Optical cable routing general survey method and system based on optical cable sonar instrument
The optical cable sonar collects and fuses acoustic signals, and uses deep learning models to identify the location of optical cable faults, solving the information loss and noise problems in the optical cable census, realizing the precise positioning and repair of optical cable faults.
Patent Information
- Application Number
- CN202411656394.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-19
- Publication Date
- 2025-01-07
AI Technical Summary
The existing optical cable census technology has problems such as information loss, environmental interference and signal noise, which makes it difficult and inaccurate to repair optical cable faults.
Using a cable-based sonar method, by applying knocking and sound to the optical cable, two sound signals are collected and fused, and deep learning models such as convolutional length and short-term memory network (ConvLSTM) are used to identify the location of the optical cable failure.
It realizes accurate identification of optical cable fault locations, improves the accuracy and efficiency of optical cable census, and reduces the impact of environmental interference and noise.
Smart Images
Figure CN119270280A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of optical cable survey, and in particular relates to an optical cable route survey method and system based on an optical cable sonar. Background Art
[0002] Communication optical cable is composed of a core and outer sheath made up of several optical fibers (generally from a few cores to thousands of cores). Optical fiber has the advantages of large transmission capacity, low attenuation, long transmission distance, small size, light weight, no electromagnetic interference, and low cost. It is currently the most promising communication transmission medium. At present, optical cables are widely used in signal transmission in various departments such as telecommunications, electricity, and broadcasting.
[0003] In specific applications, optical cables can be deployed over long distances by means of pipe burial, underground burial, or aerial installation. Due to historical reasons, there is a problem of information loss in the optical cables currently laid and in use, and there is also a problem that engineering personnel fail to update pipeline resource information in a timely manner, resulting in a mismatch between existing records and actual conditions, affecting subsequent maintenance and management decisions.
[0004] When an optical cable fails during use, the loss of information makes it difficult to repair the optical cable line failure, and the repair takes a long time and is costly. One solution in the prior art is to conduct a census of optical cable resources, retrieve the lost data, and form optical cable dumb resources for unified management.
[0005] Fiber optic distributed acoustic sensing technology is based on the principle of back Rayleigh scattering. It collects back Rayleigh scattered light. By analyzing and processing the back Rayleigh scattered light, it can effectively conduct a line inspection of optical cables. However, optical cables are usually deployed in various background environments, and there are various interferences, which makes it difficult to deeply understand the collected signals, resulting in errors in the line inspection of optical cables. Pattern recognition of the monitored signals reduces the difficulty of optical cable line inspection. However, the above method still has certain problems, mainly including: Environmental interference: Optical cables are usually deployed in complex environments and may be affected by temperature changes, mechanical vibrations, electromagnetic interference and other factors. These interferences will be superimposed on the monitored optical signals, making signal processing complicated. Difficult signal interpretation: Even if the signal can be collected using fiber optic distributed acoustic sensing technology, due to the presence of background noise, the signal is often not pure enough, difficult to accurately analyze and understand, and may lead to misjudgment. Pattern recognition challenge: Although existing technologies rely on pattern recognition technology and can help reduce the difficulty of line inspections, in actual applications, it is necessary to process a large amount of data and be able to accurately distinguish different types of sound signals, which places high demands on the robustness and accuracy of the algorithm. Summary of the invention
[0006] Technical problem to be solved by the present invention: In view of the above-mentioned problems in the prior art, a method and system for optical cable route survey based on optical cable sonar are provided. The present invention aims to solve the problem of inaccurate optical cable route survey information in the prior art and realize accurate optical cable route survey.
[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is: A method for surveying optical cable routes based on an optical cable sonar instrument comprises the following steps: S1, using an optical cable sonar to obtain two types of acoustic wave signals caused by knocking and sound after knocking and sound are applied to the optical cable under inspection; S2, fusing the two sound wave signals to obtain a fused sound wave signal; S3, using a deep learning model to detect the fused acoustic wave signal to obtain the fault location information of the optical cable, wherein the deep learning model is pre-trained to establish a mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable.
[0008] Optionally, applying knocking and sound to the checked optical cable in step S1 includes: performing distance segmentation according to a set length frame in the possible direction of the optical cable route, setting knocking points and sound wave points in each distance segment, and knocking at the knocking points with a handheld vibration hammer at certain time intervals, and emitting sounds at the sound wave points after the knocking is completed, and the energy of the knocking or sound applied is carried out in a gradually increasing or decreasing manner, so that a series of two sound wave signals of different sizes caused by knocking and sound are collected for the same distance segment within a specified time.
