decision module is used to extract features from the monitoring data and generate a stress-flatness feature vector characterizing the current comprehensive state of the optical fiber strip.

[0022] The dynamic collaborative control command generation module is used to generate pixelated illumination control commands for driving the light source based on the stress-flatness feature vector and a preset dynamic collaborative control algorithm.

[0023] The matrix UV curing execution module is used to control a preset programmable matrix UV light source composed of multiple independently addressable light-emitting units to perform matrix UV curing on the optical fiber ribbon according to the pixelated irradiation control command, so as to generate a cured optical fiber ribbon.

[0024] Compared with the prior art, the present invention has the following advantages:

[0025] 1. Real-time and precise control of the fiber ribbon curing process has been achieved. The in-situ stress-flatness monitoring unit synchronously acquires real-time stress distribution parameters and flatness parameters of the fiber ribbon. Combined with the stress-flatness feature vector generated by feature extraction, and then using a dynamic collaborative control algorithm to generate pixelated irradiation control commands, it can accurately identify stress anomaly areas and surface morphology deviations in the fiber ribbon. This avoids the stress concentration or insufficient flatness problems caused by information lag in traditional curing, thus improving the controllability of the curing process.

[0026] 2. Possesses multi-parameter collaborative optimization capabilities. Not only does it construct the curing control foundation through stress-flatness parameters, but it also integrates tension parameters from the tension control system to form a comprehensive feature vector. Simultaneously, it combines dynamic tension control to apply oscillating tension to the fiber ribbon, achieving synergistic coordination between optical curing and mechanical stress release. This multi-dimensional parameter linkage approach can counteract macroscopic and microscopic stresses within the fiber ribbon, reducing defects such as warping and wrinkling after curing, further ensuring product form stability.

[0027] 3. A closed-loop quality improvement system was constructed. After curing, an error signal is generated through flatness detection. The internal parameters of the dynamic collaborative control algorithm are adjusted based on the error signal, enabling subsequent curing processes to be continuously optimized based on previous quality feedback. This closed-loop control mode breaks through the limitations of traditional curing where parameters are set only once. It can gradually reduce product quality fluctuations, improve the consistency of fiber optic ribbons in mass production, and meet the long-term use requirements of high-precision communication fiber optic ribbons.

[0028] 4. Differentiated curing is achieved through a programmable matrix ultraviolet light source. The light source consists of multiple independently addressable light-emitting units, which can output a spatialized irradiation intensity distribution and a temporally sequenced irradiation time sequence according to pixelated irradiation control commands, applying appropriate curing energy to high-stress and low-stress areas of the fiber ribbon respectively. This differentiated irradiation method avoids the uneven curing problem caused by traditional overall irradiation, and can optimize the curing effect of different areas, further improving the overall flatness and structural strength of the fiber ribbon.

[0029] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a schematic diagram of a matrix UV curing method for improving the flatness of optical fiber ribbons according to an embodiment of the present invention.

[0032] Figure 2 This is a schematic diagram illustrating the stress-flatness feature vector construction process of a matrix UV curing method for improving the flatness of optical fiber ribbons according to an embodiment of the present invention.

[0033] Figure 3 This is a schematic diagram of the feedback learning loop and curing parameter optimization effect of a matrix UV curing method for improving the flatness of optical fiber ribbon according to an embodiment of the present invention.

[0034] Figure 4 This is a schematic diagram of a matrix UV curing system for improving the flatness of optical fiber ribbons according to an embodiment of the present invention. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] Reference Figure 1 One embodiment of the present invention proposes a matrix ultraviolet curing method to improve the flatness of optical fiber ribbons. It adopts a closed-loop control method that monitors the internal stress and surface morphology of the optical fiber ribbon in real time and drives the programmable matrix ultraviolet light source for pixelated irradiation based on the monitoring results through a dynamic collaborative control algorithm. This method can actively manage the stress and deformation during the curing process, thereby improving the flatness of the final optical fiber ribbon.

[0037] The method described in this embodiment specifically includes:

[0038] Multiple optical fibers are arranged to form a geometric reference template with a preset tension deviation;

[0039] A pre-cured material is coated onto the geometric reference template to obtain a geometric reference template coated with matrix material.

[0040] The geometric reference template coated with matrix material is exposed for a short time using preset low-power ultraviolet light, which transforms the geometric reference template coated with matrix material into a geometrically locked structure, thereby generating an optical fiber ribbon.

[0041] By real-time monitoring of the in-situ stress of the internal stress and surface morphology of the optical fiber strip, the real-time stress distribution parameters and real-time flatness parameters of the optical fiber strip are obtained, and monitoring data is generated.

[0042] Feature extraction is performed on the monitoring data to generate a stress-flatness feature vector characterizing the current overall state of the optical fiber strip;

[0043] Based on the stress-flatness feature vector, a preset dynamic collaborative control algorithm is used to generate pixelated illumination control commands for driving the light source;

[0044] According to the pixelated irradiation control command, a preset programmable matrix ultraviolet light source composed of multiple independently addressable light-emitting units is controlled to perform matrix ultraviolet curing on the optical fiber ribbon, thereby generating a cured optical fiber ribbon.

