Method, device and equipment for identifying partial same-cable optical fibers, medium and product

Abstract

The invention discloses a method, a device, equipment, a medium and a product for identifying partial optical fibers in the same cable. The method comprises the following steps: acquiring a first OTDR curve and a second OTDR curve of two optical fibers to be identified; extracting event characteristics of each event in the first OTDR curve and the second OTDR curve; identifying the same event in the two optical fibers to be identified according to the extracted event features; and performing similarity analysis on the OTDR curve of the cable section where the same event is located in the two optical fibers to be identified according to a preset similarity comparison model, and outputting a same-cable optical fiber identification result. According to the scheme, similarity analysis is carried out on the cable sections where the same events are located based on identification of the same events in the OTDR curves of the two to-be-identified optical fibers, blindness of identification of the same-cable optical fibers is reduced, identification of part of the same-cable optical fibers in a complex scene can be dealt with, and different parts of the same-cable optical fibers can be accurately and efficiently identified.

Description

Technical Field

[0001] The present invention relates to the field of optical transmission and bearing technology, and in particular to a method, device, equipment, medium and product for identifying partially co-cable optical fibers. Background Art

[0002] During the network construction and development, operators laid a large number of optical fibers. Over time, it became impossible to effectively verify whether the optical fibers belonged to the same optical cable. For a communication system with a primary and backup configuration, if the optical fiber used by the primary system and the optical fiber used by the backup system are in the same optical cable, the availability will be greatly reduced. Therefore, it is necessary to effectively identify whether the optical fibers are in the same cable, so that the service can be carried on two optical fibers with separate physical routes to improve the reliability of the system.

[0003] The existing methods for detecting and identifying optical fiber co-cables only consider the situation of complete co-cables, that is, two optical fibers are completely fused into the same optical cable. However, in the current network scenario, there are a large number of partial co-cable scenarios, that is, a certain optical fiber will be partially fused into other optical cables due to network planning and other reasons, and different optical fibers will form partial co-cables. Two optical fibers may have several different parts of co-cables. The partial co-cable situations of optical fibers in different scenarios are very complex. The existing optical fiber co-cable identification methods cannot cope with the identification of partial co-cable optical fibers in complex scenarios, and the identification efficiency and identification accuracy cannot meet the requirements. Summary of the invention

[0004] In order to solve the problem that it is difficult to identify partially co-cabled optical fibers, the present invention proposes a method, device, equipment, medium and product for identifying partially co-cabled optical fibers, which can accurately and efficiently realize the identification of partially co-cabled optical fibers in complex scenarios.

[0005] An embodiment of the present invention provides a method for identifying partially co-cable optical fibers, the method comprising:

[0006] Collecting a first OTDR curve and a second OTDR curve of two optical fibers to be identified;

[0007] Extracting event features of each event in the first OTDR curve and the second OTDR curve;

[0008] Identify the same events in the two optical fibers to be identified according to the extracted event features;

[0009] The OTDR curves of the cable segments where the same events are located in the two optical fibers to be identified are analyzed for similarity according to a preset similarity comparison model, and the identification results of the same-cable optical fibers are output.

[0010] Preferably, similarity analysis is performed on the OTDR curves of the cable segments where the same event is located in the two optical fibers to be identified according to a preset similarity comparison model, and the same-cable optical fiber identification result is output, including:

[0011] For each identical event, extracting a preceding OTDR curve between the identical event and a preceding event, and a succeeding OTDR curve between the identical event and a succeeding event from the first OTDR curve and the second OTDR curve respectively;

[0012] Using the similarity comparison model, respectively perform similarity analysis on the previous OTDR curves of the same events in the two optical fibers to be identified, and determine whether the previous cable segments at the corresponding positions of the previous OTDR curves of the same events in the two optical fibers to be identified are the same cable;

[0013] Using the similarity comparison model, respectively perform similarity analysis on the subsequent OTDR curves of each identical event in the two optical fibers to be identified, and determine whether the subsequent cable segments at corresponding positions of the subsequent OTDR curves of each identical event in the two optical fibers to be identified are the same cable;

[0014] According to whether the preceding cable segment and the succeeding cable segment of each identical event are in the same cable, a partial same-cable optical fiber identification result of the two optical fibers to be identified is determined.

[0015] Further, the similarity comparison model is used to perform similarity analysis on the previous OTDR curves of each identical event in the two optical fibers to be identified, to determine whether the previous cable segment between each identical event and the previous event is the same cable, including:

[0016] Filling the shorter of the two previous OTDR curves of the same event of the two optical fibers to be identified with zeros from the front end or the rear end to the length of the other previous OTDR curve;

[0017] The two LSTM networks in the similarity comparison model are used to respectively extract the feature representations of the two prior OTDR curves after filling;

[0018] The characteristic representations of the two previous OTDR curves are input into the similarity comparison model for similarity analysis, and it is determined based on the output results whether the previous cable segments at corresponding positions of the previous OTDR curves of the same events in the two optical fibers to be identified are the same.

[0019] Preferably, the similarity comparison model building process includes:

[0020] Acquire a training data set, wherein each set of data in the training data set includes two OTDR curves and labels corresponding to similar cables;

[0021] Filling two OTDR curves of each set of data in the training data set to the same length;

[0022] Two LSTM networks with shared parameters in the twin neural network are used to extract the feature representations of the two filled OTDR curves respectively;

[0023] Calculate the distance between two feature representations using a predefined distance function;

[0024] The calculated distance is used as input, the corresponding label is used as output, and the loss function of the twin neural network is used for optimization training;

[0025] When the preset training termination condition is reached, the twin neural network is output as the similarity comparison model.