[0009] Optionally, before fusing the two sound wave signals to obtain the fused sound wave signal in step S2, it also includes using a filter to filter and reduce noise on the two sound wave signals caused by the knocking and the sound.
[0010] Optionally, the function expression of fusing the two sound wave signals in step S2 to obtain a fused sound wave signal is: , , , In the above formula, To knock on the fusion signal, is the tapping fusion coefficient, is the acoustic signal caused by the ith knock in the tth distance segment, is the acoustic wave fusion signal, is the sound fusion coefficient, is the acoustic wave signal caused by the i-th sound in the t-th distance segment, is the fused sound wave signal, is the importance coefficient of the acoustic wave signal caused by knocking in the fusion process, is the importance coefficient of the acoustic wave signal caused by the sound in the fusion process, It is the total number of sound wave signals corresponding to the knocks or the total number of sound wave signals corresponding to the sounds generated in the same distance segment.
[0011] Optionally, the importance coefficient of the acoustic signal caused by the knocking in the fusion process and the importance coefficient of the acoustic wave signal caused by the sound in the fusion process The two have the same value.
[0012] Optionally, the deep learning model in step S3 is a convolutional long short-term memory network ConvLSTM.
[0013] Optionally, before step S3, the method further includes training a deep learning model to obtain a mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable: S101, collecting sound signals of two types of sound wave signals caused by knocking and sound with labels, wherein the labels are real fault location information of the optical cable; filtering and denoising the collected sound signals with a filter, normalizing the signals, and dividing the signals into a training set, a verification set, and a test set according to a specified ratio; S102, using the training set, the validation set and the test set to train the deep learning model, and using the specified loss function to optimize and update the model parameters of the deep learning model, and finally saving the model parameters of the deep learning model to obtain a deep learning model that is trained to establish a mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable.
[0014] In addition, the present invention also provides an optical cable route survey system based on an optical cable sonar, comprising a microprocessor and a memory connected to each other, wherein the microprocessor is programmed or configured to execute the optical cable route survey method based on the optical cable sonar.
[0015] In addition, the present invention also provides a computer-readable storage medium, in which a computer program or instruction is stored. The computer program or instruction is programmed or configured to execute the optical cable route survey method based on the optical cable sonar through a processor.
[0016] In addition, the present invention also provides a computer program product, including a computer program or an instruction, wherein the computer program or the instruction is programmed or configured to execute the optical cable route survey method based on an optical cable sonar through a processor.
[0017] Compared with the prior art, the present invention mainly has the following advantages: the optical cable route survey method based on the optical cable sonar of the present invention includes using the optical cable sonar to conduct a resource survey of the optical cable, specifically, using the optical cable sonar to access the optical cable that needs to be checked, and then vibrating and knocking and affecting the sound waves along the possible directions of the route of the optical cable, and then collecting the sound wave signals in the optical cable for analysis, and fusing the sound wave signals corresponding to the vibration and knocking and the sound wave signals corresponding to the sound wave influence during the analysis process, and analyzing the optical cable route position information after obtaining the fused signal, which can solve the problem of inaccurate optical cable survey information in the prior art and realize accurate optical cable route survey. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 Schematic diagram of the basic flow of the method of the embodiment of the present invention.
[0019] Figure 2 Detailed flowchart of the method of the embodiment of the present invention. DETAILED DESCRIPTION
[0020] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0021] like Figure 1 As shown, the optical cable route survey method based on the optical cable sonar in this embodiment includes the following steps: S1, using an optical cable sonar to obtain two types of acoustic wave signals caused by knocking and sound after knocking and sound are applied to the optical cable under inspection; S2, fusing the two sound wave signals to obtain a fused sound wave signal; S3, using a deep learning model to detect the fused acoustic wave signal to obtain the fault location information of the optical cable, wherein the deep learning model is pre-trained to establish a mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable.