[0045] Specifically, in-situ non-destructive monitoring technology is used to acquire key parameters of the internal stress distribution and external surface morphology of the optical fiber ribbon in real time during the curing process. These multi-dimensional physical quantities are then transformed into a stress-flatness feature vector that comprehensively characterizes the current state. This feature vector serves as the core input of a dynamic collaborative control algorithm, which calculates in real time the optimal curing energy distribution scheme to counteract and guide material deformation. Ultimately, the scheme is parsed into pixelated irradiation control commands, precisely driving a programmable matrix ultraviolet light source composed of numerous independently addressable light-emitting units to irradiate the optical fiber ribbon with a spatially non-uniform and temporally dynamically adjustable ultraviolet light field. This actively guides the curing process, making the stress distribution more uniform and suppressing uneven deformation. By introducing real-time monitoring and dynamic collaborative control during the curing process, refined and proactive management of the optical fiber ribbon's flatness is achieved. Compared with traditional uniform irradiation curing methods, this approach can predict and respond in real time to internal stress concentration and surface morphology fluctuations caused by material shrinkage and uneven tension. By utilizing a pixelated matrix ultraviolet light source, precise and differentiated curing energy can be applied to different regions of the fiber ribbon, effectively guiding the curing behavior of the material and counteracting deformation trends at the macroscopic level. This not only improves the surface smoothness of the fiber ribbon after final curing, reducing ripples and warping, but also optimizes the distribution of residual stress within the product, thereby improving the overall geometric accuracy and reliability of the fiber ribbon.

[0046] Optionally, the step of arranging multiple optical fibers to form a geometric reference template with a preset tension deviation includes:

[0047] Multiple optical fibers are subjected to preset independent tension to generate multi-channel tensioned optical fibers;

[0048] The multi-channel tension optical fibers are organized using a fiber splitting device to obtain a coplanar geometric reference template with uniform spacing.

[0049] Specifically, to achieve a matrix UV curing method for improving the flatness of optical fiber ribbons, multiple optical fibers first need to be arranged to form a geometric reference template with a preset tension deviation. This operation includes the following two key steps. First, for the prepared multiple optical fibers, a preset, independent tension force is applied to each of them. This is done to ensure that each fiber is under uniform and controllable preset tension, forming multi-path tensioned fibers. The magnitude and control of the tension are the foundation for achieving the final flatness, and its preset independence ensures that each fiber is under controllable stress. Next, using a specific fiber splitting device, these multi-path tensioned fibers treated with independent tension are precisely arranged and organized. The design of the fiber splitting device allows these fibers to be arranged on a common plane, maintaining a uniform and consistent spacing between them. In this way, a coplanar geometric reference template with uniform spacing is finally obtained. This template not only specifies the basic arrangement of optical fibers in the fiber ribbon, but also has a predetermined and beneficial initial effect on the subsequent curing process and the stress generated, laying a physical foundation for subsequent flatness improvement.

[0050] For example, to arrange multiple optical fibers into a geometric reference template with a preset tension deviation, the operator first prepares 24 optical fibers to be processed. These 24 fibers are then passed through 24 independent tension controllers. According to a preset tension deviation scheme, for example, the tension of fiber 12, located at the center, is set to 20 grams, decreasing sequentially towards both sides, with the tensions of fibers 1 and 24 set to 15 grams, thus creating a tension distribution that is high in the middle and low at the edges. Next, these 24 fibers with individually applied tension are guided into a precision fiber separator. This separator has 24 V-grooves spaced 250 micrometers apart. Under tension, each fiber automatically falls into its corresponding V-groove and is taut. Finally, all the fibers are arranged on the same plane with strictly consistent spacing, forming a coplanar, uniformly spaced geometric reference template.

[0051] Optionally, the generation of the fiber optic ribbon includes:

[0052] The irradiation intensity and exposure time of the low-power ultraviolet light are controlled to induce a preliminary gelation reaction on the surface of the ultraviolet-cured matrix material.

[0053] The preliminary gelation reaction forms a gel shell on the surface of the geometric reference template, resulting in a geometrically locked structure and generating an optical fiber ribbon.

[0054] Specifically, the irradiation intensity of low-power ultraviolet light is adjusted to a low level to avoid complete curing of the material and instead induce a chemical reaction on its surface. Simultaneously, the exposure time of the ultraviolet light is set to a short period. The combination of the specific irradiation intensity of the low-power ultraviolet light and the short exposure time triggers a preliminary gelation reaction on the surface of the matrix material coated on the geometric reference template. This preliminary gelation reaction results in the formation of a relatively stable gel shell on the surface of the matrix material. This gel shell effectively locks the optical fibers into a pre-defined coplanar, uniformly spaced geometric structure on the geometric reference template, ensuring precise fiber arrangement and generating an optical fiber ribbon with a geometrically locked structure.

