An on-line monitoring method, device, storage medium and equipment for the loss of single-crystal optical fiber

Abstract

This specification discloses a method, device, storage medium and equipment for on-line monitoring of the loss of single crystal optical fiber. During the process of preparing single crystal optical fiber, a sequence of target melting zone images of the single crystal optical fiber being drawn within a specified time period is collected by a preset image acquisition device. Furthermore, a preset monitoring model can be used to on-line monitor the quality of the single crystal optical fiber being drawn based on the obtained sequence of target melting zone images, and when an abnormality is detected in the drawn single crystal optical fiber, an alarm message can be sent so that the equipment for drawing single crystal optical fiber can make timely production adjustments according to the alarm message, thereby improving the yield of single crystal optical fiber and further reducing the cost of preparing single crystal optical fiber.

Description

Technical Field

[0001] This specification relates to the technical field of optical fibers, and particularly to an on-line monitoring method, device, storage medium and equipment for the loss of single-crystal optical fibers. Background Art

[0002] Single-crystal optical fiber is a single-crystal material in the form of an optical fiber. It not only has excellent physical, chemical and optical properties of bulk crystal materials, but also has unique high heat dissipation efficiency and the ability to achieve long-distance total reflection waveguides. These properties make single-crystal optical fibers widely used in aerospace and other fields.

[0003] Generally, the transmission loss of single-crystal optical fibers directly affects the output power of fiber lasers, the detection sensitivity of fiber sensors, and the transmission distance of fiber communication systems. Low transmission loss means that the energy loss of the optical signal during transmission in the single-crystal optical fiber is small, and efficient optical transmission over a longer distance can be achieved. Therefore, the transmission loss of single-crystal optical fibers is crucial for applications in high-power laser transmission, mid-infrared laser technology, high-energy ray detection, high-temperature sensing and other fields.

[0004] In the process of preparing single-crystal optical fibers by the Laser heated pedestal growth (LHPG) method, the laser beam can be precisely focused on the top of the feed rod by using a laser heating system. This focused laser beam can melt the top of the feed rod to form a molten region. Subsequently, the feed rod is pulled upward at a controlled speed by a motor drive device. During this process, the molten material in the molten region is stretched into a slender fiber. With the continuous pulling of the feed rod and the stable heating of the laser beam, the molten material gradually solidifies, and finally an optical fiber with a single-crystal structure is formed. During this process, due to the influence of factors such as environmental vibration, laser power fluctuation, and raw material uniformity, the diameter of the drawn optical fiber will fluctuate, and the transmission losses in different regions will also be different. That is, there may be local optical fibers in the drawn single-crystal optical fiber with relatively large transmission losses due to impurity contamination or insufficiently smooth surface. The transmission loss of the single-crystal optical fiber can only be measured after the entire optical fiber is pulled and removed from the system, rather than being measured in real time during the preparation process. In order to maintain the quality of the entire optical fiber when it is found that there is an abnormality in the drawing of the single-crystal optical fiber, the drawing is stopped at the abnormal part of the single-crystal optical fiber, a part of the optical fiber and the feed rod are cut off, and then the drawing is continued. As a result, the prepared single-crystal optical fiber may need to be re-prepared due to relatively large transmission losses, which in turn leads to a relatively high preparation cost of the single-crystal optical fiber.

[0005] Therefore, how to reduce the cost of preparing single-crystal optical fibers is an urgent problem to be solved. Summary of the Invention

[0006] This specification provides an online monitoring method, device, storage medium, and equipment for the loss of single-crystal optical fiber to partially solve the above problems existing in the prior art.

[0007] This specification adopts the following technical solutions:

[0008] This specification provides an online monitoring method for the loss of single-crystal optical fiber, including:

[0009] Obtain a target melting zone image sequence, where each target melting zone image in the target melting zone image sequence is an image of the melting zone of the raw material rod used for preparing the single-crystal optical fiber collected during the preparation of the single-crystal optical fiber. The target melting zone image contains bright lines formed due to the penetration of the preset illumination light inside the drawn optical fiber. Different target melting zone images are collected at different times within a specified time period;

[0010] Input the target melting zone image sequence into a pre-trained monitoring model, so as to determine the target melting zone characteristics according to the target melting zone image sequence through the monitoring model, and estimate the transmission loss value of the single-crystal optical fiber drawn within the specified time period according to the target melting zone characteristics. The target melting zone characteristics are used to reflect the uniformity of the bright lines contained in the target melting zone image, the diameter change of the drawn optical fiber, and the shape change of the melting zone;

[0011] If it is determined that the transmission loss value meets the preset alarm condition, an alarm message is generated and the alarm message is uploaded to execute tasks according to the alarm message.

[0012] Optionally, the monitoring model includes: a convolution module;

[0013] Inputting the target melting zone image sequence into a pre-trained monitoring model to determine the target melting zone characteristics according to the target melting zone image sequence through the monitoring model specifically includes:

[0014] Input the target melting zone image sequence into a pre-trained monitoring model, and through the convolution module, perform convolution processing on each target melting zone image in the target melting zone image sequence to extract the image convolution characteristics corresponding to the target melting zone image from the target melting zone image;

[0015] Determine the target melting zone characteristics according to the image convolution characteristics corresponding to each target melting zone image.