[0026] As a preferred solution, extracting event features of each event in the first OTDR curve and the second OTDR curve includes:

[0027] Identify different events in the first OTDR curve and the second OTDR curve according to type characteristics corresponding to preset event types;

[0028] The characteristic values ​​of different events of the first OTDR curve and the second OTDR curve are extracted as event features.

[0029] Furthermore, the characteristic values ​​include event type, event occurrence location, return loss and splicing loss.

[0030] Preferably, identifying the same event in the two optical fibers to be identified according to the extracted event features comprises:

[0031] Constructing event feature vectors according to event features of different events;

[0032] Calculating the event distance between each event on the first OTDR curve and each event on the second OTDR curve using a preset distance calculation model;

[0033] The corresponding identical events on the two optical fibers to be identified are identified according to the calculated event distances.

[0034] Further, using a preset distance calculation model to calculate the event distance between each event on the first OTDR curve and each event on the second OTDR curve includes:

[0035] Identify the same type of events with the same event type on the first OTDR curve and the second OTDR curve;

[0036] Calculating the distances between the characteristic values ​​in the event characteristic vectors of the same type of events on the first OTDR curve and the second OTDR curve;

[0037] The distance of each feature value is weighted according to the preset weight to obtain the event distance of the event.

[0038] As a preferred solution, identifying the corresponding identical events on the two optical fibers to be identified according to the calculated event distances includes:

[0039] When the event distance between an event on the first OTDR curve and an event on the second OTDR curve is within a preset first range, determining that the two corresponding events are the same event;

[0040] When the event distance between an event on the first OTDR curve and an event on the second OTDR curve is not within the first range, it is determined that the two corresponding events are not the same event.

[0041] The embodiment of the present invention further provides a device for identifying partially co-cable optical fibers, the device comprising:

[0042] A curve acquisition module, used for acquiring a first OTDR curve and a second OTDR curve of two optical fibers to be identified;

[0043] A feature extraction module, used to extract event features of each event in the first OTDR curve and the second OTDR curve;

[0044] An event identification module, used for identifying the same events in the two optical fibers to be identified according to the extracted event features;

[0045] The same-cable optical fiber identification module is used to perform similarity analysis on the OTDR curves of the cable sections where the same events in the two optical fibers to be identified are located according to a preset similarity comparison model, and output the same-cable optical fiber identification result.

[0046] Preferably, the same cable optical fiber identification module is specifically used for:

[0047] For each identical event, extracting a preceding OTDR curve between the identical event and a preceding event, and a succeeding OTDR curve between the identical event and a succeeding event from the first OTDR curve and the second OTDR curve respectively;

[0048] Using the similarity comparison model, respectively perform similarity analysis on the previous OTDR curves of the same events in the two optical fibers to be identified, and determine whether the previous cable segments at the corresponding positions of the previous OTDR curves of the same events in the two optical fibers to be identified are the same cable;

[0049] Using the similarity comparison model, respectively perform similarity analysis on the subsequent OTDR curves of each identical event in the two optical fibers to be identified, and determine whether the subsequent cable segments at corresponding positions of the subsequent OTDR curves of each identical event in the two optical fibers to be identified are the same cable;

[0050] According to whether the preceding cable segment and the succeeding cable segment of each identical event are in the same cable, a partial same-cable optical fiber identification result of the two optical fibers to be identified is determined.

[0051] Furthermore, the same cable optical fiber identification module is also specifically used for:

[0052] Filling the shorter of the two previous OTDR curves of the same event of the two optical fibers to be identified with zeros from the front end or the rear end to the length of the other previous OTDR curve;

[0053] The two LSTM networks in the similarity comparison model are used to respectively extract the feature representations of the two prior OTDR curves after filling;

[0054] The characteristic representations of the two previous OTDR curves are input into the similarity comparison model for similarity analysis, and it is determined based on the output results whether the previous cable segments at corresponding positions of the previous OTDR curves of the same events in the two optical fibers to be identified are the same.

[0055] Preferably, the process of constructing the similarity comparison model by the same-cable optical fiber identification module includes:

[0056] Acquire a training data set, wherein each set of data in the training data set includes two OTDR curves and labels corresponding to similar cables;

[0057] Filling two OTDR curves of each set of data in the training data set to the same length;

[0058] Two LSTM networks with shared parameters in the twin neural network are used to extract the feature representations of the two filled OTDR curves respectively;

[0059] Calculate the distance between two feature representations using a predefined distance function;

[0060] The calculated distance is used as input, the corresponding label is used as output, and the loss function of the twin neural network is used for optimization training;

[0061] When the preset training termination condition is reached, the twin neural network is output as the similarity comparison model.

[0062] Preferably, the feature extraction module is specifically used for:

[0063] Identify different events in the first OTDR curve and the second OTDR curve according to type characteristics corresponding to preset event types;

[0064] The characteristic values ​​of different events of the first OTDR curve and the second OTDR curve are extracted as event features.

[0065] Furthermore, the characteristic values ​​include event type, event occurrence location, return loss and splicing loss.

[0066] Preferably, the event identification module is specifically used for:

[0067] Constructing event feature vectors according to event features of different events;

[0068] Calculating the event distance between each event on the first OTDR curve and each event on the second OTDR curve using a preset distance calculation model;

[0069] The corresponding identical events on the two optical fibers to be identified are identified according to the calculated event distances.