[0055] For example, to generate a preliminary fiber ribbon structure, an operator transfers a geometric reference template coated with UV-curable matrix material to a low-power UV curing chamber. Here, the parameters of the UV light source need to be precisely controlled; for example, the UV irradiance is set to 5 milliwatts per square centimeter, and the exposure time is set to 2 seconds. This low-intensity, short-duration irradiation is insufficient to completely cure the material, but sufficient to trigger a photochemical reaction on its surface. Through this process, a thin and stable gel shell forms on the surface of the matrix material coated on the fiber. This gel shell acts like a mold, locking the fiber in a pre-defined, coplanar, and uniformly spaced precise structure on the geometric reference template, thus generating a preliminary fiber ribbon with a geometrically locked structure.

[0056] Optionally, the step of acquiring monitoring data includes:

[0057] The optical fiber strip is scanned by a preset optical coherence tomography sensor to obtain raw scan data containing depth and polarization information;

[0058] The original scan data is used to reconstruct images, generating a stress distribution image representing the internal stress distribution and a smoothness image representing the surface morphology.

[0059] Real-time stress distribution parameters are extracted from the stress distribution image, and real-time flatness parameters are extracted from the flatness image. The two are then combined to generate monitoring data.

[0060] Specifically, a pre-set optical coherence tomography (OCT) sensor is used to scan a geometrically locked fiber optic ribbon. This sensor is a special type of OCT device, specifically polarization-sensitive optical coherence tomography (PS-OCT), which measures the time-of-flight of light to obtain depth information and analyzes changes in the polarization state of light waves after penetrating materials. During the scanning process, the light beam emitted by the sensor penetrates the fiber optic ribbon and returns, recording raw scan data containing both depth and polarization information. Subsequently, image reconstruction is performed on this raw scan data. Image reconstruction involves two parallel processing paths. On one hand, the depth information is used to reconstruct the cross-sectional structure of the fiber optic ribbon, with its top contour accurately representing the surface morphology of the fiber optic ribbon, thus generating a flatness image. On the other hand, the polarization information is used to analyze the influence of the material's birefringence effect caused by internal stress on the polarization state of the light wave. By calculating parameters such as phase delay, the polarization change is converted into a visualized stress field distribution, thereby generating a stress distribution image. Finally, numerical real-time flatness parameters such as surface roughness, waviness, and maximum height difference are extracted from the generated flatness image; simultaneously, numerical real-time stress distribution parameters such as maximum stress value, average stress, and stress gradient are extracted from the stress distribution image. These two sets of parameters are combined to generate comprehensive monitoring data that includes the conditions inside and outside the fiber optic band.

[0061] For example, to acquire real-time monitoring data of an optical fiber ribbon, a polarization-sensitive optical coherence tomography (OCT) sensor is activated to scan the geometrically locked optical fiber ribbon. The sensor emits a low-coherence beam that scans the surface of the fiber ribbon and receives the reflected signals. This raw data, containing depth and polarization information, is processed to generate two images in parallel: one is a cross-sectional structure map of the fiber ribbon reconstructed using depth information, whose top contour visually demonstrates the surface flatness of the fiber ribbon; the other is a stress distribution map generated by analyzing the birefringence effect of the material using polarization information, where different colors represent the magnitude and direction of internal stress. Subsequently, an algorithm extracts a maximum height difference of 15 micrometers from the flatness image as a real-time flatness parameter, and a maximum stress value of 0.8 MPa from the stress distribution image as a real-time stress distribution parameter. Combining these two parameters forms a comprehensive set of monitoring data characterizing the current state of the optical fiber ribbon.

[0062] Optionally, the step of generating the stress-flatness feature vector includes:

[0063] The parameters with different physical dimensions in the monitoring data are normalized to generate standardized monitoring data;

[0064] Principal component analysis was performed on the standardized monitoring data to extract the main characteristic components used to characterize the main direction of data change;

[0065] The main feature components are combined to construct a stress-flatness feature vector.

[0066] Specifically, parameters with different physical dimensions in the monitoring data are normalized. Since the monitoring data includes real-time stress distribution parameters in Pascals and real-time smoothness parameters in micrometers, these two parameters differ significantly in numerical scale, and direct use would affect the accuracy of subsequent analyses. Therefore, a standardization method, such as Z-score standardization, is used to process each parameter and generate dimensionless standardized monitoring data:

[0067] ;