[0016] Optionally, the monitoring model includes: a sequence feature extraction module;

[0017] Determining the target melting zone characteristics according to the image convolution characteristics corresponding to each target melting zone image specifically includes:

[0018] According to the chronological order of the acquisition times of the target melt zone images, the image convolution features corresponding to the target melt zone images are sequentially input into the sequence feature extraction module, so that the sequence feature extraction module performs several rounds of processing based on the image convolution features corresponding to the target melt zone images to obtain the sequence features corresponding to the target melt zone image sequence as the target melt zone features; wherein,

[0019] For each round of processing, determine the image convolution feature of the target melt zone image input into the sequence feature extraction module in this round of processing as the target image convolution feature, and obtain the candidate target melt zone feature obtained after the previous round of processing. Fuse the target image convolution feature and the candidate target melt zone feature to obtain the initial candidate target melt zone feature in this round of processing, and screen the initial candidate target melt zone feature through a preset forgetting function to obtain the candidate target melt zone feature after this round of processing until it is determined that the preset termination condition is met. At this time, the sequence features corresponding to the target melt zone image sequence are obtained as the target melt zone features.

[0020] Optionally, training the monitoring model specifically includes:

[0021] Obtain an initial monitoring model, a sample melt zone image sequence, and the actual transmission loss value of the single crystal optical fiber drawn during the time period corresponding to the sample melt zone image sequence;

[0022] Input the sample melt zone image sequence into the initial monitoring model, so that the initial monitoring model determines the sample melt zone features according to the sample melt zone image sequence, and estimates the transmission loss value of the single crystal optical fiber drawn during the time period corresponding to the sample melt zone image sequence as the estimated transmission loss value;

[0023] Determine the target loss according to the deviation between the estimated transmission loss value and the actual transmission loss value, and train the initial monitoring model with the goal of minimizing the target loss to obtain the monitoring model. The target loss is positively correlated with the deviation between the estimated transmission loss value and the actual transmission loss value.

[0024] Optionally, the actual transmission loss value of the single crystal optical fiber drawn during the time period corresponding to the sample melt zone image sequence is measured and calculated by a spectrometer connected to the end of the single crystal optical fiber drawn during the time period corresponding to the sample melt zone image sequence.

[0025] This specification provides an on-line monitoring device for the loss of single crystal optical fiber, including:

[0026] An acquisition module for acquiring a target melt zone image sequence, where each target melt zone image in the target melt zone image sequence is an image of the melt zone of a raw material rod used for preparing a single crystal optical fiber, which is collected during the process of preparing the single crystal optical fiber. The target melt zone image contains bright lines formed due to the penetration of a preset illumination light inside the drawn optical fiber. Different target melt zone images are collected at different times within a specified time period;

[0027] A prediction module for inputting the target melt zone image sequence into a pre-trained monitoring model, so as to determine target melt zone features according to the target melt zone image sequence through the monitoring model, and estimate the transmission loss value of the single crystal optical fiber drawn within the specified time period according to the target melt zone features. The target melt zone features are used to reflect the uniformity of the bright lines contained in the target melt zone image, the diameter change of the drawn optical fiber, and the shape change of the melt zone;

[0028] An execution module for generating an alarm message when it is determined that the transmission loss value meets a preset alarm condition, and uploading the alarm message to perform a task according to the alarm message.

[0029] Optionally, the monitoring model includes: a convolution module;

[0030] Specifically, the prediction module is used to input the target melt zone image sequence into a pre-trained monitoring model, so as to perform convolution processing on each target melt zone image in the target melt zone image sequence through the convolution module to extract image convolution features corresponding to the target melt zone image from the target melt zone image; determine target melt zone features according to the image convolution features corresponding to each target melt zone image.

[0031] Optionally, the monitoring model includes: a sequence feature extraction module;

[0032] Specifically, the prediction module is configured to, according to the chronological order of the acquisition times of the target melt zone images, sequentially input the image convolution features corresponding to the target melt zone images into the sequence feature extraction module, so that the sequence feature extraction module performs several rounds of processing based on the image convolution features corresponding to the target melt zone images to obtain the sequence features corresponding to the target melt zone image sequence as the target melt zone features. Wherein, for each round of processing, the image convolution feature of the target melt zone image input into the sequence feature extraction module in this round of processing is determined as the target image convolution feature, and the candidate target melt zone feature obtained after the previous round of processing is acquired. The target image convolution feature and the candidate target melt zone feature are fused to obtain the initial candidate target melt zone feature in this round of processing, and the initial candidate target melt zone feature is screened through a preset forgetting function to obtain the candidate target melt zone feature after this round of processing until it is determined that the preset termination condition is met. At this time, the sequence features corresponding to the target melt zone image sequence are obtained as the target melt zone features.

[0033] This specification provides a computer-readable storage medium storing a computer program, which when executed by a processor, implements the above-mentioned on-line monitoring method for the loss of single-crystal optical fiber.

[0034] This specification provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the above-mentioned on-line monitoring method for the loss of single-crystal optical fiber.

[0035] At least one of the above technical solutions adopted in this specification can achieve the following beneficial effects:

[0036] In the on-line monitoring method for the loss of single-crystal optical fiber provided in this specification, a target melt zone image sequence is acquired, where the target melt zone image sequence includes target melt zone images, which are images of the melt zone of the raw material rod for preparing the single-crystal optical fiber collected during the preparation of the single-crystal optical fiber. The target melt zone image contains bright lines formed by the penetration of the preset illumination light inside the drawn optical fiber. Different target melt zone images are collected at different times within a specified time period. The target melt zone image sequence is input into a pre-trained monitoring model, so that the monitoring model determines the target melt zone features according to the target melt zone image sequence, and estimates the transmission loss value of the single-crystal optical fiber drawn within the specified time period based on the target melt zone features. The target melt zone features are used to reflect the uniformity of the bright lines contained in the target melt zone image, the diameter change of the drawn optical fiber, and the shape change of the melt zone. If it is determined that the transmission loss value meets the preset alarm condition, an alarm message is generated and uploaded to perform tasks according to the alarm message.