[0022] In step S1 of this embodiment, applying knocking and sound to the checked optical cable includes: performing distance segmentation according to a set length frame in the possible direction of the optical cable route, setting knocking points and sound wave points in each distance segment, using a handheld vibration hammer to knock at the knocking point at a certain time interval, and making sounds at the sound wave points after the knocking is completed, and the energy of the knocking or sound is gradually increased or gradually decreased (for example, the knocking is performed in a manner that the knocking force gradually increases or gradually decreases, and the sound power is sounded in a manner that gradually increases or gradually decreases), so that a series of two sound wave signals of different sizes caused by knocking and sound are collected for the same distance segment within a specified time. The knocking force and sound power change regularly. Here, the two fusion systems are also preset with reference to this regular change to improve the fit, thereby providing the accuracy of the analysis. The optical cable sonar will receive the sound wave signals caused by knocking and sound through the optical fiber, and convert these signals into sound wave signals in the form of electrical signals.
[0023] As an optional implementation, before fusing the two sound wave signals to obtain the fused sound wave signal in step S2, it also includes using a filter to filter and reduce noise on the two sound wave signals caused by the knocking and the sound.
[0024] Signal fusion is to fuse the sound wave signal corresponding to the knocking and the sound wave signal corresponding to the sound. In step S2 of this embodiment, the function expression of fusing the two sound wave signals to obtain the fused sound wave signal is: , , , In the above formula, To knock on the fusion signal, is the tapping fusion coefficient, is the acoustic signal caused by the ith knock in the tth distance segment, is the acoustic wave fusion signal, is the sound fusion coefficient, is the acoustic wave signal caused by the i-th sound in the t-th distance segment, is the fused sound wave signal, is the importance coefficient of the acoustic wave signal caused by knocking in the fusion process, is the importance coefficient of the acoustic wave signal caused by the sound in the fusion process, is the total number of sound wave signals corresponding to the knocks or the total number of sound wave signals corresponding to the sounds generated in the same distance segment. As an optional implementation, the importance coefficient of the sound wave signal caused by the knocks in the fusion process in this embodiment is and the importance coefficient of the acoustic wave signal caused by the sound in the fusion process The two have the same value.
[0025] Based on the fused signal, the location of the optical cable is identified through a deep learning model to determine the routing direction of the optical cable.
[0026] As an optional implementation, the deep learning model in step S3 of this embodiment is a convolutional long short-term memory network ConvLSTM. The convolutional long short-term memory network ConvLSTM combines the advantages of convolutional neural network (CNN) in spatial feature extraction and the advantages of long short-term memory network (LSTM) in time series modeling. The convolutional long short-term memory network ConvLSTM includes an input layer: the shape of the input data is (batch_size, time_steps, height, width, channels), where height and width are the spatial dimensions of the signal, and channels is the number of channels of the signal. ConvLSTM layer: multiple ConvLSTM layers are used to extract spatiotemporal features. Regression layer: used for the final regression task. Among them, the ConvLSTM layer includes: a convolutional LSTM layer, which uses multiple ConvLSTM units to process spatiotemporal data; an activation function, which usually uses ReLU or tanh activation function; Dropout, a Dropout layer is added between ConvLSTM layers to prevent overfitting. The regression layer includes: Flatten layer: flattens the output of the ConvLSTM layer into a one-dimensional vector. Fully connected layer: adds one or more fully connected layers for the final regression task. Output layer: uses a linear activation function.
[0027] In this embodiment, before step S3, the process also includes training a deep learning model to determine the mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable: S101, collecting sound signals of two types of sound wave signals caused by knocking and sound with labels, wherein the labels are real fault location information of the optical cable; filtering and denoising the collected sound signals with a filter, normalizing the signals, and dividing the signals into a training set, a verification set, and a test set according to a specified ratio; S102, using the training set, the validation set and the test set to train the deep learning model, and using the specified loss function to optimize and update the model parameters of the deep learning model, and finally saving the model parameters of the deep learning model to obtain a deep learning model that is trained to establish a mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable.
[0028] In step S101 of this embodiment, the collected sound signals are filtered, denoised and normalized using filters, and divided into training sets, validation sets and test sets according to specified ratios, including: Normalization: normalizing the sample signals so that they are in the same numerical range (for example, between 0 and 1). Then, the sample signals corresponding to the knocks and sounds at the same position are fused using the above-mentioned signal fusion formula to obtain the fused sample signals; Filtering: using appropriate filters to remove noise. Segmentation: dividing the fused sample signals into time windows of fixed lengths, each time window as a sample. Label processing: ensuring that each sample signal has a corresponding position label. Data division: dividing the data set into training sets, validation sets and test sets. The usual ratios are 70% (training), 15% (validation), and 15% (testing).