[0068] in, Here, is the normalized parameter value, x is the original parameter measurement value in the monitoring data, μ is the average value of the parameter over a period of time or within a batch, and σ is the corresponding standard deviation. Both μ and σ can be obtained through statistical calculations of continuously collected monitoring data. After normalization, all parameters are transformed into comparable standardized monitoring data. Next, principal component analysis (PCA) is performed on the standardized monitoring data. PCA is a statistical dimensionality reduction technique that aims to identify the key directions that explain the greatest data variability among numerous interrelated parameters—the principal characteristic components. Through this analysis, high-dimensional standardized monitoring data can be projected onto a few mutually orthogonal principal characteristic components, which condense the core information in the original data. Finally, these extracted principal characteristic components are combined according to their importance to construct a low-dimensional but information-rich vector, namely the stress-smoothness characteristic vector. The stress-smoothness characteristic vector comprehensively describes the overall stress and smoothness state of the fiber optic strip at the current moment in a compact form. See [link to stress-smoothness characteristic vector construction process] for details. Figure 2 The original parameters are normalized to eliminate dimensional differences, and principal component analysis is used to extract core features, ultimately forming a low-dimensional, high-information-density feature vector, which provides a basis for the generation of subsequent control commands.

[0069] For example, the monitoring data is normalized. Because the units for real-time stress distribution parameters (MPa) and real-time flatness parameters (micrometers) are different, they cannot be directly compared. Therefore, the Z-score normalization method is used. The difference between the original measurement value and the historical average is divided by the standard deviation of the historical data to obtain a normalized value representing how much the current value deviates from the average. For example, calculations of the 100 most recently produced fiber optic strips show that the average maximum height difference is 12 micrometers and the average maximum stress value is 0.7 MPa. This is the standard deviation corresponding to that parameter. For example, the standard deviation of the batch mentioned above is 2 micrometers and 0.1 megapascals. μ and All values ​​can be obtained through statistical calculations on continuously collected monitoring data. For example, for a flatness parameter of 15 micrometers, its normalized value is (15-12) / 2=1.5. For a stress parameter of 0.8 MPa, its normalized value is (0.8-0.7) / 0.1=1.0. In this way, parameters with different dimensions are converted into comparable dimensionless values. After normalization, the system performs principal component analysis on all standardized monitoring data to identify two main characteristic components that can explain the data variability. Finally, these two main characteristic components are combined to construct a two-dimensional stress-flatness feature vector, such as [1.5, 1.0]. This vector compactly and comprehensively describes the current stress and flatness status of the fiber optic strip.

[0070] Optionally, the step of generating pixelated illumination control instructions includes:

[0071] Based on the stress-smoothness feature vector, phase-modulated curing path parameters are generated to guide the curing reaction sequence;

[0072] Based on the phase modulation curing path parameters, a dynamic scanning irradiation pattern that varies with time is generated;

[0073] Based on the dynamic scanning illumination mode, pixelated illumination control commands are generated through the dynamic collaborative control algorithm.

[0074] Specifically, using the stress-flatness feature vector, a preset dynamic collaborative control algorithm generates phase-modulated curing path parameters to guide the entire curing reaction sequence. These parameters define the reaction activation sequence and rate in different regions during curing, aiming to predict and compensate for potential stress distribution to achieve the final flatness target. For example, if the feature vector indicates potential tensile stress causing bulging in a certain region, the phase modulation parameters may instruct that region to be cured later or at a specific rate to guide the material to shrink towards the target state. Next, based on the phase-modulated curing path parameters, a time-varying dynamic scanning irradiation pattern is generated. This pattern describes the irradiation requirements of different regions on the fiber ribbon surface at different time points throughout the curing cycle, including the division of the irradiation area, the distribution of irradiation intensity over time, and the scanning sequence, aiming to macroscopically regulate the overall stress balance and deformation by precisely controlling the local curing state. Based on the dynamic scanning irradiation pattern, a preset dynamic collaborative control algorithm is used to calculate specific pixel-level control signals according to the pattern. This process transforms the region, intensity, and timing information in the dynamic scanning pattern into specific instructions required to control each independent emitting unit of the programmable matrix ultraviolet light source. These instructions are output in the form of pixelated irradiation control instructions, which include the precise address of each light-emitting unit, the desired irradiation intensity, and the time signal to control its switching or brightness, ensuring that the fiber ribbon can be dynamically and collaboratively irradiated throughout the curing process, thereby precisely guiding its deformation.

[0075] For example, to generate pixelated illumination control instructions, firstly, based on the previously generated stress-flatness feature vector [1.5, 1.0], a set of phase-modulated curing path parameters is generated through a dynamic collaborative control algorithm. These parameters indicate that, due to large flatness deviations, the curing process should proceed gradually from the edge to the center to balance the stress. Based on this, a dynamic scanning illumination pattern is generated, such as a spiral scanning pattern that contracts from the outside in, and a function is defined for the illumination intensity of each region over time. Finally, this dynamic pattern is converted into specific pixel-level instructions. For example, the instructions might specify that at t=0 seconds, the emitting unit at coordinate (10, 10) is turned on with 50% intensity; at t=0.1 seconds, the emitting unit at coordinate (11, 10) is turned on with 50% intensity, while the intensity of the emitting unit at (10, 10) decreases to 30%. This set of instructions, containing address, intensity, and time signals, constitutes the final pixelated illumination control instructions.