[0070] Furthermore, the event identification module is also specifically used for:

[0071] Identify the same type of events with the same event type on the first OTDR curve and the second OTDR curve;

[0072] Calculating the distances between the characteristic values ​​in the event characteristic vectors of the same type of events on the first OTDR curve and the second OTDR curve;

[0073] The distance of each feature value is weighted according to the preset weight to obtain the event distance of the event.

[0074] Preferably, the event identification module is further specifically used for:

[0075] When the event distance between an event on the first OTDR curve and an event on the second OTDR curve is within a preset first range, determining that the two corresponding events are the same event;

[0076] When the event distance between an event on the first OTDR curve and an event on the second OTDR curve is not within the first range, it is determined that the two corresponding events are not the same event.

[0077] An embodiment of the present invention also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method for identifying partially co-cable optical fibers as described in any of the above embodiments is implemented.

[0078] An embodiment of the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the method for identifying partially co-cable optical fibers as described in any of the above embodiments.

[0079] An embodiment of the present invention further provides a computer program product, including a computer program / instruction, which implements the steps of any of the above methods when executed by a processor.

[0080] The present invention provides a method, device, equipment, medium and product for identifying part of the same cable optical fiber, by collecting the first OTDR curve and the second OTDR curve of two optical fibers to be identified; extracting the event features of each event in the first OTDR curve and the second OTDR curve; identifying the same event in the two optical fibers to be identified according to the extracted event features; performing similarity analysis on the OTDR curves of the cable segments where the same events in the two optical fibers to be identified are located according to a preset similarity comparison model, and outputting the same cable optical fiber identification result. The present application scheme is based on the identification of the same events in the OTDR curves of the two optical fibers to be identified, and performing similarity analysis on the cable segments where the same events are located, thereby reducing the blindness of the same cable optical fiber identification, being able to cope with the identification of part of the same cable optical fiber in complex scenarios, and being able to accurately and efficiently identify different parts of the same cable optical fiber. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 is a schematic diagram of a portion of co-cable optical fibers provided by an embodiment of the present invention in a first scenario;

[0082] Figure 2 is a schematic diagram of a portion of co-cable optical fibers provided by an embodiment of the present invention in a second scenario;

[0083] Figure 3 Schematic diagram of a portion of co-cable optical fibers provided by an embodiment of the present invention in a third scenario

[0084] Figure 4 It is a flow chart of a method for identifying partially co-cable optical fibers provided by an embodiment of the present invention;

[0085] Figure 5 It is a schematic diagram of an OTDR curve and a curve of a corresponding event provided by an embodiment of the present invention;

[0086] Figure 6 It is a schematic structural diagram of a device for identifying some optical fibers in the same cable provided by an embodiment of the present invention;

[0087] Figure 7 It is a structural diagram of a terminal device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0088] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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.

[0089] The difficulty in managing optical fiber dumb resources is due to the low degree of digital management, single resource monitoring methods, and insufficient intelligence. When a fault occurs, due to the lack of multi-parameter accurate perception and modeling analysis capabilities, slow-changing faults such as device aging and performance degradation are difficult to predict, and sudden faults such as optical fiber breakage are difficult to accurately locate. The optical network's "sensing, intelligence, and control" capabilities are insufficient, dumb resources are difficult to manage, and the reliability of the optical network is difficult to ensure.

[0090] According to the operators' operation and maintenance experience in the existing network, due to improper management of fiber dumb resources, the existing network fiber co-routing problems frequently occur. Operators have never stopped checking the co-routing and co-cable of the existing network in the past decade, and co-routing can easily lead to network failures. However, in some cases, due to manual asynchrony after service switching (fiber splicing / fiber jumper errors), co-routing risks are generated without knowing it, thus laying hidden dangers.

[0091] During the network construction and development, operators laid a large number of optical fibers. Over time, it became impossible to effectively verify whether the optical fibers belonged to the same cable. The existing solution can only identify the scenario where two optical fibers are completely in the same cable, and cannot identify optical fibers that are not in the same cable. The manual inspection method to sort out the optical cable route requires manual tapping and manual monitoring at the same time, and the identification efficiency and accuracy are low.

[0092] The existing methods for detecting and identifying optical fibers in the same cable only consider the case of completely in the same cable, that is, two optical fibers are completely in the same cable. In this scenario, the OTDR curves of the two optical fibers are usually compared for similarity to determine whether they are in the same cable.

[0093] In fact, there are a large number of partial co-cable scenarios in the existing network, that is, the optical fibers in the same optical cable will be fused to other optical cables due to network planning and other reasons.

[0094] See also Figure 1 , is a schematic diagram of a portion of co-cable optical fibers provided by an embodiment of the present invention in a first scenario, Figure 1 In the scenario, among the multiple optical fibers between site A and site B, a co-cable segment appears on the side close to site B. Figure 2 , is a schematic diagram of a portion of co-cable optical fibers provided by an embodiment of the present invention in a second scenario, Figure 2In the scenario, the optical fiber between site A and site B, and the optical fiber between site A and site C, have a co-cable section on the side close to site A. Figure 3 , is a schematic diagram of a portion of co-cable optical fibers provided by an embodiment of the present invention in a third scenario, Figure 3 In the scenario in, the optical fiber between site A and site B, and the optical fiber between site A and site C, have co-cable sections in the middle of the sites. It can be seen that the scenarios faced by the identification of co-cabled optical fibers are complex. It is not sufficient to only consider the situation of complete co-cabledness. It is necessary to accurately identify the co-cabled parts and non-co-cabled parts of partially co-cabled optical fibers. Different optical fibers form partial co-cables, and two optical fibers may have several different parts of optical fiber co-cabled. The partial co-cabled situation of optical fibers in different scenarios is very complex. The existing optical fiber co-cabled identification method cannot cope with the partial co-cabled optical fiber identification in complex scenarios, and the identification efficiency and identification accuracy cannot meet the requirements. In view of the above-mentioned defects, an embodiment of the present invention provides a method for identifying partially co-cabled optical fibers, see. Figure 4 , is a flow chart of a method for identifying partially co-cable optical fibers provided by an embodiment of the present invention, the method comprising the following steps:

[0095] Step S1, collecting a first OTDR curve and a second OTDR curve of two optical fibers to be identified;

[0096] Step S2, extracting event features of each event in the first OTDR curve and the second OTDR curve;

[0097] Step S3, identifying the same events in the two optical fibers to be identified according to the extracted event features;

[0098] Step S4, performing similarity analysis on the OTDR curves of the cable segments where the same event is located in the two optical fibers to be identified according to a preset similarity comparison model, and outputting the identification result of the same-cable optical fibers.

[0099] When this embodiment is specifically implemented, when identifying optical fibers in the same cable, it is necessary to obtain the OTDR curves of the two optical fibers to be identified. Specifically, a distributed sensing device is used to collect the OTDR curves of the optical fibers to be measured.

[0100] OTDR (Optical Time Domain Reflectometer) is a fiber optic instrument used for characterization, troubleshooting, and maintenance of optical communication networks. It contains a laser diode light source, an avalanche diode, and a highly accurate timing circuit (or time base).

[0101] When collecting OTDR curves, a pulse signal of a specific length is emitted into the optical fiber to be tested through the optical time domain reflectometer transmitter. The pulse signal propagates along the optical fiber and generates Rayleigh scattering. The injected light pulse is scattered in all directions at each point along the length of the optical fiber, and part of the light will return back to the detection unit of the OTDR. The OTDR will collect and measure this backscattered light. The backscattered signal will decrease exponentially as the optical fiber distance increases. By converting the backscattered signal into the dB exponential space, the OTDR sampling curve can be linearly and evenly decreased, so that the backscattered signal can be detected as the OTDR curve.

[0102] In this case, when performing the specific detection, the two optical fibers to be identified are respectively tested by OTDR curves, that is, a pulse signal is emitted to the first end of the first optical fiber to be identified by an optical time domain reflectometer, and the scattered light signal reflected back is received. According to the intensity of the scattered light, a first OTDR curve is generated, and a second OTDR curve can be obtained in the same way.

[0103] In the OTDR curve, when there is a joint, fusion loss or damage on the optical fiber link, the sampling signal curve at the corresponding position on the OTDR curve will show a nonlinear mutation. By identifying the nonlinear change of this signal, the optical fiber event can be automatically identified.

[0104] According to the statistical characteristics of the variability of different events, each event in the first OTDR curve and the second OTDR curve is identified, and the characteristics of each event are extracted.

[0105] By comparing and analyzing the event features of each event extracted from the first OTDR curve and the event features of each event extracted from the second OTDR curve, the same events in the two optical fibers to be identified are identified.

[0106] Specifically, the same event refers to a nonlinear mutation in the OTDR curves of the two optical cables to be identified due to the same or similar reasons. The specific reasons may include the appearance of joints, fusion loss or damage on the optical fiber link.

[0107] When two optical fibers are indistinguishably in the same cable, the events of the optical fiber parts in the same cable will be identical, that is, the optical fiber joints of the optical fiber parts in the same cable will appear at the same time, and the damage will also appear synchronously, so the same event will be identified on the OTDR curves of the two optical fibers to be identified. Even if there are no optical fiber joints and damage events in the optical fiber parts in the same cable, the optical fiber parts in the same cable of the two optical fibers to be identified cannot appear out of thin air, and some optical fibers in the same cable will definitely produce the fusion points and bifurcation points of the two optical fibers, so through the identification of the same events, it is possible to quickly search for the same optical fiber parts that may appear in the two optical fibers to be identified.

[0108] The identified identical events are used as the basis for partial search of the same-cable optical fibers, and similarity analysis is performed on the cable segments where the identical events are located in the first optical fiber to be identified and the second optical fiber to be identified, to determine whether the cable segments where the identical events are located are the same-cable optical fibers, to achieve partial identification of the same-cable optical fibers, and to output the same-cable optical fiber identification results of the two optical fibers to be identified.

[0109] It should be noted that a similarity comparison model is specifically used when performing similarity analysis. The similarity comparison model is an analysis model pre-trained for optical fibers with different similarities, and can perform similarity analysis on two different cable segments and identify the same cable segment.

[0110] This case is based on the identification of the same events in the OTDR curves of the two optical fibers to be identified, and the similarity analysis of the cable segments of the same events is performed to reduce the blindness of the identification of optical fibers in the same cable. It can cope with the identification of partial optical fibers in the same cable in complex scenarios, and can accurately and efficiently identify different parts of optical fibers in the same cable.

[0111] In another embodiment provided by the present invention, when the above step S3 is specifically performed, the following steps are specifically performed:

[0112] Identify the same events in the first OTDR curve and the second OTDE curve, and record the number of events in the first OTDR curve that are the same as those in the second OTDE curve as K, and the number of events in the second OTDR curve that are the same as those in the first OTDE curve as P.