[0037] As can be seen from the above method, during the process of preparing single-crystal optical fiber, a sequence of target molten zone images of the single-crystal optical fiber being drawn within a specified time period can be collected by a preset image acquisition device. Furthermore, a preset monitoring model can be used to online monitor the quality of the single-crystal optical fiber being drawn based on the obtained sequence of target molten zone images. When an abnormality is detected in the drawn single-crystal optical fiber, an alarm message can be sent, so that the device for drawing single-crystal optical fiber can make production adjustments in a timely manner according to the alarm message, thereby improving the yield of single-crystal optical fiber and reducing the cost of preparing single-crystal optical fiber. Description of the Drawings

[0038] The drawings described herein are used to provide a further understanding of the present specification and form a part of the present specification. The schematic embodiments of the present specification and their descriptions are used to explain the present specification and do not constitute an improper limitation to the present specification. In the drawings:

[0039] Figure 1 is a schematic diagram of the principle of the LHPG system provided in the present specification;

[0040] Figure 2 is a schematic flowchart of an online monitoring method for the loss of single-crystal optical fiber provided in the present specification;

[0041] Figure 3 is a schematic diagram of the target molten zone image provided in the present specification;

[0042] Figure 4 is a schematic diagram of the determination process of the transmission loss value of the single-crystal optical fiber provided in the present specification;

[0043] Figure 5 is a schematic diagram of the measurement process of the actual transmission loss value provided in the present specification;

[0044] Figure 6 is a schematic diagram of an online monitoring device for the loss of single-crystal optical fiber provided in the present specification;

[0045] Figure 7 is provided in the present specification corresponding to Figure 1 schematic diagram of the electronic device. Detailed Embodiments

[0046] To make the objectives, technical solutions, and advantages of the present specification clearer, the technical solutions of the present specification will be clearly and completely described below in conjunction with specific embodiments of the present specification and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present specification, rather than all of the embodiments. Based on the embodiments in the present specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present specification.

[0047] At present, common methods for preparing single-crystal fibers include the guiding mode method, the micro-pulling down method, the laser-heated pedestal growth (LHPG) method, etc. Among them, the LHPG method is particularly good at preparing ultra-fine single-crystal fibers with a diameter of less than 100 microns, while other methods such as the micro-pulling down method (μ-PD) and the edge-defined film-fed growth (EFG) method are usually more used to prepare fibers with a diameter in the range of 0.5 to 2 mm. This makes the LHPG method widely used in national defense and people's livelihood fields such as high-energy lasers, high-temperature sensing, radiation detection, information communication, etc., specifically as Figure 1 shown.

[0048] Figure 1 is a schematic diagram of the principle of the LHPG system provided in this specification.

[0049] Combined with Figure 1 it can be seen that in the process of preparing single-crystal fibers through the LHPG system, a laser beam can be transmitted through a carefully designed optical path and then focused on the feed rod located at the bottom of the growth system. The high energy of the laser melts the material at the top of the feed rod, forming a molten region. At the same time, a motor drive system pulls the fiber upward at a certain speed, while another motor drives the feed rod to feed upward at a corresponding speed. In this way, the material in the molten region gradually solidifies to form a single-crystal fiber as the feed rod is continuously fed and the fiber is continuously pulled up. By precisely controlling the laser power, the feed speed, and the pulling speed, the diameter and quality of the single-crystal fiber can be effectively controlled.

[0050] However, in the above process of preparing single-crystal fibers, due to factors such as environmental vibration, laser power fluctuation, and raw material uniformity (e.g., local regions in the feed rod are contaminated by impurities) that the LHPG system may be affected by, the transmission losses of the grown single-crystal fibers are not the same. Since the transmission loss of the single-crystal fiber can only be measured after the entire fiber is pulled out from the LHPG system, when it is determined that the prepared single-crystal fiber does not meet the standard, it is often necessary to re-prepare the single-crystal fiber. Therefore, in the process of preparing single-crystal fibers, it is particularly important to monitor the quality of the single-crystal fiber being pulled in real time.

[0051] At present, the commonly used on-line monitoring technology for single-crystal fibers is to collect the image of the melting zone during the process of preparing single-crystal fibers through the LHPG system by a preset image acquisition device and measure the diameter of the just-pulled single-crystal fiber. When it is monitored that the diameter of the just-pulled single-crystal fiber becomes larger, the pulling speed of the upper motor can be appropriately increased to reduce the fiber diameter. Through such monitoring and control, the diameter fluctuation of the prepared single-crystal fiber can be kept within a certain range, thereby improving the quality of the prepared single-crystal fiber.

[0052] However, the diameter fluctuation of the single-crystal fiber only affects one aspect of the fiber quality. What ultimately determines the quality of the fiber is the transmission loss of the single-crystal fiber. For example: In actual production, it is very likely that there may be local minute impurity contamination in the raw material rod sometimes, which will result in an abnormal increase in the transmission loss of a small section in the drawn single-crystal fiber. Although the diameter fluctuation of the whole fiber meets the manufacturing requirements, due to the high loss of this small section, the transmission loss of the whole fiber will also increase accordingly, affecting the performance of the fiber.

[0053] Therefore, if the transmission loss of the single-crystal fiber during the drawing process can be monitored in real time, measures can be taken promptly when abnormalities are detected. For example: Stop the drawing, cut off the problematic fiber section and the raw material rod, and then continue the drawing. This can not only ensure the quality of the whole fiber but also effectively reduce the cost increase caused by quality problems.