[0076] Optionally, the dynamic cooperative control algorithm includes:

[0077] Based on the dynamic scanning irradiation mode, the irradiation area, intensity gradient, and scanning timing parameters that change over time in the mode are analyzed.

[0078] Establish a spatial mapping relationship between the dynamic scanning irradiation mode and the light-emitting unit array of the programmable matrix ultraviolet light source, map the irradiation area to the physical coordinates of the light-emitting unit, and generate a spatial pixelation mapping table;

[0079] Based on the spatial pixelation mapping table, the intensity gradient of the dynamic scanning illumination mode is decomposed into temporal and spatial values ​​to generate an initial illumination intensity sequence for each luminescent unit.

[0080] Based on the stress-smoothness feature vector, the deviation between the current stress-smoothness feature vector and the preset target feature vector is calculated, and the initial irradiation intensity sequence is collaboratively corrected according to the deviation value.

[0081] The corrected illumination parameters of each light-emitting unit are encapsulated into pixelated illumination control instructions, which include the address code, intensity value and time control signal of each light-emitting unit.

[0082] Specifically, the dynamic scanning irradiation pattern is analyzed, including accurately identifying the time-varying irradiation areas defined in the pattern, clarifying the gradient changes of irradiation intensity in different areas, and extracting detailed scanning timing parameters, such as the start time, duration, and scanning order of each area. A spatial mapping relationship is established between the dynamic scanning irradiation pattern and the array of light-emitting units in a preset programmable matrix ultraviolet light source composed of multiple independently addressable light-emitting units. Through this mapping, the irradiation areas defined in the pattern are accurately mapped to the coordinate positions of the light-emitting units in the physical array. This process generates a spatial pixelated mapping table, which records in detail the correspondence between each abstract irradiation area in the pattern and the specific light-emitting units in the matrix light source. Based on the spatial pixelated mapping table, the algorithm performs spatiotemporal decomposition of the intensity gradient information in the dynamic scanning irradiation pattern. For each light-emitting unit in the matrix light source, an initial irradiation intensity sequence is calculated based on its corresponding irradiation area and the intensity information in the pattern, describing the change of the preset irradiation intensity of the light-emitting unit over time throughout the curing cycle. Based on the current stress-flatness feature vector, the deviation between it and the preset target feature vector representing the ideal stress and flatness state is calculated. The deviation value quantifies the difference between the current fiber ribbon state and the target state. Using this deviation value, the previously generated initial illumination intensity sequence is collaboratively corrected to produce a corrected illumination intensity sequence. The purpose of this correction is to enable the illumination strategies on different light-emitting units to work synergistically. By precisely controlling the spatial distribution and temporal cumulative effect of light intensity, the redistribution of stress within the fiber ribbon is guided to minimize the deviation and approach the preset flatness target. The corrected illumination parameters for each light-emitting unit, including its address encoding (e.g., row and column coordinates), the corrected intensity value, and necessary time control signals, are encapsulated and integrated into pixelated illumination control instructions. These instructions can directly drive each individual light-emitting unit in the programmable matrix ultraviolet light source, precisely controlling its illumination intensity and duration at specific times.

[0083] For example, suppose the dynamic collaborative control algorithm receives a dynamic scanning illumination pattern that defines a scanning process from left to right, in two rows. Parsing this pattern, four time-varying illumination regions are identified: upper left region A, upper right region B, lower left region C, and lower right region D. The start time, duration, and scanning order of each region are determined. Next, the algorithm establishes a spatial mapping relationship between this pattern and a 2x2 programmable matrix UV light source, generating a spatial pixelated mapping table. Region A is mapped to the emitting unit (1,1), region B to (1,2), region C to (2,1), and region D to (2,2). Based on this mapping table, the algorithm generates an initial illumination intensity sequence for each emitting unit; for example, at the start of the curing cycle, the initial illumination intensity of all four emitting units is set to 50%. At this point, the algorithm obtains the current stress-flatness feature vector as [1.5, 1.0], while the preset target feature vector is [0.0, 0.0]. To quantify the difference between the current state and the ideal state, the algorithm calculates the deviation value between them. The Euclidean distance formula is used for calculation here:

[0084] ;

[0085] In this formula, D is the calculated deviation value. and These are the two components of the current stress-smoothness eigenvector. and These are the two components of the target feature vector. These components are dimensionless normalized values, therefore the calculated deviation value D is also dimensionless. Substituting the values ​​into the calculation... Subsequently, the algorithm uses this deviation value to collaboratively correct the initial irradiation intensity sequence, using the following correction formula:

[0086] ;

[0087] in, This is the corrected irradiation intensity. Here, D is the initial irradiation intensity, D is the previously calculated deviation value, and K is a preset correction coefficient in percentage form, used to convert the dimensionless deviation value into an adjustable intensity percentage. For example, if K is set to 10%, the corrected irradiation intensity will be... The corrected illumination parameters for each light-emitting unit are encapsulated to generate the final pixelated illumination control instructions, such as: address (1,1), intensity 32%, time t=0; address (1,2), intensity 32%, time t=0.1 seconds; address (2,1), intensity 32%, time t=0.2 seconds; address (2,2), intensity 32%, time t=0.3 seconds. These instructions will be sent to the programmable matrix ultraviolet light source to perform precise curing operations.