[0029] When the specified loss function is used to optimize and update the model parameters of the deep learning model in step S102 of this embodiment, the loss function may use mean square error (MSE), root mean square error (RMSE), etc. After obtaining a deep learning model that has been trained to establish a mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable, it can be used for real-time signal processing: the trained deep learning model is applied to the processing of the above-mentioned fused signal, and the acoustic wave signal generated by the new knocking and sound is pattern recognized to determine the location information of the optical cable. The predicted route information of the optical cable is constructed based on the optical cable location information obtained in the possible direction of the above-mentioned optical cable route, and the census of the optical cable route is completed.
[0030] like Figure 2As shown, the overall process of the method of this embodiment includes: S110, collecting sample data to train the convolutional long short-term memory network (ConvLSTM) to obtain a trained convolutional long short-term memory network (ConvLSTM) model. S120, connecting the optical cable sonar to the optical cable to be tested; S130, obtaining the possible routing direction of the optical cable, dividing the optical cable into distance segments according to the possible routing direction of the optical cable, and setting a knocking point and a sound wave point in each distance segment; S140, using a handheld vibration hammer to knock at the knocking point at a certain time interval, and the knocking force increases or decreases according to a certain rule; S150, using a sound generator to sound at the sound wave point at a certain time interval, and the power of the sound generator increases or decreases according to a certain rule when sounding; S160, while knocking and sounding, using the optical cable sonar to collect the corresponding sound waves, and convert these signals into electrical signals. S170. Use the above-mentioned signal fusion formula to fuse the converted electrical signal with the acoustic signal to obtain a fused signal; S180. Input the fused signal into a trained convolutional long short-term memory network (ConvLSTM) model to obtain the corresponding optical cable location information; S180. Construct the predicted routing information of the optical cable based on the optical cable location information obtained in the possible directions of the above-mentioned optical cable routing, and complete the census of the optical cable routing.
[0031] The generation method of the two sound waves and the subsequent fusion method in the method of this embodiment need to be combined, which is different from the single sound wave and the randomness of the sound wave emission in the prior art: first, two groups of sound wave signals are formed for each distance segment in the possible direction of the optical cable route, which can more accurately locate each distance segment of the optical cable; second, knocking first and then emitting sound waves, or emitting sound waves first and then knocking, can effectively improve the optical cable sonar to obtain two sound wave signals with regular changes; third, the regular changes of the two sound wave signals can help in the subsequent signal fusion process, and set different numerical values for each sound wave signal. Fusion system, specifically, the knocking force can be gradually increased, then the knocking fusion coefficient is also pre-examined in a gradually increasing manner. Usually, the sound wave signal with a large knocking force is easier to obtain, so when setting the value of the knocking fusion coefficient, it is set to be relatively large, so that the accuracy of the acquired signal can be reflected from different values, and the accuracy of the fused sound wave signal can be improved. The above explains the comparison effect of general knocking and the sound wave signal generation method of this application, the general fusion method, and the fusion method adopted by the method of this embodiment. In addition, the method of this embodiment fully considers the spatial and temporal advantages of ConvLSTM, and is used in the analysis of the two sound wave signals generated by the method of this embodiment. It can not only extract the corresponding spatial features well, but also consider the temporal information of the sound wave signals well, and can play a very accurate analysis and modeling role for the fused sound wave signals.
[0032] In summary, the method of this embodiment processes the possible routing directions of the optical cable in segments, obtains the sound wave signal corresponding to the knocking and the sound wave signal corresponding to the sound in each segment according to the preset knocking and sound emission method, describes the fusion of multiple signals on the same segment, and uses the fused signal for position identification, which can greatly improve the accuracy of position identification. At the same time, the method of this embodiment constructs a model formula for signal fusion, taking into account the importance of the fusion of the sound wave signal corresponding to the knocking and the sound wave signal corresponding to the sound, thereby improving the position identification based on the fused signal. Finally, the recognition mode (deep learning model) adopted by the method of this embodiment is the convolutional long short-term memory network (ConvLSTM), which performs pattern recognition on the fused signal to determine the location information of the optical cable. ConvLSTM combines the advantages of convolutional neural network (CNN) in spatial feature extraction and the advantages of long short-term memory network (LSTM) in time series modeling, which can well consider the spatial information of the optical cable layout and the time information of the signal, thereby greatly improving the accuracy of the model.
[0033] In addition, this embodiment also provides an optical cable route survey system based on an optical cable sonar, comprising a microprocessor and a memory connected to each other, wherein the microprocessor is programmed or configured to execute the optical cable route survey method based on the optical cable sonar.