[0113] It should be noted that the number of identical events in the first OTDR curve and the second OTDR curve may be different, that is, one event in the first OTDR curve may be identical to the corresponding two events in the second OTDR curve. In this case, the identical events recorded should record the event in the first OTDR curve and the event in the second OTDR curve that are identical to each other.

[0114] For the vth identical event, corresponding to the kth event in the first OTDR curve, the prior OTDR curve of the preceding cable segment between the kth event and the k-1th event is extracted from the first OTDR curve, and the subsequent OTDR curve of the subsequent cable segment between the kth event and the k+1th event is extracted from the first OTDR curve. For the identical event, which may correspond to the pth event in the second OTDR curve, the subsequent OTDR curve of the subsequent cable segment between the pth event and the p-1th event is extracted from the second OTDR curve, and the subsequent OTDR curve of the subsequent cable segment between the pth event and the p+1th event is extracted from the second OTDR curve.

[0115] It can be seen that when performing similar cable segment analysis, the cable segment between the same event and the previous event and / or the cable segment between the same event and the next event can be used as the cable segment for similarity analysis.

[0116] It should be noted that when k is 1, the corresponding preceding cable segment is the OTDR curve portion before the first event. When p is the last event on the first OTDR curve or the second OTDR curve, the corresponding succeeding cable segment is the OTDR curve portion after the last event.

[0117] The subsequent OTDR curves and the previous OTDR curves of the same events extracted from the first OTDR curve and the second OTDR curve correspond to each other. That is, the previous OTDR curve of the kth event in the first OTDR curve corresponds to the previous OTDR curve of the pth event in the second OTDR curve. When the similarity analysis is performed later, the two are the corresponding curves to be analyzed. That is, the subsequent OTDR curve of the kth event in the first OTDR curve corresponds to the subsequent OTDR curve of the pth event in the second OTDR curve. When the similarity analysis is performed later, the two are the corresponding curves to be analyzed.

[0118] It should be noted that if the kth event in the first OTDR curve has more than one identical event in the second OTDR curve, namely p1 to pn, p1 to pn are all recorded as corresponding curves to be analyzed, and the cable segment of each event in p1 to pn is respectively analyzed for similarity with the cable segment of the kth event.

[0119] For the vth identical event, a similarity comparison model is used to perform similarity analysis on the previous OTDR curve of the kth event in the first OTDR curve and the previous OTDR curve of the pth event in the second OTDR curve to determine whether the previous cable segment between each identical event and the previous event is the same cable; a similarity comparison model is used to perform similarity analysis on the subsequent OTDR curve of the kth event in the first OTDR curve and the subsequent OTDR curve of the pth event in the second OTDR curve to determine whether the previous cable segment between each identical event and the previous event is the same cable, thereby identifying whether partial common cable occurs in the cable segment where the vth identical event is located.

[0120] It should be noted that this case also uses the similarity comparison model to perform similarity analysis on the previous OTDR curve of each identical event in each optical fiber to be identified and the subsequent OTDR curve of the corresponding identical event in another optical cable to be identified, and outputs whether the optical cables of the identical events in different directions are in the same cable. That is, the similarity comparison model is used to perform similarity analysis on the previous OTDR curve of the kth event in the first OTDR curve and the subsequent OTDR curve of the pth event in the second OTDR curve to determine whether the previous cable segment of each identical event in a certain optical fiber is in the same cable as the subsequent cable segment in another optical fiber, thereby identifying whether the cable segment where the vth identical event is located has partial same cable. Through the identification of different-direction cable segments, it is possible to identify optical fibers that have some optical fibers in the same cable but with opposite transmission directions, thereby realizing the identification of different-direction but same-cable optical fibers.

[0121] During specific implementation, a traversal analysis is performed on each cable segment where the same event is located to identify all existing optical fiber segments in the same cable.

[0122] The cable segment for similarity analysis is determined through the same events, and then traversal similarity analysis is performed to achieve accurate and efficient identification of some optical fibers in the same cable.

[0123] In another embodiment provided by the present invention, when performing cable segment similarity analysis, the following steps are specifically performed:

[0124] For the searched and located part of the same cable segment, two time series of OTDR curves are intercepted, with lengths of N1 and N2 respectively.

[0125] The lengths of the preceding cable segments determined by the same event in the first OTDR curve and the second OTDR curve may not be the same, so the maximum length of the two time series is found, and the shorter sequence is padded with zeros to the same length.

[0126] The two filled OTDR curves are input into the feature representations extracted by the two LSTM networks respectively;

[0127] The characteristic representations of the two previous OTDR curves are input into the similarity comparison model to perform similarity analysis, and it is determined based on the output results whether the previous cable segment between each identical event and the previous event is the same cable.

[0128] It should be noted that, in this embodiment, only the analysis process of whether the preceding cable segment in the first OTDR curve and the preceding cable segment in the second OTDR curve are the same cable is provided. In other embodiments, the analysis process of whether the following cable segment in the first OTDR curve and the following cable segment in the second OTDR curve are the same cable is the same, only the input curves are different, and the recognition principle is the same.

[0129] It should be noted that, in this embodiment, when the similarity comparison model performs similarity analysis, it is specifically performed based on the extracted feature representation. During pre-training, training analysis is also performed based on different curves of corresponding labels to obtain the corresponding relationship between the input feature representation and the output recognition result.