[0088] Optionally, the step of controlling the programmable matrix ultraviolet light source for matrix ultraviolet curing includes:

[0089] The pixelated illumination control command is analyzed to obtain a spatialized illumination intensity distribution and a temporally sequenced illumination time series.

[0090] Based on the irradiation intensity distribution and the irradiation time sequence, an independent driving signal is sent to each light-emitting unit of the programmable matrix ultraviolet light source;

[0091] During the irradiation process, the drive signal is adjusted in real time according to the continuously updated pixelated irradiation control command to dynamically maintain the flatness of the fiber ribbon.

[0092] Specifically, the pixelated illumination control commands are parsed. During parsing, each data packet in the command contains a unique identifier (address code) for an addressable light-emitting unit, along with the specific illumination intensity value it needs to withstand and the corresponding illumination time control signal. These commands are converted into two key pieces of information: first, a spatialized illumination intensity distribution, i.e., the specific intensity superposition required to illuminate different areas of the fiber ribbon surface at a given moment; and second, a temporally sequenced illumination time series, i.e., the specific plan for how the illumination intensity of each light-emitting unit evolves over time during the entire curing process. Based on the parsed spatialized illumination intensity distribution and temporally sequenced illumination time series, independent drive signals are sent to each light-emitting unit of the programmable matrix ultraviolet light source. The drive signals directly control the light output of each light-emitting unit, including its on / off state, brightness level, and operation for the specific time length required by the command, thereby forming a complex, pixelated illumination pattern on the fiber ribbon surface. Throughout the illumination process, the drive signals sent to each light-emitting unit are adjusted in real time based on the continuously updated pixelated illumination control commands. In this way, the flatness of the fiber ribbon can be dynamically maintained during the curing process. For example, if a deformation trend is detected in a certain area due to uneven curing, the control system will immediately receive a new, corrected instruction and adjust the irradiation intensity or time of the corresponding light-emitting unit accordingly to counteract or compensate for this deformation.

[0093] For example, to control the programmable matrix UV light source to complete the matrix UV curing of the fiber ribbon, the previously generated pixelated irradiation control instructions are first parsed. For example, the instruction set contains multiple data packets with similar addresses (10,10), 50% intensity, and a duration of 0.1 seconds. These instructions are converted into a spatialized irradiation intensity distribution map and a time-sequential control flow. Subsequently, an independent drive signal is sent to each light-emitting unit of the programmable matrix UV light source. For example, it sends a drive voltage to light-emitting unit (10,10) to make it emit light at 50% intensity for 0.1 seconds. During irradiation, if the PS-OCT sensor detects abnormal stress accumulation near the (10,10) region in real time, it will immediately receive an updated instruction to adjust the intensity of light-emitting unit (10,10) to 30% to dynamically compensate for the deformation trend in that region, thereby dynamically maintaining the flatness of the fiber ribbon throughout the curing process.

[0094] Optionally, the method further includes:

[0095] After curing, the flatness of the optical fiber ribbon is tested to generate test results characterizing the quality of the final product;

[0096] The detection results are compared with a preset quality target value to generate an error signal;

[0097] Based on the error signal, the internal parameters of the dynamic cooperative control algorithm are adjusted to generate optimized control algorithm parameters.

[0098] Specifically, high-precision surface topography measurement equipment is used to inspect the flatness of the fully cured optical fiber ribbon, generating a set of quantitative inspection results that characterize the final product quality, such as the average surface roughness and maximum profile peak-to-valley difference of the optical fiber ribbon. These inspection results, containing actual quality data, are compared with a pre-set quality target value representing the ideal product specifications. Through this comparison, a clear error signal can be calculated. This error signal can be expressed by the following formula:

[0099] ;

[0100] Where E is the calculated error signal. These are the actual product quality parameter values ​​obtained through flatness testing, while These are preset ideal quality target values. These two parameters have the same physical dimension, such as micrometers, so they can be directly subtracted. Based on the magnitude and direction of the error signal, the internal parameters of the dynamic cooperative control algorithm are fine-tuned. These internal parameters include weighting coefficients used to calculate pixelated illumination control commands, response gain, or coefficients of the prediction model, generating optimized control algorithm parameters.