[0054] The following will, in conjunction with the accompanying drawings, elaborate on the technical solutions provided by each embodiment of this specification in detail.

[0055] Figure 2 It is a schematic flowchart of an on-line monitoring method for the loss of a single-crystal fiber provided in this specification, including the following steps:

[0056] S201: Obtain a sequence of target melt zone images. The sequence of target melt zone images contains each target melt zone image, which is an image of the melt zone of the raw material rod used for preparing the single-crystal fiber collected during the preparation of the single-crystal fiber. The target melt zone image contains bright lines formed due to the penetration of the preset illumination light inside the drawn fiber. Different target melt zone images are collected at different moments within a specified time period.

[0057] In this specification, the service platform can obtain a sequence of target melt zone images, and based on the obtained sequence of target melt zone images through a preset monitoring model, predict the transmission loss values of the single-crystal fibers corresponding to each target melt zone image included in the sequence of target melt zone images. Furthermore, control instructions for preparing the single-crystal fiber can be generated according to the predicted transmission loss values, so as to control the stop of the preparation of the single-crystal fiber according to the control instructions.

[0058] Among them, the above sequence of target melt zone images contains each target melt zone image, and different target melt zone images are collected by a preset image acquisition device at different moments within a specified time period.

[0059] For example, it is possible to set the specified time period for collecting each target melt zone image in the target melt zone image sequence to 30 seconds. During this period, an image can be collected every 3 seconds through a preset image acquisition device, and finally 10 target melt zone images are obtained, constituting the target melt zone image sequence. Such a collection method can ensure the uniform distribution of each target melt zone image in the target melt zone image sequence in terms of time, providing a continuous and time-resolved image sequence for subsequent analysis.

[0060] The above-mentioned image acquisition device can be a hardware device capable of capturing and recording images, such as electronic components like sensors, digital cameras, cameras, industrial cameras, etc.

[0061] The above-mentioned target melt zone image contains bright lines formed due to the penetration of the preset illumination light inside the drawn optical fiber, specifically as Figure 3 shown.

[0062] Figure 3 This is a schematic diagram of the target melt zone image provided in this specification.

[0063] In Figure 3 , the left image is the target melt zone image corresponding to a single-crystal optical fiber with a relatively high transmission loss value, and the right image is the target melt zone image corresponding to a single-crystal optical fiber with a relatively low transmission loss value. Among them, after passing through a pre-set camera and illumination, it can be observed that a bright line is formed in the center of the single-crystal optical fiber due to the penetration of the illumination optical fiber. This bright line is formed by the illumination light transmitted by the optical fiber. Since the illumination light itself is uniform, the unevenness of the light spots on the bright line in the center of the single-crystal optical fiber can reflect the transmission loss of the single-crystal optical fiber. In the regions with higher and lower losses, there will be obvious differences in the brightness and distribution of the light spots, that is, in Figure 3 , the white line in the single-crystal optical fiber region in the left image is in a discontinuous state, while in Figure 3 , the white line in the single-crystal optical fiber region in the left image is in a uniform and continuous state. Therefore, the target melt zone image can be recognized through a monitoring model to estimate the transmission loss of the single-crystal optical fiber corresponding to the target melt zone image.

[0064] In this specification, the execution entity for implementing the online monitoring method of single-crystal optical fiber loss can refer to a specified device set in a service platform such as a server or an Internet of Things (IOT) device, or it can also refer to a terminal device installed with a control application program of the LHPG system such as a desktop computer or a laptop computer. For the sake of convenience of description, below only the case where the terminal device is the execution entity is taken as an example to illustrate the online monitoring method of single-crystal optical fiber loss provided in this specification.

[0065] S202: Input the target molten zone image sequence into a pre-trained monitoring model, so as to determine the target molten zone features according to the target molten zone image sequence through the monitoring model, and estimate the transmission loss value of the single crystal optical fiber drawn within the specified time period according to the target molten zone features. The target molten zone features are used to reflect the uniformity of the bright lines included in the target molten zone image, the diameter change of the drawn optical fiber, and the shape change of the molten zone.

[0066] In this specification, after obtaining the target molten zone image sequence, the terminal device can input the obtained target molten zone image sequence into a pre-trained monitoring model, so as to determine the target molten zone features according to the target molten zone image sequence through the preset monitoring model, and estimate the transmission loss value of the single crystal optical fiber drawn within the specified time period according to the target molten zone features.

[0067] Among them, the above-mentioned target molten zone features are used to reflect the uniformity of the bright lines included in the target molten zone image, the diameter change of the drawn optical fiber, and the shape change of the molten zone.

[0068] Specifically, a convolutional module (Convolutional Neural Network, CNN) can be provided in the above-mentioned monitoring module. At this time, the terminal device can input the target molten zone image sequence into a pre-trained monitoring model, so as to perform convolutional processing on each target molten zone image in the target molten zone image sequence through the convolutional module, extract the image convolutional features corresponding to the target molten zone image from the target molten zone image, and determine the target molten zone features according to the image convolutional features corresponding to each target molten zone image. Furthermore, the transmission loss value of the single crystal optical fiber drawn within the specified time period can be estimated according to the target molten zone features.

[0069] In an actual application scenario, in order to be able to extract the dynamic features of each target molten zone image in the target molten zone image sequence over time, a sequence feature extraction module can also be provided in the above-mentioned monitoring model. At this time, after the terminal device determines the image convolutional features corresponding to each target molten zone image through the convolutional module of the monitoring model, it can also input the image convolutional features corresponding to each target molten zone image into the sequence feature extraction module in sequence according to the order of the acquisition times of each target molten zone image, so as to perform several rounds of processing on the image convolutional features corresponding to each target molten zone image through the sequence feature extraction module to obtain the sequence features corresponding to the target molten zone image sequence as the target molten zone features.