[0130] For some of the same cable segments that have been searched and located, it is necessary to use the OTDR curve feature similarity comparison algorithm to further analyze the curve feature similarity. By filling the data to ensure that the data length is the same, the influence of the curve length difference during measurement on the similarity detection can be avoided, thus improving the accuracy of the similarity analysis.

[0131] In another embodiment provided by the present invention, the similarity comparison model building process specifically performs the following steps:

[0132] The method and process of using Siamese LSTM to compare the similarity of unequal length curves are as follows:

[0133] The input data preparation of the model requires two time series data (curves) of unequal length. The two time series obtained based on the partial co-cable segment search and positioning method are 'sequence1' and 'sequence2'. Each set of data in the training data set includes two OTDR curves and labels corresponding to similar co-cables. The labels can be binary labels (similar / dissimilar).

[0134] The Siamese network consists of two LSTM networks that share the same parameters. Each LSTM network processes a time series and compares the outputs to calculate similarity.

[0135] Create an LSTM network that receives a time series as input and extracts features through multiple LSTM layers and fully connected layers.

[0136] Create two LSTM networks with shared weights and define two inputs, corresponding to two time series.

[0137] The two inputs are passed to the LSTM network with shared weights respectively to obtain two feature representations.

[0138] Define a function to compute the Euclidean distance between two feature representations.

[0139] Compute the Euclidean distance between feature representations:

[0140] It should be noted that, in this embodiment, the Euclidean distance calculation equation is calculated as the distance function, and in other embodiments, other distance functions may be used to calculate the distance.

[0141] The above components are combined together to form a complete Siamese network model.

[0142] The model is compiled using mean squared error as the loss function and Adam as the optimizer.

[0143] The calculated distance is used as input, the corresponding label is used as output, and the loss function of the twin neural network is used for optimization training;

[0144] When the preset training termination condition is reached, the twin neural network is output as the similarity comparison model.

[0145] By using the twin neural network model to compare curve feature similarity, cables with similar feature representations can be accurately identified.

[0146] In another embodiment provided by the present invention, when performing event feature recognition, the following steps are specifically performed:

[0147] Identify different events in the first OTDR curve and the second OTDR curve. When performing event identification, identify different events in the curve according to the type characteristics corresponding to different event types. Figure 5 , is a schematic diagram of an OTDR curve and a corresponding event curve provided by an embodiment of the present invention.

[0148] In the OTDR curve, the horizontal and vertical axes are distance (km) and corresponding reflectivity (db), respectively. Different events can be identified by their type characteristics in the OTDR curve.

[0149] exist Figure 5 The events identified in the include positive splice loss events, negative splice loss events, and reflection events.

[0150] It should be noted that the same type of event may correspond to different type features in different sets, see Figure 2 In Figure 1, there are two different positive splice loss events.

[0151] It should be noted that event types also include connector events, break events, damage events, splice events, and fiber end events.

[0152] To facilitate subsequent feature comparison, different event types on the OTDR are mapped to corresponding numbers.

[0153] For different events, in order to analyze event-specific event features, characteristic values ​​of different events of the first OTDR curve and the second OTDR curve are extracted as event features.

[0154] Based on the recognition of different events and the extraction of feature values, accurate event features are extracted.

[0155] In another embodiment provided by the present invention, the identified event feature values ​​specifically include event type, event occurrence location, return loss, and splicing loss.

[0156] Specifically, event types are represented by different digital labels, and the same event is identified based on the identified different types of events. The same event must be the same type of event identified.

[0157] Based on the occurrence location of the event, the specific location of the time in the optical fiber can be located, and then the specific location of the optical fiber in the same cable can be determined according to the event.

[0158] Return loss and splice loss are loss values ​​calculated based on the OTDR curve when an event occurs, and are used to distinguish different events.

[0159] By reading different characteristic values ​​of different events, different events can be characterized to accurately distinguish the events.

[0160] In another embodiment of the present invention, when the same event is identified in step S3, the following steps are specifically performed:

[0161] A feature vector is constructed based on the feature value extracted from each event. For each identical event, the feature vector of the event corresponding to the first OTDR curve is extracted. Correspondingly, the feature vector of the event corresponding to the same event in the second OTDR curve is extracted.

[0162] For a specific event feature, the four selected feature values ​​are recorded as T, P, L, and S for the event type, event occurrence location, return loss, and splicing loss, respectively, to construct an event feature vector, where the event type is represented by a number.

[0163] The event distance between the event feature vector of the event on the first OTDR curve and the event feature vector of the event on the second OTDR curve is calculated using a preset distance calculation model.

[0164] Identical events are identified based on the calculated event distances.

[0165] In another embodiment provided by the present invention, the process of calculating the event distance specifically includes:

[0166] According to the event types in the event feature vectors on the first OTDR curve and the second OTDR curve, different types of events are identified and similar events are determined.

[0167] That is, an event on the first OTDR curve and an event of the same type on the second OTDR curve are marked as the same type of events.

[0168] It should be noted that a certain event on the first OTDR curve may have multiple events of the same type on the second OTDR curve, and each event of the same type on the second OTDR curve needs to be marked as a same type event.

[0169] The event distance is calculated based on the identified similar events, that is, the event distance between each event feature vector of the similar event on the first OTDR curve and each feature vector appearing on the second OTDR curve is calculated.

[0170] Calculate event distances for similar events, screen out the calculation of event distances between events of different event types, reduce the calculation of event distances, and improve calculation efficiency.