[0034] In addition, this embodiment also provides a computer-readable storage medium, in which a computer program or instruction is stored. The computer program or instruction is programmed or configured to execute the optical cable route survey method based on the optical cable sonar through a processor.
[0035] In addition, this embodiment also provides a computer program product, including a computer program or instructions, which are programmed or configured to execute the optical cable route survey method based on the optical cable sonar through a processor.
[0036] Those skilled in the art should understand that the technical solutions provided by the embodiments of the present application may be in the form of methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the process Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the functions specified in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0037] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.
Claims
1. A method for surveying optical cable routes based on an optical cable sonar, characterized in that: The steps include: S1, using an optical cable sonar to obtain two types of acoustic wave signals caused by knocking and sound after knocking and sound are applied to the optical cable under inspection; S2, fusing the two sound wave signals to obtain a fused sound wave signal; S3, using a deep learning model to detect the fused acoustic wave signal to obtain the fault location information of the optical cable, wherein the deep learning model is pre-trained to establish a mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable.
2. The optical cable route survey method based on optical cable sonar according to claim 1 is characterized in that: The step S1 of applying knocking and sound to the checked optical cable includes: performing distance segmentation according to a set length frame in the possible direction of the optical cable route, setting knocking points and sound wave points in each distance segment, and knocking at the knocking points with a handheld vibration hammer at certain time intervals, and making sounds at the sound wave points after the knocking is completed, and the energy of the knocking or sound is gradually increased or decreased, so that a series of two sound wave signals of different sizes caused by knocking and sound are collected for the same distance segment within a specified time.
3. The optical cable route survey method based on optical cable sonar according to claim 1 is characterized in that: Before fusing the two sound wave signals to obtain the fused sound wave signal in step S2, the process also includes using a filter to filter and reduce noise on the two sound wave signals caused by the knocking and the sound.
4. The optical cable route survey method based on optical cable sonar according to claim 1 is characterized in that: The function expression of fusing the two sound wave signals in step S2 to obtain the fused sound wave signal is: ， ， ， In the above formula, To knock on the fusion signal, is the tapping fusion coefficient, is the acoustic signal caused by the ith knock in the tth distance segment, is the acoustic wave fusion signal, is the sound fusion coefficient, is the acoustic wave signal caused by the i-th sound in the t-th distance segment, is the fused sound wave signal, is the importance coefficient of the acoustic wave signal caused by knocking in the fusion process, is the importance coefficient of the acoustic wave signal caused by the sound in the fusion process, It is the total number of sound wave signals corresponding to the knocks or the total number of sound wave signals corresponding to the sounds generated in the same distance segment.
5. The optical cable route survey method based on optical cable sonar according to claim 4 is characterized in that: The importance coefficient of the acoustic signal caused by knocking in the fusion process and the importance coefficient of the acoustic wave signal caused by the sound in the fusion process The two have the same value.
6. The optical cable route survey method based on optical cable sonar according to claim 1 is characterized in that: The deep learning model in step S3 is a convolutional long short-term memory network ConvLSTM.
7. The optical cable route survey method based on optical cable sonar according to claim 1 is characterized in that: Before step S3, the process also includes training a deep learning model to determine the mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable: S101, collecting sound signals of two types of sound wave signals caused by knocking and sound with labels, wherein the labels are real fault location information of the optical cable; filtering and denoising the collected sound signals with a filter, normalizing the signals, and dividing the signals into a training set, a verification set, and a test set according to a specified ratio; S102, using the training set, the validation set and the test set to train the deep learning model, and using the specified loss function to optimize and update the model parameters of the deep learning model, and finally saving the model parameters of the deep learning model to obtain a deep learning model that is trained to establish a mapping relationship between the fused acoustic wave signal and the fault location information of the optical cable.
8. An optical cable route survey system based on an optical cable sonar, comprising a microprocessor and a memory connected to each other, characterized in that: The microprocessor is programmed or configured to execute the optical cable route survey method based on optical cable sonar as described in any one of claims 1 to 7.
9. A computer-readable storage medium having a computer program or instruction stored therein, characterized in that: The computer program or instruction is programmed or configured to execute the optical cable route survey method based on optical cable sonar as claimed in any one of claims 1 to 7 through a processor.
10. A computer program product comprising a computer program or instructions, characterized in that The computer program or instruction is programmed or configured to execute the optical cable route survey method based on optical cable sonar as claimed in any one of claims 1 to 7 through a processor.