[0070] Among them, for each round of processing, according to the chronological order of the acquisition times of the target melting zone images, one target melting zone image is selected from the target melting zone images as the image to be processed in this round of processing, and the image convolution feature of the image to be processed in this round of processing is input into the sequence feature extraction module as the target image convolution feature in this round of processing, specifically as follows Figure 4 as shown.

[0071] Figure 4 is a schematic diagram of the determination process of the transmission loss value of the single crystal fiber provided in this specification.

[0072] Combined with Figure 4 it can be seen that for each round of processing, the terminal device can determine the image convolution feature of the target melting zone image input into the sequence feature extraction module in this round of processing as the target image convolution feature, and obtain the candidate target melting zone feature obtained after the previous round of processing. Thus, the target image convolution feature and the candidate target melting zone feature can be fused to obtain the initial candidate target melting zone feature in this round of processing, and the initial candidate target melting zone feature is screened through a preset forgetting function to obtain the candidate target melting zone feature after this round of processing until the preset termination condition is met. At this time, the sequence feature corresponding to the target melting zone image sequence is obtained as the target melting zone feature.

[0073] In the above content, there are various methods for the terminal device to fuse the target image convolution feature and the candidate target melting zone feature to obtain the initial candidate target melting zone feature in this round of processing. For example: weighted fusion of the target image convolution feature and the candidate target melting zone feature according to the weights corresponding to the preset target image convolution feature and the candidate target melting zone feature.

[0074] In the above content, the preset forgetting function can be an activation function, and the output value is between 0 and 1. A value close to 0 indicates "forgetting", and a value close to 1 indicates "retaining". The process of screening the initial candidate target melting zone feature through the preset forgetting function can refer to the following formula:

[0075]

[0076]

[0077] In the above formula, is the activation value through the forgetting function, is the weight corresponding to the preset target image convolution feature, is the weight of the preset candidate target melting zone feature, is the target image convolution feature, is the candidate target melting zone feature, is the bias parameter, This is the candidate target melt zone feature after this round of processing.

[0078] Further, after the terminal device obtains the sequence feature corresponding to the target melt zone image sequence as the target melt zone feature, it can input the target melt zone feature into the decoding layer of the monitoring model. Then, through the decoding layer of the monitoring model, according to the input target melt zone feature, the target melt zone feature activation value corresponding to the target melt zone image sequence can be obtained. Furthermore, the transmission loss value of the single crystal optical fiber drawn within the specified time period can be estimated based on the target melt zone feature activation value corresponding to the target melt zone image sequence.

[0079] The above-mentioned monitoring model can be deployed to the terminal device only after being trained. The training method of the monitoring model here can be that the terminal device obtains the initial monitoring model, the sample melt zone image sequence, and the actual transmission loss value of the single crystal optical fiber drawn within the time period corresponding to the sample melt zone image sequence. Then, the pre-obtained sample melt zone image sequence is input into the initial monitoring model. Through the initial monitoring model, according to the sample melt zone image sequence, the sample melt zone feature is determined, and based on the sample melt zone feature, the transmission loss value of the single crystal optical fiber drawn within the time period corresponding to the sample melt zone image sequence is estimated as the estimated transmission loss value. Furthermore, based on the deviation between the estimated transmission loss value and the actual transmission loss value, the target loss is determined, and the initial monitoring model is trained with the goal of minimizing the target loss to obtain the monitoring model.

[0080] Among them, the above-mentioned target loss is positively correlated with the deviation between the estimated transmission loss value and the actual transmission loss value.

[0081] The terminal device obtains the actual transmission loss value of the single crystal optical fiber drawn within the time period corresponding to the sample melt zone image sequence, which can be measured and calculated by a spectrometer connected to the end of the single crystal optical fiber drawn within the time period corresponding to the sample melt zone image sequence. Specifically, as Figure 5 shown.

[0082] Figure 5 This is a schematic diagram of the measurement process of the actual transmission loss value provided in this specification.

[0083] Combined with Figure 5 It can be seen that during the process of drawing a single crystal optical fiber, a section of seed optical fiber can be left at the end of the single crystal optical fiber being drawn, and the end face of the seed optical fiber is ground and prepared into a standard connector and connected to a preset spectrometer. Thus, the spectral data of the single crystal optical fiber being drawn can be measured by the spectrometer. Furthermore, by calculating the change value between the spectral data of the single crystal optical fiber drawn before the time period corresponding to the sample melt zone image sequence and the spectral data of the single crystal optical fiber drawn after the time period corresponding to the sample melt zone image sequence, the actual transmission loss value of the single crystal optical fiber drawn within the time period corresponding to the sample melt zone image sequence can be obtained.

[0084] In an actual application scenario, since the spectral data of the single-crystal optical fiber being drawn measured by a spectrometer often fluctuates up and down (that is, the spectral data is not a continuously decreasing or continuously increasing curve, but a curve that fluctuates up and down), therefore, there may be a certain error in the actual transmission loss value of the single-crystal optical fiber drawn during the corresponding time period of the sample melt zone image sequence measured by the spectrometer.