[0171] When calculating the characteristic distance, for each event characteristic vector of the same type of event on the first OTDR curve and each event characteristic vector appearing on the second OTDR curve, the distance between each corresponding characteristic value in the two vectors is calculated, and the different characteristic values ​​are weighted and summed using preset weights to obtain the event distance.

[0172] Specifically, the distance calculation equation between the eigenvalues ​​may adopt the distance calculation equation in the existing solution, for example, the distance between the eigenvalues ​​may be determined by using the Euclidean distance calculation method.

[0173] It should be noted that the scheme of screening by events of the same type in this case is a preferred scheme. In other embodiments, the distance of each event between all events on the first OTDR curve and the second OTDR curve may be calculated by traversal. Specifically:

[0174] Let i and j represent the i-th event on the first OTDR curve and the j-th event on the second OTDR curve respectively. Let the function D(·) represent the distance between the event feature values.

[0175] Dist ij =D T (T i ,T j )·[δ P D P (P i ,P j )+δ L D L (L i ,L j )+δ S D S (S i ,S j )].

[0176] Among them, D T (·) is used to measure T i With Tj Whether the values are equal, and record the four selected eigenvalues: the event type, the event occurrence location, the return loss, and the fusion loss as T, P, L, and S respectively. For example, D T (·) can be an exclusive-NOR operation or a function to determine whether the difference between T i and T j is 0. D P (·), D L (·), D S (·) is used to calculate the distance between two optical fibers, such as the Euclidean distance. δ P , δ L , δ S is used to normalize the weights of the three features, δ P +δ L +δ S = 1.

[0177] By using different event types as coefficients of the event distance, the effect of screening the same type of events can also be achieved.

[0178] By using the event type as a screening condition for screening the same events, the calculation amount can be reduced.

[0179] In another embodiment provided by the present invention, when identifying the same event by the event distance, when the event distance between an event on the first OTDR curve and an event on the second OTDR curve is within a preset first range, it is determined that the two corresponding events are the same event;

[0180] If Dist ij < r, it is determined that the event features of the i-th event of the first OTDR curve and the j-th event of the second OTDR curve being compared are the same, and it is determined that the two events are the same event.

[0181] If Dist ij ≥ r, it is determined that the event features of the i-th event of the first OTDR curve and the j-th event of the second OTDR curve being compared are different, and it is determined that the two corresponding events are not the same event.

[0182] Among them, r can be obtained through the empirical values of multiple calculations.

[0183] By setting the range of the event distance, the identification of the same event in different events is realized.

[0184] See Figure 6 , which is a schematic structural diagram of an identification device for partially co-cabled optical fibers provided by an embodiment of the present invention. The device includes:

[0185] A curve acquisition module, configured to acquire a first OTDR curve and a second OTDR curve of the two optical fibers to be identified;

[0186] A feature extraction module, used to extract event features of each event in the first OTDR curve and the second OTDR curve;

[0187] An event identification module, used for identifying the same events in the two optical fibers to be identified according to the extracted event features;

[0188] The same-cable optical fiber identification module is used to perform similarity analysis on the OTDR curves of the cable sections where the same events in the two optical fibers to be identified are located according to a preset similarity comparison model, and output the same-cable optical fiber identification result.

[0189] The device for identifying partially co-cabled optical fibers provided in this embodiment can execute all steps and functions of the method for identifying partially co-cabled optical fibers provided in any of the above embodiments, and the specific functions of the device will not be described in detail herein.

[0190] See also Figure 7 , is a schematic diagram of the structure of a terminal device provided by an embodiment of the present invention. The terminal device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a program for identifying partially co-cabled optical fibers. When the processor executes the computer program, the steps of each of the above-mentioned embodiments of the method for identifying partially co-cabled optical fibers are implemented, such as Figure 2 Alternatively, the processor implements the functions of each module in the above-mentioned device embodiments when executing the computer program.

[0191] Exemplarily, the computer program may be divided into one or more modules, and the one or more modules are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments that can complete specific functions, and the instruction segments are used to describe the execution process of the computer program in the device for identifying partially co-cabled optical fibers. For example, the computer program may be divided into several modules, and the specific functions of each module have been described in detail in the method for identifying partially co-cabled optical fibers provided in any of the above embodiments, and the specific functions of the device will not be repeated here.

[0192] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art may understand that the schematic diagram is only an example of a terminal device and does not constitute a limitation on a device for identifying partially co-cabled optical fibers. The device may include more or fewer components than shown in the diagram, or may combine certain components, or different components. For example, the terminal device may also include input and output devices, network access devices, buses, etc.

[0193] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the device for identifying partially co-cabled optical fibers, and uses various interfaces and lines to connect the various parts of the device for identifying partially co-cabled optical fibers.

[0194] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the device for identifying partially co-cable optical fibers by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0195] Wherein, if the module integrated with the identification device of a part of the same cable optical fiber is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0196] An embodiment of the present invention further provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of implementing the method by the functional network element in the above embodiment.

[0197] The computer program product provided in this embodiment can execute all the steps and functions of the method for identifying some co-cable optical fibers provided in any of the above embodiments, and the specific functions of the product will not be described in detail here.

[0198] It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for identifying partially co-cable optical fibers, characterized in that: The method comprises: Collecting a first OTDR curve and a second OTDR curve of two optical fibers to be identified; Extracting event features of each event in the first OTDR curve and the second OTDR curve; Identify the same events in the two optical fibers to be identified according to the extracted event features; The OTDR curves of the cable segments where the same events are located in the two optical fibers to be identified are analyzed for similarity according to a preset similarity comparison model, and the identification results of the same-cable optical fibers are output.