[0101] For example, to continuously optimize the entire matrix UV curing method, after producing a batch of fiber optic ribbons, operators use a high-precision laser profilometer to perform a final flatness check on the cured ribbons. The check results show that the average surface roughness of this batch of fiber optic ribbons is 12 micrometers. This actual measurement result is compared with the preset quality target value of 10 micrometers, and an error signal is calculated. The formula for calculating this error signal is the difference between the actual product quality parameter value and the preset ideal quality target value. For example, It is 12 micrometers. If the error is 10 micrometers, then the error signal E is 12-10=2 micrometers. This positive error signal indicates that the currently produced product is rougher than expected. This error signal is then input into the adaptive adjustment module. Based on this error, the module fine-tunes the weighting coefficients in the dynamic collaborative control algorithm, for example, slightly reducing the overall irradiation intensity during the initial curing stage to slow down the material shrinkage rate. This optimized parameter will be saved and applied to the production of the next batch of fiber ribbons to obtain a smoother product. See [link to documentation] for feedback learning and parameter optimization results. Figure 3 After 10 batches of optimization, the actual quality parameters decreased from 1.2μm to 0.6μm, the error approached 0, and the control parameters stabilized at the optimal value, confirming the continuous effect of closed-loop optimization on improving smoothness.

[0102] Reference Figure 4 Based on the same inventive concept, the present invention also provides a matrix UV curing system for improving the flatness of optical fiber ribbons, the system comprising:

[0103] The geometric reference template construction module is used to organize multiple optical fibers to form a geometric reference template with a preset tension deviation;

[0104] A matrix material coating module is used to coat a preset curing material onto the geometric reference template to obtain a geometric reference template coated with matrix material.

[0105] The geometric locking structure generation module is used to expose the geometric reference template coated with matrix material to preset low-power ultraviolet light for a short time, transforming the geometric reference template coated with matrix material into a geometric locking structure and generating an optical fiber ribbon;

[0106] The in-situ stress and flatness monitoring module is used to obtain the real-time stress distribution parameters and real-time flatness parameters of the optical fiber by real-time monitoring of the in-situ stress of the internal stress and surface morphology of the optical fiber, and to generate monitoring data.

[0107] The stress-flatness feature extraction and decision module is used to extract features from the monitoring data and generate a stress-flatness feature vector characterizing the current comprehensive state of the optical fiber strip.

[0108] The dynamic collaborative control command generation module is used to generate pixelated illumination control commands for driving the light source based on the stress-flatness feature vector and a preset dynamic collaborative control algorithm.

[0109] The matrix UV curing execution module is used to control a preset programmable matrix UV light source composed of multiple independently addressable light-emitting units to perform matrix UV curing on the optical fiber ribbon according to the pixelated irradiation control command, so as to generate a cured optical fiber ribbon.

[0110] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0111] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A matrix UV curing method for improving the flatness of optical fiber ribbons, characterized in that: Multiple optical fibers are arranged to form a geometric reference template with a preset tension deviation; A pre-cured material is coated onto the geometric reference template to obtain a geometric reference template coated with matrix material. The geometric reference template coated with matrix material is exposed for a short time using preset low-power ultraviolet light to transform the geometric reference template coated with matrix material into a geometrically locked structure, thereby generating a preliminary fiber ribbon. By real-time monitoring of the in-situ stress of the internal stress and surface morphology of the preliminary optical fiber strip, the real-time stress distribution parameters and real-time flatness parameters of the preliminary optical fiber strip are obtained, and monitoring data is generated. Feature extraction is performed on the monitoring data to generate a stress-flatness feature vector characterizing the current overall state of the preliminary optical fiber strip; Based on the stress-flatness feature vector, a preset dynamic collaborative control algorithm is used to generate pixelated illumination control commands for driving the light source; According to the pixelated irradiation control command, a preset programmable matrix ultraviolet light source composed of multiple independently addressable light-emitting units is controlled to perform matrix ultraviolet curing on the preliminary fiber ribbon to generate the final cured fiber ribbon.

2. The matrix UV curing method for improving the flatness of optical fiber ribbons according to claim 1, characterized in that, The step of organizing multiple optical fibers to form a geometric reference template with a preset tension deviation includes: Multiple optical fibers are subjected to preset independent tension to generate multi-channel tensioned optical fibers; The multi-channel tension optical fibers are organized using a fiber splitting device to obtain a coplanar geometric reference template with uniform spacing.

3. The matrix UV curing method for improving the flatness of optical fiber ribbons according to claim 1, characterized in that, The generation of the initial fiber ribbon includes: The irradiation intensity and exposure time of the low-power ultraviolet light are controlled to induce a preliminary gelation reaction in the matrix material. The preliminary gelation reaction forms a gel shell on the surface of the geometric reference template, resulting in a geometrically locked structure and generating a preliminary optical fiber ribbon.