[0085] Based on this, another method for the above terminal device to obtain the actual transmission loss value of the single-crystal optical fiber drawn during the corresponding time period of the sample melt zone image sequence is to cut off the single-crystal optical fiber drawn during the corresponding time period of the sample melt zone image sequence to obtain the finally drawn single-crystal optical fiber during the corresponding time period of the sample melt zone image sequence as the sample single-crystal optical fiber. Thus, a laser can be input at one end of the sample single-crystal optical fiber, and the laser power of the laser output from the other end of the sample single-crystal optical fiber can be measured. Furthermore, a specified length of the sample single-crystal optical fiber can be cut off at the other end of the sample single-crystal optical fiber, and the laser power of the laser output from the other end of the remaining sample single-crystal optical fiber can be measured again. Then, based on the difference in the laser power obtained from the two measurements, the transmission loss value of the cut-off part of the sample single-crystal optical fiber can be determined, and the above operations can be repeated until the transmission loss value of the entire sample single-crystal optical fiber is determined as the actual transmission loss value of the single-crystal optical fiber drawn during the corresponding time period of the sample melt zone image sequence.

[0086] S203: If it is determined that the transmission loss value meets the preset alarm condition, an alarm message is generated and the alarm message is uploaded to perform a task according to the alarm message.

[0087] In this specification, if the terminal device determines that the transmission loss value of the single-crystal optical fiber drawn during the specified time period corresponding to the predicted target melt zone image sequence output by the monitoring model meets the preset alarm condition, an alarm message can be generated. Furthermore, the alarm message can be uploaded so that the control system for controlling the drawing of the single-crystal optical fiber can generate a control instruction according to the alarm message, and the device for drawing the single-crystal optical fiber can be controlled to stop drawing, cut off the problematic optical fiber part and the raw material rod, and then continue to draw the single-crystal optical fiber.

[0088] Among them, the above alarm condition can be set according to actual needs. For example: if it is determined that the transmission loss value of the single-crystal optical fiber drawn during the specified time period corresponding to the predicted target melt zone image sequence output by the monitoring model is higher than the preset threshold, it can be determined that the preset alarm condition is met.

[0089] For another example, if it is determined that the difference between the transmission loss value of the single-crystal optical fiber drawn during a specified time period corresponding to the predicted target melt zone image sequence output by the monitoring model and the historical transmission loss value output by the monitoring model exceeds a preset threshold, it can be determined that the preset alarm condition is met.

[0090] It should be noted that in an actual application scenario, the terminal device can also obtain target melt zone video data and input the target melt zone video data into a preset monitoring model, so that the monitoring model can predict the transmission loss value of the single-crystal optical fiber corresponding to the target melt zone video data based on each frame of the target melt zone image included in the target melt zone video data, and then an alarm message can be generated according to the predicted transmission loss value.

[0091] As can be seen from the above method, the terminal device can, during the process of preparing the single-crystal optical fiber, collect a target melt zone image sequence of the single-crystal optical fiber drawn during a specified time period through a preset image acquisition device, and then can, based on the obtained target melt zone image sequence, perform online monitoring on the quality of the single-crystal optical fiber being drawn through a preset monitoring model, and can send an alarm message when it is detected that there is an abnormality in the drawn single-crystal optical fiber, so that the device for drawing the single-crystal optical fiber can make production adjustments in a timely manner according to the alarm message, thereby improving the yield of the single-crystal optical fiber and further reducing the cost of preparing the single-crystal optical fiber.

[0092] The above is one or more embodiments of the method for online monitoring of single-crystal optical fiber loss in this specification. Based on the same idea, this specification also provides a corresponding device for online monitoring of single-crystal optical fiber loss, as Figure 6 shown.

[0093] Figure 6 The following is a schematic diagram of a device for online monitoring of single-crystal optical fiber loss provided in this specification, including:

[0094] An acquisition module 601, configured to acquire a target melt zone image sequence, where each target melt zone image in the target melt zone image sequence is an image of the melt zone of the raw material rod for preparing the single-crystal optical fiber collected during the process of preparing the single-crystal optical fiber, and the target melt zone image includes bright lines formed due to the penetration of preset illumination light inside the drawn optical fiber, and different target melt zone images are collected at different times within a specified time period;

[0095] A prediction module 602, configured to input the target molten zone image sequence into a pre-trained monitoring model, so as to determine, according to the target molten zone image sequence by the monitoring model, target molten zone features, and estimate a transmission loss value of a single-crystal optical fiber drawn within the specified time period according to the target molten zone features, where the target molten zone features are used to reflect the uniformity of the bright lines included in the target molten zone image, the diameter change of the drawn optical fiber, and the shape change of the molten zone;

[0096] An execution module 603, configured to generate an alarm message when it is determined that the transmission loss value meets a preset alarm condition, and upload the alarm message, so as to perform a task according to the alarm message.

[0097] Optionally, the monitoring model includes: a convolution module;

[0098] The prediction module 602 is specifically configured to input the target molten zone image sequence into a pre-trained monitoring model, so as to, through the convolution module, perform convolution processing on each target molten zone image in the target molten zone image sequence to extract image convolution features corresponding to the target molten zone image from the target molten zone image; and determine target molten zone features according to the image convolution features corresponding to each target molten zone image.

[0099] Optionally, the monitoring model includes: a sequence feature extraction module;

[0100] The prediction module 602 is specifically configured to, according to the chronological order of the acquisition times of the target molten zone images, sequentially input the image convolution features corresponding to the target molten zone images into the sequence feature extraction module, so as to perform several rounds of processing by the sequence feature extraction module according to the image convolution features corresponding to the target molten zone images to obtain sequence features corresponding to the target molten zone image sequence as target molten zone features; where, for each round of processing, determine the image convolution features of the target molten zone image input into the sequence feature extraction module in this round of processing as target image convolution features, and obtain the candidate target molten zone features obtained after the previous round of processing, fuse the target image convolution features with the candidate target molten zone features to obtain the initial candidate target molten zone features in this round of processing, and screen the initial candidate target molten zone features through a preset forgetting function to obtain the candidate target molten zone features after this round of processing until it is determined that a preset termination condition is met, at this time, obtain the sequence features corresponding to the target molten zone image sequence as target molten zone features.