2. The method for identifying partially co-cabled optical fibers according to claim 1, characterized in that: According to a preset similarity comparison model, similarity analysis is performed on the OTDR curves of the cable segments where the same event is located in the two optical fibers to be identified, and the same-cable optical fiber identification results are output, including: For each identical event, extracting a preceding OTDR curve between the identical event and a preceding event, and a succeeding OTDR curve between the identical event and a succeeding event from the first OTDR curve and the second OTDR curve respectively; Using the similarity comparison model, respectively perform similarity analysis on the previous OTDR curves of the same events in the two optical fibers to be identified, and determine whether the previous cable segments at the corresponding positions of the previous OTDR curves of the same events in the two optical fibers to be identified are the same cable; Using the similarity comparison model, respectively perform similarity analysis on the subsequent OTDR curves of each identical event in the two optical fibers to be identified, and determine whether the subsequent cable segments at corresponding positions of the subsequent OTDR curves of each identical event in the two optical fibers to be identified are identical; According to whether the preceding cable segment and the succeeding cable segment of each identical event are in the same cable, a partial same-cable optical fiber identification result of the two optical fibers to be identified is determined.

3. The method for identifying partially co-cabled optical fibers according to claim 2, characterized in that: The similarity comparison model is used to perform similarity analysis on the previous OTDR curves of the same events in the two optical fibers to be identified, respectively, to determine whether the previous cable segments at the corresponding positions of the previous OTDR curves of the same events in the two optical fibers to be identified are the same cable, including: Filling the shorter of the two previous OTDR curves of the same event of the two optical fibers to be identified with zeros from the front end or the rear end to the length of the other previous OTDR curve; The two LSTM networks in the similarity comparison model are used to respectively extract the feature representations of the two prior OTDR curves after filling; The characteristic representations of the two previous OTDR curves are input into the similarity comparison model for similarity analysis, and it is determined based on the output results whether the previous cable segments at corresponding positions of the previous OTDR curves of the same events in the two optical fibers to be identified are the same.

4. The method for identifying partially co-cabled optical fibers according to claim 1, characterized in that: The similarity comparison model building process includes: Acquire a training data set, wherein each set of data in the training data set includes two OTDR curves and labels corresponding to similar cables; Filling two OTDR curves of each set of data in the training data set to the same length; Two LSTM networks with shared parameters in the twin neural network are used to extract the feature representations of the two filled OTDR curves respectively; Calculate the distance between two feature representations using a predefined distance function; The calculated distance is used as input, the corresponding label is used as output, and the loss function of the twin neural network is used for optimization training; When the preset training termination condition is reached, the twin neural network is output as the similarity comparison model.

5. The method for identifying partially co-cabled optical fibers according to claim 1, characterized in that: Extracting event features of each event in the first OTDR curve and the second OTDR curve includes: Identify different events in the first OTDR curve and the second OTDR curve according to type characteristics corresponding to preset event types; The characteristic values ​​of different events of the first OTDR curve and the second OTDR curve are extracted as event features.

6. The method for identifying partially co-cabled optical fibers according to claim 5, characterized in that: The characteristic values ​​include event type, event occurrence location, return loss and splicing loss.

7. The method for identifying partially co-cabled optical fibers according to claim 1, characterized in that: Identifying the same events in the two optical fibers to be identified according to the extracted event features includes: Constructing event feature vectors according to event features of different events; Calculating the event distance between each event on the first OTDR curve and each event on the second OTDR curve using a preset distance calculation model; The corresponding identical events on the two optical fibers to be identified are identified according to the calculated event distances.

8. The method for identifying partially co-cabled optical fibers according to claim 7, characterized in that: Calculating the event distance between each event on the first OTDR curve and each event on the second OTDR curve using a preset distance calculation model includes: Identify the same type of events with the same event type on the first OTDR curve and the second OTDR curve; Calculating the distances between the characteristic values ​​in the event characteristic vectors of the same type of events on the first OTDR curve and the second OTDR curve; The distance of each feature value is weighted according to the preset weight to obtain the event distance of the event.

9. The method for identifying partially co-cabled optical fibers according to claim 7, characterized in that: Identifying corresponding identical events on the two optical fibers to be identified according to the calculated event distances, comprising: When the event distance between an event on the first OTDR curve and an event on the second OTDR curve is within a preset first range, determining that the two corresponding events are the same event; When the event distance between an event on the first OTDR curve and an event on the second OTDR curve is not within the first range, it is determined that the two corresponding events are not the same event.

10. A device for identifying partially co-cabled optical fibers, characterized in that: The device comprises: A curve acquisition module, used for acquiring a first OTDR curve and a second OTDR curve of two optical fibers to be identified; A feature extraction module, used to extract event features of each event in the first OTDR curve and the second OTDR curve; An event identification module, used for identifying the same events in the two optical fibers to be identified according to the extracted event features; The same-cable optical fiber identification module is used to perform similarity analysis on the OTDR curves of the cable sections where the same events in the two optical fibers to be identified are located according to a preset similarity comparison model, and output the same-cable optical fiber identification result.

11. A terminal device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the method for identifying partially co-cabled optical fibers as claimed in any one of claims 1 to 9 when executing the computer program.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method for identifying partially co-cabled optical fibers according to any one of claims 1 to 9.

13. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of any method described in claims 1 to 9 are implemented.

Images