4. The matrix UV curing method for improving the flatness of optical fiber ribbons according to claim 1, characterized in that, The generated monitoring data includes: The preliminary fiber optic band is scanned by a preset optical coherence tomography sensor to obtain raw scan data containing depth and polarization information; The original scan data is used to reconstruct images, generating a stress distribution image representing the internal stress distribution and a smoothness image representing the surface morphology. Real-time stress distribution parameters are extracted from the stress distribution image, and real-time flatness parameters are extracted from the flatness image. The two are then combined to generate monitoring data.

5. The matrix UV curing method for improving the flatness of optical fiber ribbons according to claim 1, characterized in that, The generation of the stress-flatness feature vector characterizing the current synthetic state of the preliminary fiber strip includes: The parameters with different physical dimensions in the monitoring data are normalized to generate standardized monitoring data; Principal component analysis was performed on the standardized monitoring data to extract the main characteristic components used to characterize the main direction of data change; The main feature components are combined to construct a stress-flatness feature vector.

6. The matrix UV curing method for improving the flatness of optical fiber ribbons according to claim 1, characterized in that, The generation of pixelated illumination control commands for driving the light source includes: Based on the stress-smoothness feature vector, phase-modulated curing path parameters are generated to guide the curing reaction sequence; Based on the phase modulation curing path parameters, a dynamic scanning irradiation pattern that varies with time is generated; Based on the dynamic scanning illumination mode, pixelated illumination control commands are generated through the dynamic collaborative control algorithm.

7. The matrix UV curing method for improving the flatness of optical fiber ribbons according to claim 6, characterized in that, The dynamic cooperative control algorithm includes: Based on the dynamic scanning irradiation mode, the irradiation area, intensity gradient, and scanning timing parameters that change over time in the mode are analyzed. Establish a spatial mapping relationship between the dynamic scanning irradiation mode and the light-emitting unit array of the programmable matrix ultraviolet light source, map the irradiation area to the physical coordinates of the light-emitting unit, and generate a spatial pixelation mapping table; Based on the spatial pixelation mapping table, the intensity gradient of the dynamic scanning illumination mode is decomposed into temporal and spatial values ​​to generate an initial illumination intensity sequence for each luminescent unit. Based on the stress-smoothness feature vector, the deviation between the current stress-smoothness feature vector and the preset target feature vector is calculated, and the initial irradiation intensity sequence is collaboratively corrected according to the deviation value. The corrected illumination parameters of each light-emitting unit are encapsulated into pixelated illumination control instructions, which include the address code, intensity value and time control signal of each light-emitting unit.

8. The matrix UV curing method for improving the flatness of optical fiber ribbons according to claim 1, characterized in that, The control of a preset programmable matrix ultraviolet light source composed of multiple independently addressable light-emitting units to perform matrix ultraviolet curing on the preliminary fiber ribbon includes: The pixelated illumination control command is analyzed to obtain a spatialized illumination intensity distribution and a temporally sequenced illumination time series. Based on the irradiation intensity distribution and the irradiation time sequence, an independent driving signal is sent to each light-emitting unit of the programmable matrix ultraviolet light source; During the irradiation process, the drive signal is adjusted in real time according to the continuously updated pixelated irradiation control command to dynamically maintain the flatness of the fiber ribbon.

9. The matrix UV curing method for improving the flatness of optical fiber ribbons according to claim 1, characterized in that, The method further includes: After curing, the flatness of the optical fiber ribbon is tested to generate test results characterizing the quality of the final product; The detection results are compared with a preset quality target value to generate an error signal; Based on the error signal, the internal parameters of the dynamic cooperative control algorithm are adjusted to generate optimized control algorithm parameters.

10. A matrix UV curing system for improving the flatness of optical fiber ribbons, applied to the matrix UV curing method for improving the flatness of optical fiber ribbons as described in any one of claims 1-9, characterized in that, The system includes: The geometric reference template construction module is used to organize multiple optical fibers to form a geometric reference template with a preset tension deviation; A matrix material coating module is used to coat a preset curing material onto the geometric reference template to obtain a geometric reference template coated with matrix material. The geometric locking structure generation module is used to expose the geometric reference template coated with matrix material to preset low-power ultraviolet light for a short time, transforming the geometric reference template coated with matrix material into a geometric locking structure and generating a preliminary fiber ribbon; The in-situ stress and flatness monitoring module is used to obtain the real-time stress distribution parameters and real-time flatness parameters of the fiber ribbon by real-time monitoring of the in-situ stress of the internal stress and surface morphology of the preliminary fiber ribbon, and to generate monitoring data. The stress-flatness feature extraction and decision module is used to extract features from the monitoring data and generate a stress-flatness feature vector characterizing the current comprehensive state of the optical fiber strip. The dynamic collaborative control command generation module is used to generate pixelated illumination control commands for driving the light source based on the stress-flatness feature vector and a preset dynamic collaborative control algorithm. The matrix UV curing execution module is used to control a preset programmable matrix UV light source composed of multiple independently addressable light-emitting units to perform matrix UV curing on the optical fiber ribbon according to the pixelated irradiation control command, so as to generate the final cured optical fiber ribbon.

Images