[0101] Optionally, the apparatus further includes: a training module 604;

[0102] The training module 604 is specifically configured to obtain an initial monitoring model, a sample melt zone image sequence, and the actual transmission loss value of the single crystal optical fiber drawn during the time period corresponding to the sample melt zone image sequence; input the sample melt zone image sequence into the initial monitoring model, so that the initial monitoring model determines sample melt zone features according to the sample melt zone image sequence, and estimates the transmission loss value of the single crystal optical fiber drawn during the time period corresponding to the sample melt zone image sequence as the estimated transmission loss value; determine the target loss according to the deviation between the estimated transmission loss value and the actual transmission loss value, and train the initial monitoring model with the goal of minimizing the target loss to obtain a monitoring model, where the target loss is positively correlated with the deviation between the estimated transmission loss value and the actual transmission loss value.

[0103] Optionally, the actual transmission loss value of the single crystal optical fiber drawn during the time period corresponding to the sample melt zone image sequence is measured and calculated by a spectrometer connected to the end of the single crystal optical fiber drawn during the time period corresponding to the sample melt zone image sequence.

[0104] This specification also provides a computer-readable storage medium storing a computer program that can be used to execute the above Figure 1 provided on-line monitoring method for single crystal optical fiber loss.

[0105] This specification also provides Figure 7 the schematic structural diagram of an electronic device corresponding to Figure 1 as shown. As Figure 7 described, at the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory. Of course, other hardware required for other services may also be included. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to implement the above Figure 1 on-line monitoring method for single crystal optical fiber loss. Of course, in addition to the software implementation method, this specification does not exclude other implementation methods, such as logic devices or a combination of software and hardware. That is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or a logic device.

[0106] Improvements to a technology can be clearly distinguished as hardware improvements (e.g., improvements to circuit structures such as diodes, transistors, switches, etc.) or software improvements (improvements to method flows). However, with the development of technology, many improvements to method flows today can be regarded as direct improvements to hardware circuit structures. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that an improvement to a method flow cannot be implemented with a hardware entity module. For example, a programmable logic device (PLD) (e.g., a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user programming the device. Designers can program themselves to "integrate" a digital system on a single PLD without having to ask a chip manufacturer to design and fabricate a dedicated integrated circuit chip. Moreover, nowadays, instead of manually fabricating integrated circuit chips, this programming is mostly implemented using "logic compiler" software, which is similar to the software compiler used in program development and writing. The original code before compilation also has to be written in a specific programming language, which is called a hardware description language (HDL), and there is not just one type of HDL, but many types, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones currently are VHDL (Very-High-Speed Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also be aware that by simply performing a little logical programming on the method flow using the above-mentioned several hardware description languages and programming it into the integrated circuit, it is easy to obtain the hardware circuit that implements the logical method flow.

[0107] The controller can be implemented in any suitable manner. For example, the controller can take the form of, for example, a microprocessor or a processor and a computer-readable medium storing computer-readable program code (such as software or firmware) executable by the (micro)processor, logic gates, switches, an application specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of the controller include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art also know that in addition to implementing the controller in the form of pure computer-readable program code, it is entirely possible to logically program the method steps to enable the controller to be implemented in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. to achieve the same function. Therefore, such a controller can be considered a hardware component, and the devices included therein for implementing various functions can also be regarded as the structures within the hardware component. Or even, the devices for implementing various functions can be regarded as either software modules for implementing the method or structures within the hardware component.

[0108] The systems, devices, modules, or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0109] For the convenience of description, when describing the above devices, they are described separately as various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0110] Those skilled in the art should understand that the embodiments of this specification can be provided as a method, a system, or a computer program product. Therefore, this specification can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, this specification can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code.

[0111] This specification is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the specification. It should be understood that each flow and / or block in the flowchart and / or block diagram, and combinations of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to produce a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices produce means for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0112] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0113] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0114] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0115] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0116] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0117] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0118] It should be understood by those skilled in the art that the embodiments of this specification may be provided as methods, systems or computer program products. Therefore, this specification may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, this specification may take the form of a computer program product implemented 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.

[0119] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0120] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other, and the differences between each embodiment and other embodiments are emphasized. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and for the relevant parts, reference can be made to the description of the method embodiment.

[0121] The above description is only for the embodiments of this specification and is not intended to limit this specification. For those skilled in the art, various changes and modifications can be made to this specification. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification shall be included within the scope of the claims of this specification.

Claims

1. A method for online monitoring of single crystal optical fiber loss, characterized in that: include: Acquire a target melting zone image sequence, wherein each target melting zone image is an image of a melting zone of a raw material rod for preparing a single crystal optical fiber collected during the process of preparing the single crystal optical fiber, wherein the target melting zone image contains bright lines formed inside the drawn optical fiber due to the penetration of a preset illumination light, and different target melting zone images are collected at different times within a specified time period; Inputting the target melt zone image sequence into a pre-trained monitoring model, so as to determine the target melt zone characteristics according to the target melt zone image sequence through the monitoring model, and estimate the transmission loss value of the single crystal optical fiber drawn within the specified time period according to the target melt zone characteristics, wherein the target melt zone characteristics are used to reflect the uniformity of the bright lines contained in the target melt zone image, the diameter change of the drawn optical fiber, and the shape change of the melt zone; If it is determined that the transmission loss value meets the preset alarm condition, an alarm message is generated and the alarm message is uploaded so as to perform a task according to the alarm message.

2. The method according to claim 1, characterized in that The monitoring model includes: a convolution module; Inputting the target melt zone image sequence into a pre-trained monitoring model, so as to determine the target melt zone characteristics according to the target melt zone image sequence through the monitoring model, specifically including: Inputting the target melt zone image sequence into a pre-trained monitoring model, so as to perform convolution processing on each target melt zone image in the target melt zone image sequence through the convolution module, so as to extract image convolution features corresponding to the target melt zone image from the target melt zone image; The target melting zone features are determined according to the image convolution features corresponding to each target melting zone image.

3. The method according to claim 2, characterized in that The monitoring model includes: a sequence feature extraction module; According to the image convolution features corresponding to each target melt zone image, the target melt zone features are determined, specifically including: According to the sequence of acquisition time of each target melt zone image, the image convolution features corresponding to each target melt zone image are sequentially input into the sequence feature extraction module, so that the sequence feature extraction module performs several rounds of processing according to the image convolution features corresponding to each target melt zone image, so as to obtain the sequence features corresponding to the target melt zone image sequence as the target melt zone features; wherein, For each round of processing, the image convolution features of the target melt zone image input to the sequence feature extraction module in the round of processing are determined as the target image convolution features, and the candidate target melt zone features obtained after the previous round of processing are obtained, and the target image convolution features are fused with the candidate target melt zone features to obtain the initial candidate target melt zone features in the current round of processing, and the initial candidate target melt zone features are screened by a preset forgetting function to obtain the candidate target melt zone features after the current round of processing, until it is determined that the preset termination condition is met, at which time, the sequence features corresponding to the target melt zone image sequence are obtained as the target melt zone features.

4. The method according to claim 1, characterized in that Training monitoring models, including: Acquire an initial monitoring model, a sample melt zone image sequence, and an actual transmission loss value of a single crystal optical fiber drawn within a time period corresponding to the sample melt zone image sequence; Inputting the sample melt zone image sequence into the initial monitoring model, so as to determine the sample melt zone characteristics according to the sample melt zone image sequence through the initial monitoring model, and estimating the transmission loss value of the single crystal optical fiber drawn in the time period corresponding to the sample melt zone image sequence according to the sample melt zone characteristics as the estimated transmission loss value; According to the deviation between the estimated transmission loss value and the actual transmission loss value, a target loss is determined, and the initial monitoring model is trained with the goal of minimizing the target loss to obtain a monitoring model, wherein the target loss is positively correlated with the deviation between the estimated transmission loss value and the actual transmission loss value.

5. The method according to claim 4, characterized in that The actual transmission loss value of the single crystal optical fiber drawn within the time period corresponding to the sample melt zone image sequence is measured and calculated by a spectrometer connected to the end of the single crystal optical fiber drawn within the time period corresponding to the sample melt zone image sequence.

6. A single crystal optical fiber loss online monitoring device, characterized in that: include: An acquisition module, used to acquire a target melting zone image sequence, wherein each target melting zone image is an image of a melting zone of a raw material rod for preparing a single crystal optical fiber acquired during the process of preparing the single crystal optical fiber, wherein the target melting zone image includes bright lines formed inside the drawn optical fiber due to the penetration of a preset illumination light, and different target melting zone images are acquired at different times within a specified time period; A prediction module, used for inputting the target melt zone image sequence into a pre-trained monitoring model, so as to determine the target melt zone characteristics according to the target melt zone image sequence through the monitoring model, and estimate the transmission loss value of the single crystal optical fiber drawn within the specified time period according to the target melt zone characteristics, wherein the target melt zone characteristics are used to reflect the uniformity of the bright lines contained in the target melt zone image, the diameter change of the drawn optical fiber, and the shape change of the melt zone; The execution module is used to generate alarm information when it is determined that the transmission loss value meets the preset alarm condition, and upload the alarm information to perform tasks according to the alarm information.

7. The device according to claim 6, characterized in that The monitoring model includes: a convolution module; The prediction module is specifically used to input the target melt zone image sequence into a pre-trained monitoring model, so as to perform convolution processing on each target melt zone image in the target melt zone image sequence through the convolution module, so as to extract the image convolution feature corresponding to the target melt zone image from the target melt zone image; and determine the target melt zone feature according to the image convolution feature corresponding to each target melt zone image.

8. The device according to claim 7, characterized in that The monitoring model includes: a sequence feature extraction module; The prediction module is specifically used to input the image convolution features corresponding to each target melt zone image into the sequence feature extraction module in sequence according to the sequence of acquisition time of each target melt zone image, so that the sequence feature extraction module performs several rounds of processing according to the image convolution features corresponding to each target melt zone image, so as to obtain the sequence features corresponding to the target melt zone image sequence as the target melt zone features; wherein, for each round of processing, the image convolution features of the target melt zone image input into the sequence feature extraction module in the round of processing are determined as the target image convolution features, and the candidate target melt zone features obtained after the previous round of processing are obtained, the target image convolution features are fused with the candidate target melt zone features to obtain the initial candidate target melt zone features in the current round of processing, and the initial candidate target melt zone features are screened by a preset forgetting function to obtain the candidate target melt zone features after the current round of processing, until it is determined that the preset termination condition is met, at which time, the sequence features corresponding to the target melt zone image sequence are obtained as the target melt zone features.

9. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any one of claims 1 to 5 is implemented.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method described in any one of claims 1 to 5 is implemented.

Images